swf-spec-19.txt

SWF FILE FORMAT SPECIFICATION
VERSION 19

Copyright © 2006-2012 Adobe Systems Incorporated. All rights reserved. This manual may not be copied,
photocopied, reproduced, translated, or converted to any electronic or machine-readable form in whole or in
part without written approval from Adobe Systems Incorporated. Notwithstanding the foregoing, a person
obtaining an electronic version of this manual from Adobe may print out one copy of this manual provided that
no part of this manual may be printed out, reproduced, distributed, resold, or transmitted for any other
purposes, including, without limitation, commercial purposes, such as selling copies of this documentation or
providing paid-for support services.
Trademarks
Adobe, ActionScript, Flash, Flash Media Server, Adobe Media Server, Flash Player, PostScript, and XMP are either
registered trademarks or trademarks of Adobe Systems Incorporated and may be registered in the United States
or in other jurisdictions including internationally. Other product names, logos, designs, titles, words, or phrases
mentioned within this publication may be trademarks, service marks, or trade names of Adobe Systems
Incorporated or other entities and may be registered in certain jurisdictions including internationally. No right or
license is granted to any Adobe trademark.
Third-Party Information
This guide contains links to third-party websites that are not under the control of Adobe Systems Incorporated,
and Adobe Systems Incorporated is not responsible for the content on any linked site. If you access a third-party
website mentioned in this guide, then you do so at your own risk. Adobe Systems Incorporated provides these
links only as a convenience, and the inclusion of the link does not imply that Adobe Systems Incorporated
endorses or accepts any responsibility for the content on those third- party sites. No right, license or interest is
granted in any third party technology referenced in this guide.
NOTICE: THIS PUBLICATION AND THE INFORMATION HEREIN IS FURNISHED “AS IS”, IS SUBJECT TO CHANGE
WITHOUT NOTICE, AND SHOULD NOT BE CONSTRUED AS A COMMITMENT BY ADOBE SYSTEMS INCORPORATED.
ADOBE SYSTEMS INCORPORATED ASSUMES NO RESPONSIBILITY OR LIABILITY FOR ANY ERRORS OR
INACCURACIES, MAKES NO WARRANTY OF ANY KIND (EXPRESS, IMPLIED, OR STATUTORY) WITH RESPECT TO
THIS PUBLICATION, AND EXPRESSLY DISCLAIMS ANY AND ALL WARRANTIES OF MERCHANTABILITY, FITNESS FOR
PARTICULAR PURPOSES, AND NONINFRINGEMENT OF THIRD PARTY RIGHTS.
Adobe Systems Incorporated - Published 2012

Contents
Introduction ............................................................................................................................................................. 12
Chapter 1: Basic Data Types .................................................................................................................................... 14
Coordinates and twips ......................................................................................................................................... 14
Integer types and byte order............................................................................................................................... 14
Fixed-point numbers ........................................................................................................................................... 15
Floating-point numbers ....................................................................................................................................... 16
Encoded integers ................................................................................................................................................. 16
Bit values ............................................................................................................................................................. 17
Using bit values.................................................................................................................................................... 18
String values ........................................................................................................................................................ 19
Language code ..................................................................................................................................................... 20
RGB color record ................................................................................................................................................. 21
RGBA color with alpha record ............................................................................................................................. 21
ARGB color with alpha record ............................................................................................................................. 21
Rectangle record ................................................................................................................................................. 22
MATRIX record..................................................................................................................................................... 22
Color transform record........................................................................................................................................ 24
Color transform with alpha record ...................................................................................................................... 25
Chapter 2: SWF Structure Summary ................................................................................................................. 27
The SWF header .................................................................................................................................................. 27
SWF file structure ................................................................................................................................................ 28
Tag format ........................................................................................................................................................... 29
Definition and control tags .................................................................................................................................. 29
Tag ordering in SWF files ..................................................................................................................................... 30

The dictionary ...................................................................................................................................................... 30
Processing a SWF file ........................................................................................................................................... 32
File compression strategy.................................................................................................................................... 32
Summary.............................................................................................................................................................. 32
Chapter 3: The Display List ...................................................................................................................................... 33
Clipping layers ..................................................................................................................................................... 33
Using the display list ............................................................................................................................................ 33
Display list tags .................................................................................................................................................... 34
PlaceObject ...................................................................................................................................................... 34
PlaceObject2 .................................................................................................................................................... 35
PlaceObject3 .................................................................................................................................................... 38
Color Matrix filter ............................................................................................................................................ 42
Convolution filter............................................................................................................................................. 42
Blur filter .......................................................................................................................................................... 44
Drop Shadow filter........................................................................................................................................... 45
Glow filter ........................................................................................................................................................ 46
Bevel filter ....................................................................................................................................................... 46
Gradient Glow and Gradient Bevel filters ....................................................................................................... 47
ClipEventFlags .................................................................................................................................................. 48
RemoveObject ................................................................................................................................................. 50
RemoveObject2 ............................................................................................................................................... 50
ShowFrame ...................................................................................................................................................... 51
Chapter 4: Control Tags ........................................................................................................................................... 52
SetBackgroundColor ............................................................................................................................................ 52
FrameLabel .......................................................................................................................................................... 52

Protect ................................................................................................................................................................. 53
End ....................................................................................................................................................................... 53
ExportAssets ........................................................................................................................................................ 53
ImportAssets........................................................................................................................................................ 54
EnableDebugger .................................................................................................................................................. 55
EnableDebugger2 ................................................................................................................................................ 55
ScriptLimits .......................................................................................................................................................... 56
SetTabIndex ......................................................................................................................................................... 56
FileAttributes ....................................................................................................................................................... 57
ImportAssets2...................................................................................................................................................... 58
SymbolClass ......................................................................................................................................................... 59
Metadata ............................................................................................................................................................. 59
DefineScalingGrid ................................................................................................................................................ 60
DefineSceneAndFrameLabelData ........................................................................................................................ 62
Chapter 5: Actions ................................................................................................................................................... 63
SWF 3 action model ............................................................................................................................................. 63
SWF 3 actions .................................................................................................................................................. 64
SWF 4 action model ............................................................................................................................................. 67
The program counter....................................................................................................................................... 68
SWF 4 actions .................................................................................................................................................. 68
SWF 5 action model ............................................................................................................................................. 89
SWF 5 actions .................................................................................................................................................. 89
ScriptObject actions......................................................................................................................................... 90
Type actions ................................................................................................................................................... 100
Math actions .................................................................................................................................................. 102

Stack operator actions ................................................................................................................................... 103
SWF 6 action model ........................................................................................................................................... 108
SWF 6 actions ................................................................................................................................................ 108
SWF 7 action model ........................................................................................................................................... 111
SWF 7 actions ................................................................................................................................................ 111
SWF 9 action model ........................................................................................................................................... 117
DoABC ............................................................................................................................................................ 117
SWF 10 action model ......................................................................................................................................... 118
Chapter 6: Shapes.................................................................................................................................................. 119
Shape overview ................................................................................................................................................. 119
Shape example .................................................................................................................................................. 120
Shape structures ................................................................................................................................................ 121
Fill styles ........................................................................................................................................................ 121
Line styles ...................................................................................................................................................... 123
Shape structures ............................................................................................................................................ 125
Shape tags...................................................................................................................................................... 131
Chapter 7: Gradients ............................................................................................................................................. 134
Gradient transformations .................................................................................................................................. 134
Gradient control points ..................................................................................................................................... 135
Gradient structures ........................................................................................................................................... 135
GRADIENT ...................................................................................................................................................... 135
FOCALGRADIENT............................................................................................................................................ 136
GRADRECORD ................................................................................................................................................ 136
Chapter 8: Bitmaps ................................................................................................................................................ 137
DefineBits .......................................................................................................................................................... 137

JPEGTables ......................................................................................................................................................... 137
DefineBitsJPEG2................................................................................................................................................. 138
DefineBitsJPEG3................................................................................................................................................. 139
DefineBitsLossless ............................................................................................................................................. 139
DefineBitsLossless2 ........................................................................................................................................... 142
DefineBitsJPEG4................................................................................................................................................. 143
Chapter 9: Shape Morphing .................................................................................................................................. 144
DefineMorphShape ........................................................................................................................................... 145
DefineMorphShape2 ......................................................................................................................................... 146
Morph fill styles ................................................................................................................................................. 148
MORPHFILLSTYLEARRAY ................................................................................................................................ 148
MORPHFILLSTYLE ........................................................................................................................................... 148
Morph gradient values ...................................................................................................................................... 149
MORPHGRADIENT ......................................................................................................................................... 149
MORPHGRADRECORD ................................................................................................................................... 149
Morph line styles ............................................................................................................................................... 150
MORPHLINESTYLEARRAY ............................................................................................................................... 150
MORPHLINESTYLE .......................................................................................................................................... 150
MORPHLINESTYLE2 ........................................................................................................................................ 150
Chapter 10: Fonts and Text ................................................................................................................................... 152
Glyph text and device text................................................................................................................................. 152
Static text and dynamic text .............................................................................................................................. 152
Glyph text .......................................................................................................................................................... 153
Glyph definitions ........................................................................................................................................... 153
The EM square ............................................................................................................................................... 154

Converting TrueType fonts to SWF glyphs .................................................................................................... 154
Kerning and advance values .......................................................................................................................... 155
Advanced text rendering engine ................................................................................................................... 155
DefineFont and DefineText ........................................................................................................................... 156
Static glyph text example .............................................................................................................................. 157
Font tags ............................................................................................................................................................ 158
DefineFont ..................................................................................................................................................... 158
DefineFontInfo............................................................................................................................................... 158
Western indirect fonts................................................................................................................................... 160
Japanese indirect fonts .................................................................................................................................. 160
DefineFontInfo2............................................................................................................................................. 161
DefineFont2 ................................................................................................................................................... 162
DefineFont3 ................................................................................................................................................... 164
DefineFontAlignZones ................................................................................................................................... 166
Kerning record ............................................................................................................................................... 167
DefineFontName ........................................................................................................................................... 168
Static text tags ................................................................................................................................................... 168
DefineText ..................................................................................................................................................... 168
Text records ................................................................................................................................................... 169
Dynamic text tags .............................................................................................................................................. 171
DefineEditText ............................................................................................................................................... 171
CSMTextSettings ............................................................................................................................................ 174
DefineFont4 ................................................................................................................................................... 175
Chapter 11: Sounds ............................................................................................................................................... 177
Audio coding formats ........................................................................................................................................ 177

Event sounds ..................................................................................................................................................... 178
DefineSound .................................................................................................................................................. 178
StartSound ..................................................................................................................................................... 179
StartSound2 ................................................................................................................................................... 180
Sound styles ................................................................................................................................................... 180
Streaming sound................................................................................................................................................ 181
SoundStreamHead......................................................................................................................................... 181
SoundStreamHead2....................................................................................................................................... 183
SoundStreamBlock ........................................................................................................................................ 184
Frame subdivision for streaming sound ........................................................................................................ 184
ADPCM compression ......................................................................................................................................... 186
ADPCM sound data........................................................................................................................................ 187
MP3 compression .............................................................................................................................................. 188
MP3 sound data............................................................................................................................................. 188
MP3 frame ..................................................................................................................................................... 190
Nellymoser compression ................................................................................................................................... 192
Speex compression ............................................................................................................................................ 192
Chapter 12: Buttons .............................................................................................................................................. 193
Button states ..................................................................................................................................................... 193
Button tracking .................................................................................................................................................. 193
Events, state transitions, and actions................................................................................................................ 194
Button tags ...................................................................................................................................................... 195
Buttonrecord ................................................................................................................................................. 195
DefineButton ................................................................................................................................................. 196
DefineButton2 ............................................................................................................................................... 197

DefineButtonCxform ..................................................................................................................................... 199
DefineButtonSound ....................................................................................................................................... 199
Chapter 13: Sprites and Movie Clips ..................................................................................................................... 201
Sprite names ...................................................................................................................................................... 201
DefineSprite ....................................................................................................................................................... 202
Chapter 14: Video .................................................................................................................................................. 204
Sorenson H.263 Bitstream Format .................................................................................................................... 204
Summary of differences from H.263 ............................................................................................................. 204
Video packet .................................................................................................................................................. 205
Macro block ................................................................................................................................................... 206
Block data ...................................................................................................................................................... 207
Screen Video bitstream format ......................................................................................................................... 208
Block format .................................................................................................................................................. 208
Video packet .................................................................................................................................................. 209
Image block.................................................................................................................................................... 210
Screen Video V2 bitstream format .................................................................................................................... 210
V2 Colorspace ................................................................................................................................................ 210
V2 Video Packet ............................................................................................................................................. 211
Image Block V2 .............................................................................................................................................. 212
Image format ................................................................................................................................................. 212
Image block diff position ............................................................................................................................... 213
Image block prime position ........................................................................................................................... 213
On2 Truemotion VP6 bitstream format ............................................................................................................ 213
VP6 FLV video packet .................................................................................................................................... 215
VP6 FLV Alpha video packet .......................................................................................................................... 215

VP6 SWF video packet ................................................................................................................................... 216
VP6 SWF Alpha video packet ......................................................................................................................... 216
SWF video tags .................................................................................................................................................. 216
DefineVideoStream ....................................................................................................................................... 217
VideoFrame ................................................................................................................................................... 217
Chapter 15: Metadata........................................................................................................................................... 219
FileAttributes ..................................................................................................................................................... 219
EnableTelemetry ............................................................................................................................................... 220
DefineBinaryData .............................................................................................................................................. 220
Appendix A: SWF Uncovered: A Simple SWF File Dissected ................................................................................. 221
Appendix B: Reverse index of tag values............................................................................................................... 235
Appendix C: Screen Video v2 Palette .................................................................................................................... 238

Introduction
The SWF (pronounced “swiff ”) file format delivers vector graphics, text, video, and sound over the Internet and
is supported by Adobe® Flash® Player software. The SWF file format is designed to be an efficient delivery
format, not a format for exchanging graphics between graphics editors. It is designed to meet the following
goals:
•

On-screen display—The format is primarily intended for on-screen display and supports anti-aliasing,
fast rendering to a bitmap of any color format, animation, and interactive buttons.

•

Extensibility—The format is a tagged format, so it can be evolved with new features while maintaining
backward compatibility with earlier versions of Flash Player.

•

Network delivery—The format can travel over a network with limited and unpredictable bandwidth. The
files are compressed to be small and support incremental rendering through streaming. The SWF file
format is a binary format and is not human readable like HTML. The SWF file format uses techniques
such as bit-packing and structures with optional fields to minimize file size.

•

Simplicity—The format is simple so that Flash Player is small and easily ported. Also, Flash Player
depends upon a limited set of operating system features only.

•

File independence—The files display with minimal dependence on external resources such as fonts.

•

Scalability—The files work well on limited hardware, and can take advantage of better hardware when it
is available. This ability is important because computers have different monitor resolutions and bit
depths.

•

Speed—The graphics described by SWF files render quickly.

•

Scriptability—The format includes tags that provide sequences of byte codes to be interpreted by a
stack machine. The byte codes support the ActionScript® language. Flash Player provides a runtime
ActionScript object model that allows interaction with drawing primitives, servers, and features of Flash
Player.

SWF files have the extension .swf and a MIME type of application/x-shockwave-flash.
The SWF format has evolved through several versions. Through SWF 5, substantial additions were made to the
SWF tag set. Starting with SWF 6 and later, the SWF format changes less, as more new features are implemented
partly or entirely at the ActionScript level. Starting with SWF 9, the ActionScript 3.0 language, which employs the
new ActionScript Virtual Machine 2 (AVM2) can be used. Anyone planning to generate SWF file content that
uses newer features should become familiar with the ActionScript object model that Flash Player exposes. Some
references for this information are in the ActionScript Developer's Guide (see
http://help.adobe.com/en_US/as3/dev/index.html), ActionScript 3.0 Language Reference (see
http://help.adobe.com/en_US/FlashPlatform/reference/actionscript/3/), and the Adobe ActionScript Virtual
12

Machine 2 Overview (PDF file) at www.adobe.com/go/avm2overview.
If you need to knows the mapping between Flash Player / AIR names to SWF version (for example, 11.4 is SWF
17), look up the features in a given release, or check out the latest in what’s new in the Flash Player and SWF
features, visit: Flash Player and Air Feature List.
Adobe seriously considers all feedback to the SWF file format specification. E-mail any unclear or potentially
erroneous information within the specification to Adobe at flashformat@adobe.com. All such email submissions
shall be subject to the Submitted Materials guidelines in the Terms of Use at
www.adobe.com/misc/copyright.html.
Summary of post-ship changes:
•

13

April 23, 2013 – Fixed wrong order in ActionStackSwap

Chapter 1: Basic Data Types
This section describes the basic data types that make up the more complex data structures in the SWF file
format. All other structures in the SWF file format are built on these fundamental types.

Coordinates and twips
The SWF file format stores all x-y coordinates as integers, usually in a unit of measurement called a twip. In the
SWF format, a twip is 1/20th of a logical pixel. A logical pixel is the same as a screen pixel when the file is played
at 100%—that is, without scaling.
For example, a rectangle 800 twips wide by 400 twips high is rendered as 40 by 20 logical pixels. Fractional pixel
sizes are approximated with anti-aliasing. A rectangle 790 by 390 twips (39.5 by 19.5 pixels) appears to have
slightly blurred edges.
Twips are a good compromise between size and precision. They provide sub-pixel accuracy for zooming and
precise placement of objects, while consuming very few bits per coordinate.
Coordinates in the SWF file format use the traditional graphics axes: x is horizontal and proceeds from minimum
values at the left to maximum values at the right, and y is vertical and proceeds from minimum values at the top
to maximum values at the bottom.

Integer types and byte order
The SWF file format uses 8-bit, 16-bit, 32-bit, 64-bit, signed, and unsigned integer types. All integer values are
stored in the SWF file by using little-endian byte order: the least significant byte is stored first, and the most
significant byte is stored last, in the same way as the Intel x86 architecture. The bit order within bytes in the
SWF file format is big-endian: the most significant bit is stored first, and the least significant bit is stored last.
For example:
•
•

The 32-bit value 0x456e7120 is stored as 20 71
The 16-bit value 0xe712 is stored as 12 e7.

6e 45.

All integer types must be byte-aligned. That is, the first bit of an integer value must be stored in the first bit of a
byte in the SWF file.
Signed integers are represented by using traditional 2’s-complement bit patterns. These are the signed integer
representations used on most modern computer platforms. In the 2’s complement system, negative numbers
have 1 as the first bit, and zero and positive numbers have 0 as the first bit. A negative number, -n, is
represented as the bitwise opposite of the positive-zero number n-1.

14

Integer Types
Type

Comment

SI8

Signed 8-bit integer value

SI16

Signed 16-bit integer value

SI32

Signed 32-bit integer value

SI8[n]

Signed 8-bit array—n is the number of array elements

SI16[n]

Signed 16-bit array—n is the is number of array elements

UI8

Unsigned 8-bit integer value

UI16

Unsigned 16-bit integer value

UI32

Unsigned 32-bit integer value

UI8[n]

Unsigned 8-bit array—n is the number of array elements

UI16[n]

Unsigned 16-bit array—n is the number of array elements

UI24[n]

Unsigned 24-bit array—n is the number of array elements

UI32[n]

Unsigned 32-bit array—n is the number of array elements

UI64[n]

Unsigned 64-bit array—n is the number of array elements

Fixed-point numbers
The SWF file format supports two types of fixed-point numbers: 32-bit and 16-bit.
The 32-bit fixed-point numbers are 16.16. That is, the high 16 bits represent the number before the decimal
point, and the low 16 bits represent the number after the decimal point. FIXED values are stored like 32-bit
integers in the SWF file (using little-endian byte order) and must be byte aligned.
For example: The real value 7.5 is equivalent to: 0x0007.8000. This value is stored in the SWF file as: 00 80 07
00.
SWF 8 and later supports 16-bit 8.8 signed, fixed-point numbers. The high 8 bits represent the number before
the decimal point, and the low 8 bits represent the number after the decimal point. FIXED8 values are stored like
16-bit integers in the SWF file (using little- endian byte order) and must be byte aligned.

15

Fixed-Point Types
Type

Comment

FIXED

32-bit 16.16 fixed-point number

FIXED8

16-bit 8.8 fixed-point number

Floating-point numbers
SWF 8 and later supports the use of IEEE Standard 754 compatible floating-point types. Three types of floatingpoint numbers are supported.
Floating-Point Types
Type

Comment

FLOAT16

Half-precision (16-bit) floating-point number

FLOAT

Single-precision (32-bit) IEEE Standard 754 compatible

DOUBLE

Double-precision (64-bit) IEEE Standard 754 compatible

FLOAT16 is identical to the characteristics of FLOAT except for changes to the number of bits allocated to the
exponent and mantissa:
•
•
•

1 bit for the sign
5 bits for the exponent, with an exponent bias of 16
10 bits for the mantissa

Encoded integers
SWF 9 and later supports the use of integers encoded with a variable number of bytes. One type of encoded
integer is supported.
Floating-Point Types
Type

Comment

EncodedU32

Variable length encoded 32-bit unsigned integer

This is a 32-bit unsigned integer value encoded with a variable number of bytes to save space. All EncodedU32's
are encoded as 1-5 bytes depending on the value (larger values need more space). The encoding method is if the
hi bit in the current byte is set, then the next byte is also part of the value. Each bit in a byte contributes 7 bits to
the value, with the hi bit telling us whether to use the next byte, or if this is the last byte for the value.
16

This is the algorithm for parsing an EncodedU32:
int GetEncodedU32(unsigned char*& pos)
{
int result = pos[0];
if (!(result & 0x00000080))
{
pos++;
return result;
}
result = (result & 0x0000007f) | pos[1]<<7;
if (!(result & 0x00004000))
{
pos += 2;
return result;
}
result = (result & 0x00003fff) | pos[2]<<14;
if (!(result & 0x00200000))
{
pos += 3;
return result;
}
result = (result & 0x001fffff) | pos[3]<<21;
if (!(result & 0x10000000))
{
pos += 4;
return result;
}
result = (result & 0x0fffffff) | pos[4]<<28;
pos += 5;
return result;
}

Bit values
Bit values are variable-length bit fields that can represent three types of numbers:
1. Unsigned integers
2. Signed integers
3. Signed 16.16 fixed-point values.
Bit values do not have to be byte aligned. Other types (such as UI8 and UI16) are always byte aligned. If a bytealigned type follows a bit value, the last byte that contains the bit value is padded with zeros.
The following example is a stream of 64 bits. The 64 bits represent 9 values of varying bit length, followed by a
UI16 value:

The bit stream begins with a 6-bit value (BV1), followed by a 5-bit value (BV2) that is spread across Byte1 and
17

Byte2. BV3 is spread across Byte2 and Byte3, while BV4 is wholly contained within Byte3. Byte 5 contains two bit
values: BV7 and BV8. BV9 is followed by a byte- aligned type (UI16), so the last four bits of Byte 6 are padded
with zeros.
Bit Values
Type

Comment

SB[nBits]

Signed-bit value (nBits is the number of bits used to store the value)

UB[nBits]

Unsigned-bit value (nBits is the number of bits used to store the value)

FB[nBits]

Signed, fixed-point bit value (nBits is the number of bits used to store the value)

When an unsigned-bit value is expanded into a larger word size, the leftmost bits are filled with zeros. When a
signed-bit value is expanded into a larger word size, the high bit is copied to the leftmost bits.
This expansion is called sign extension. For example, the 4-bit unsigned value UB[4] = 1110 would be expanded
to a 16-bit value like this: 0000000000001110 = 14. The same value interpreted as a signed value, SB[4] = 1110
would be expanded to 1111111111111110 = –2.
Signed-bit values are similar but must take account of the sign bit. The signed value of 35 is represented as SB[7]
= 0100011. The extra zero bit is required; otherwise the high bit is sign extended and the value is interpreted as
negative.
Fixed-point bit values are 32-bit 16.16 signed, fixed-point numbers. That is, the high 16 bits represent the
number before the decimal point, and the low 16 bits represent the number after the decimal point. A fixedpoint bit value is identical to a signed-bit value, but the interpretation is different. For example, a 19-bit, signedbit value of 0x30000 is interpreted as
196608 decimal. The 19-bit, fixed-point bit value 0x30000 is interpreted as 3.0. The format of this value is
effectively 3.16 rather than 16.16.

Using bit values
Bit values are stored by using the minimum number of bits possible for the range needed. Most bit value fields
use a fixed number of bits. Some use a variable number of bits, but in all such cases, the number of bits to be
used is explicitly stated in another field in the same structure. In these variable-length cases, applications that
generate SWF files must determine the minimum number of bits necessary to represent the actual values that
will be specified. For signed-bit values, if the number to be encoded is positive, an extra bit is necessary to
preserve the leading 0; otherwise sign extension changes the bit value into a negative number.

18

As an example of variable-sized bit values, consider the RECT structure:
Field

Type

Comment

Nbits

UB[5]

Bits in each rect value field

Xmin

SB[Nbits]

x minimum position for rect

Xmax

SB[Nbits]

x maximum position for rect

Ymin

SB[Nbits]

y minimum position for rect

Ymax

SB[Nbits]

y maximum position for rect

The Nbits field determines the number of bits used to store the coordinate values Xmin, Xmax, Ymin, and Ymax.
Say the coordinates of the rectangle are as follows:
Xmin = 127
Xmax = 260
Ymin = 15
Ymax = 514

decimal = 1111111 binary
decimal = 100000100 binary
decimal = 1111 binary
decimal = 1000000010 binary

Nbits is calculated by finding the coordinate that requires the most bits to represent. In this case, that value is
514 (01000000010 binary) which requires 11 bits to represent. The rectangle is stored as the following table
shows:
Field

Type and Value

Comment

Nbits

UB[5] = 01011

Bits required (11)

Xmin

SB[11] = 00100000100

x minimum in twips (127)

Xmax

SB[11] = 00001111111

x maximum in twips (260)

Ymin

SB[11] = 00000001111

y minimum in twips (15)

Ymax

SB[11] = 01000000010

y maximum in twips (514)

String values
A string value represents a null-terminated character string. The format for a string value is a sequential list of
bytes terminated by the null character byte.
Field

Type

Comment

String

UI8[zero or more]

Non-null string character data

19

StringEnd

UI8

Marks end of string; always zero

In SWF 5 or earlier, STRING values are encoded using either ANSI (which is a superset of ASCII) or shift-JIS (a
Japanese encoding). You cannot indicate the encoding that is used; instead, the decoding choice is made
according to the locale in which Flash Player is running. This means that text content in SWF 5 or earlier can only
be encoded in ANSI or shift-JIS, and the target audience must be known during authoring—otherwise garbled
text results.
In SWF 6 or later, STRING values are always encoded by using the Unicode UTF-8 standard. This is a multibyte
encoding; each character is composed of between one and four bytes. UTF-8 is a superset of ASCII; the byte
range 0 to 127 in UTF-8 exactly matches the ASCII mapping, and all ASCII characters 0 to 127 are represented by
just one byte. UTF-8 guarantees that whenever a character other than character 0 (the null character) is
encoded by using more than one byte, none of those bytes are zero. This avoids the appearance of internal null
characters in UTF-8 strings, meaning that it remains safe to treat null bytes as string terminators, just as for ASCII
strings.

Language code
A language code identifies a spoken language that applies to text. Language codes are associated with font
specifications in the SWF file format.
Note: A language code does not specify a text encoding; it specifies a spoken language.
Field

Type

Comment

LanguageCode

UI8

Language code (see following)

Flash Player uses language codes to determine line-breaking rules for dynamic text, and to choose backup fonts
when a specified device font is unavailable. Other uses for language codes may be found in the future.
A language code of zero means no language. This code results in behavior that is dependent on the locale in
which Flash Player is running.
At the time of writing, the following language codes are recognized by Flash Player:

20

•

1 = Latin (the western languages covered by Latin-1: English, French, German, and so on)

•

2 = Japanese

•

3 = Korean

•

4 = Simplified Chinese

•

5 = Traditional Chinese

RGB color record
The RGB record represents a color as a 24-bit red, green, and blue value.
Field

Type

Comment

Red

UI8

Red color value

Green

UI8

Green color value

Blue

UI8

Blue color value

RGBA color with alpha record
The RGBA record represents a color as 32-bit red, green, blue and alpha value. An RGBA color with an alpha
value of 255 is completely opaque. An RGBA color with an alpha value of zero is completely transparent. Alpha
values between zero and 255 are partially transparent.
Field

Type

Comment

Red

UI8

Red color value

Green

UI8

Green color value

Blue

UI8

Blue color value

Alpha

UI8

alpha value defining opacity

ARGB color with alpha record
The ARGB record behaves exactly like the RGBA record, but the alpha value for the ARGB record is in the first
byte.
Field

Type

Comment

Alpha

UI8

alpha value defining opacity

Red

UI8

Red color value

Green

UI8

Green color value

Blue

UI8

Blue color value

21

Rectangle record
A rectangle value represents a rectangular region defined by a minimum x- and y-coordinate position and a
maximum x- and y-coordinate position. The RECT record must be byte aligned.
Field

Type

Comment

Nbits

UB[5]

Bits used for each subsequent field

Xmin

SB[Nbits]

x minimum position for rectangle in twips

Xmax

SB[Nbits]

x maximum position for rectangle in twips

Ymin

SB[Nbits]

y minimum position for rectangle in twips

Ymax

SB[Nbits]

y maximum position for rectangle in twips

MATRIX record
The MATRIX record represents a standard 2x3 transformation matrix of the sort commonly used in 2D graphics.
It is used to describe the scale, rotation, and translation of a graphic object. The MATRIX record must be byte
aligned.
Field

Type

Comment

HasScale

UB[1]

Has scale values if equal to 1

NScaleBits

If HasScale = 1, UB[5]

Bits in each scale value field

ScaleX

If HasScale = 1, FB[NScaleBits]

x scale value

ScaleY

If HasScale = 1, FB[NScaleBits]

y scale value

HasRotate

UB[1]

Has rotate and skew values if equal to 1

NRotateBits

If HasRotate = 1, UB[5]

Bits in each rotate value field

RotateSkew0

If HasRotate = 1, FB[NRotateBits]

First rotate and skew value

RotateSkew1

If HasRotate = 1, FB[NRotateBits]

Second rotate and skew value

NTranslateBits

UB[5]

Bits in each translate value field

TranslateX

SB[NTranslateBits]

x translate value in twips

TranslateY

SB[NTranslateBits]

y translate value in twips

22

The ScaleX, ScaleY, RotateSkew0 and RotateSkew1 fields are stored as 16.16 fixed-point values. The TranslateX
and TranslateY values are stored as signed values in twips.
The MATRIX record is optimized for common cases such as a matrix that performs a translation only. In this case,
the HasScale and HasRotate flags are zero, and the matrix only contains the TranslateX and TranslateY fields.
The mapping from the MATRIX fields to the 2x3 matrix is as follows:
ScaleX

RotateSkew0

RotateSkew1

ScaleY

TranslateX

TranslateY

For any coordinates (x, y), the transformed coordinates (x', y') are calculated as follows:
x' = x * ScaleX + y * RotateSkew1 + TranslateX
y' = x * RotateSkew0 + y * ScaleY + TranslateY

The following table describes how the members of the matrix are used for each type of operation:
ScaleX

RotateSkew0

RotateSkew1

ScaleY

Rotation

Cosine

Sine

Negative sine

Cosine

Scaling

Horizontal scaling
component

Nothing

Nothing

Vertical scaling
component

Shear

Nothing

Horizontal
proportionality
constant

Vertical
proportionality
constant

Nothing

Reflection

Horizontal reflection
component

Nothing

Nothing

Vertical reflection
component

23

Color transform record
The CXFORM record defines a simple transform that can be applied to the color space of a graphic object. The
following are the two types of transform possible:
•

Multiplication transforms

•

Addition transforms

Multiplication transforms multiply the red, green, and blue components by an 8.8 fixed-point multiplication
term. The fixed-point representation of 1.0 is 0x100 or 256 decimal.
For any color (R,G,B), the transformed color (R', G', B') is calculated as follows:
R' = (R * RedMultTerm)
G' = (G * GreenMultTerm)
B' = (B * BlueMultTerm)

/ 256
/ 256
/ 256

Addition transforms add an addition term (positive or negative) to the red, green, and blue components of the
object being displayed. If the result is greater than 255, the result is clamped to 255. If the result is less than
zero, the result is clamped to zero.
For any color (R,G,B), the transformed color (R', G', B') is calculated as follows:
R' = max(0, min(R + RedAddTerm,
G' = max(0, min(G + GreenAddTerm,
B' = max(0, min(B + BlueAddTerm,

255))
255))
255))

Addition and multiplication transforms can be combined as follows. The multiplication operation is performed
first:
R' = max(0, min(((R * RedMultTerm)
G' = max(0, min(((G * GreenMultTerm)
B' = max(0, min(((B * BlueMultTerm)

/ 256) + RedAddTerm,
/ 256) + GreenAddTerm,
/ 256) + BlueAddTerm,

255))
255))
255))

The CXFORM record must be byte aligned.
Field

Type

Comment

HasAddTerms

UB[1]

Has color addition values if equal to 1

HasMultTerms

UB[1]

Has color multiply values if equal to 1

Nbits

UB[4]

Bits in each value field

RedMultTerm

If HasMultTerms = 1, SB[Nbits]

Red multiply value

GreenMultTerm

If HasMultTerms = 1, SB[Nbits]

Green multiply value

24

BlueMultTerm

If HasMultTerms = 1, SB[Nbits]

Blue multiply value

RedAddTerm

If HasAddTerms = 1, SB[Nbits]

Red addition value

GreenAddTerm

If HasAddTerms = 1, SB[Nbits]

Green addition value

BlueAddTerm

If HasAddTerms = 1, SB[Nbits]

Blue addition value

Color transform with alpha record
The CXFORMWITHALPHA record extends the functionality of CXFORM by allowing color transforms to be applied
to the alpha channel, as well as the red, green, and blue channels.
The following are the two types of transform possible:
•

Multiplication Transforms

•

Addition Transforms

Multiplication transforms multiply the red, green, blue, and alpha components by an 8.8 fixed-point value. The
fixed-point representation of 1.0 is 0x100 or 256 decimal. Therefore, the result of a multiplication operation
should be divided by 256.
For any color (R,G,B,A), the transformed color (R', G', B', A') is calculated as follows:
R' = (R * RedMultTerm) / 256
G' = (G * GreenMultTerm) / 256
B' = (B * BlueMultTerm) / 256
A' = (A * AlphaMultTerm) / 256
The CXFORMWITHALPHA record is most commonly used to display objects as partially transparent, achieved by
multiplying the alpha channel by some value between zero and 256.
Addition transforms add a fixed value (positive or negative) to the red, green, blue, and alpha components of
the object being displayed. If the result is greater than 255, the result is clamped to 255. If the result is less than
zero, the result is clamped to zero.
For any color (R,G,B,A), the transformed color (R', G', B', A') is calculated as follows:
R' = max(0, min(R + RedAddTerm,
255))
G' = max(0, min(G + GreenAddTerm,
255))
B' = max(0, min(B + BlueAddTerm, 255))
A' = max(0, min(A + AlphaAddTerm,
255))

25

Addition and multiplication transforms can be combined as follows. The multiplication operation is performed
first:
R' = max(0, min(((R * RedMultTerm) / 256) + RedAddTerm, 255))
G' = max(0, min(((G * GreenMultTerm) / 256) + GreenAddTerm, 255))
B' = max(0, min(((B * BlueMultTerm) / 256) + BlueAddTerm, 255))
A' = max(0, min(((A * AlphaMultTerm) / 256) + AlphaAddTerm, 255))
Like the CXFORM record, the CXFORMWITHALPHA record is byte aligned.
Field

Type

Comment

HasAddTerms

UB[1]

Has color addition values if equal to 1

HasMultTerms

UB[1]

Has color multiply values if equal to 1

Nbits

UB[4]

Bits in each value field

RedMultTerm

If HasMultTerms = 1, SB[Nbits]

Red multiply value

GreenMultTerm

If HasMultTerms = 1, SB[Nbits]

Green multiply value

BlueMultTerm

If HasMultTerms = 1, SB[Nbits]

Blue multiply value

AlphaMultTerm

If HasMultTerms = 1, SB[Nbits]

Alpha multiply value

RedAddTerm

If HasAddTerms = 1, SB[Nbits]

Red addition value

GreenAddTerm

If HasAddTerms = 1, SB[Nbits]

Green addition value

BlueAddTerm

If HasAddTerms = 1, SB[Nbits]

Blue addition value

AlphaAddTerm

If HasAddTerms = 1, SB[Nbits]

Transparency addition value

26

Chapter 2: SWF Structure Summary
This chapter provides a summary of the elements that comprise a SWF file.

The SWF header
All SWF files begin with the following header. The types are defined in Chapter 1: Basic Data Types.
Field

Type

Comment

Signature

UI8

Signature byte:
“F” indicates uncompressed
“C” indicates a zlib compressed SWF (SWF 6
and later only)
“Z” indicates a LZMA compressed SWF (SWF
13 and later only)

Signature

UI8

Signature byte always “W”

Signature

UI8

Signature byte always “S”

Version

UI8

Single byte file version (for example, 0x06 for
SWF 6)

FileLength

UI32

Length of entire file in bytes

FrameSize

RECT

Frame size in twips

FrameRate

UI16

Frame delay in 8.8 fixed number of frames
per second

FrameCount

UI16

Total number of frames in file

The header begins with a three-byte signature of one of:
•
•

•

27

0x46, 0x57, 0x53 (“FWS”). An FWS signature indicates an uncompressed SWF file.
0x43, 0x57, 0x53 (“CWS”). A CWS indicates that the entire file after the first 8 bytes (that is,
after the FileLength field) was compressed by using the ZLIB open standard. The data format
that the ZLIB library uses is described by Request for Comments (RFCs) documents 1950 to
1952. CWS file compression is permitted in SWF 6 or later only.
0x5a, 0x57, 0x53 (“ZWS”). A ZWS indicates that the entire file after the first 8 bytes (that is,
after the FileLength field) was compressed by using the LZMA open standard: http://www.7-

zip.org/sdk.html. ZWS file compression is permitted in SWF 13 or later only.
A one-byte version number follows the signature. The version number is not an ASCII character, but an 8-bit
number. For example, for SWF 4, the version byte is 0x04, not the ASCII character “4” (0x34).
The FileLength field is the total length of the SWF file, including the header. If this is an uncompressed SWF file
(FWS signature), the FileLength field should exactly match the file size. If this is a compressed SWF file (CWS
signature), the FileLength field indicates the total length of the file after decompression, and thus generally does
not match the file size. Having the uncompressed size available can make the decompression process more
efficient.
The FrameSize field defines the width and height of the on-screen display. This field is stored as a RECT
structure, meaning that its size may vary according to the number of bits needed to encode the coordinates. The
FrameSize RECT always has Xmin and Ymin value of 0; the Xmax and Ymax members define the width and height
(see Using bit values).
The FrameRate is the desired playback rate in frames per second. This rate is not guaranteed if, for example,
Flash Player is running on a slow or busy CPU.
The FrameCount is the total number of frames in this SWF file.

SWF file structure
Following the header is a series of tagged data blocks. All tags share a common format, so any program parsing a
SWF file can skip over blocks it does not understand. Data inside the block can point to offsets within the block,
but can never point to an offset in another block. This ability enables tags to be removed, inserted, or modified
by tools that process a SWF file.
The FileAttributes tag is only required for SWF 8 and later.
Header

28

FileAttributes
tag

Tag

Tag

…

End tag

Tag format
Each tag begins with a tag type and a length. The tag-header formats can be either short or long. Short tag
headers are used for tags with 62 bytes of data or less. Long tag headers, with a signed 32-bit length field, can be
used for any tag size up to 2GB, far larger than is presently practical.
RECORDHEADER (short)
Field

Type

Comment

TagCodeAndLength

UI16

Upper 10 bits: tag type Lower 6 bits: tag length

Note: The TagCodeAndLength field is a two-byte word, not a bit field of 10 bits followed by a bit field of 6 bits.
The little-endian byte ordering of a SWF file makes these two layouts different.
The length specified in the TagCodeAndLength field does not include the RECORDHEADER that starts a tag.
If the tag is 63 bytes or longer, it is stored in a long tag header. The long tag header consists of a short tag
header with a length of 0x3f, followed by a 32-bit length.
RECORDHEADER (long)
Field

Type

Comment

TagCodeAndLength

UI16

Tag type and length of 0x3F Packed together as in short
header

Length

UI32

Length of tag

Definition and control tags
The two categories of tags in a SWF file are as follows:

29

•

Definition tags define the content of the SWF file—the shapes, text, bitmaps, sounds, and so on. Each
definition tag assigns a unique ID called a character ID to the content it defines. Flash Player then stores
the character in a repository called the dictionary. Definition tags, by themselves, do not cause anything
to be rendered.

•

Control tags create and manipulate rendered instances of characters in the dictionary, and control the
flow of the file.

Tag ordering in SWF files
Generally speaking, tags in a SWF can occur in any order. However, you must observe the following rules:
•

The FileAttributes tag must be the first tag in the SWF file for SWF 8 and later.

•

A tag should only depend on tags that come before it. A tag should never depend on a tag that comes
later in the file.

•

A definition tag that defines a character must occur before any control tag that refers to that character.

•

Streaming sound tags must be in order. Out-of-order streaming sound tags result in the sound being
played out of order.

•

The End tag is always the last tag in the SWF file.

The dictionary
The dictionary is a repository of characters that are defined, and are available for control tags to use. The
process of building and using the dictionary is as follows:
1. The definition tag defines some content, such as a shape, font, bitmap, or sound.
2. The definition tag assigns a unique CharacterId to the content.
3. The content is saved in the dictionary under the CharacterId.
4. A control tag uses the CharacterId to retrieve the content from the dictionary, and performs some
action on the content, such as displaying a shape, or playing a sound.
Every definition tag must specify a unique ID. Duplicate IDs are not allowed. Typically, the first CharacterId is 1,
the second CharacterId is 2, and so on. The number zero (0) is special and is considered a null character.
Control tags are not the only tags that reference the dictionary. Definition tags can use characters from the
dictionary to define more complex characters. For example, the DefineButton and DefineSprite tags refer to
other characters to define their contents. The DefineText tag can refer to font characters to select different
fonts for the text.

30

The following diagram illustrates a typical interaction between definition tags, control tags, and the dictionary:
Tags in SWF file

Dictionary

DefineShape as character 1

Character 1
Shape

DefineSound as character 2
DefineFont as character 3

Character 2
Sound

PlaceObject character 1
Add shape to display list*

Character 3
Font

DefineText as character 4
Uses font defined as character 3

Character 4
Text

PlaceObject character 4
Add text to display list*

Character 5
Morph

ShowFrame
Render contents of the display*
DefineMorphShape as character 5
StartSound character 2
PlaceObject character 5
Add Morph to display list*
ShowFrame
Render contents of the display*

31

Contol tag
Definition tag
Character

Processing a SWF file
Flash Player processes all of the tags in a SWF file until a ShowFrame tag is encountered. At this point, the
display list is copied to the screen and Flash Player is idle until it is time to process the next frame. The contents
of the first frame are the cumulative effect of performing all of the control tag operations before the first
ShowFrame tag. The contents of the second frame are the cumulative effect of performing all of the control tag
operations from the beginning of the file to the second ShowFrame tag, and so on.

File compression strategy
Since SWF files are frequently delivered over a network connection, they should be as compact as possible.
Several techniques are used to accomplish this, including the following items:
•

Reuse—The structure of the character dictionary makes it easy to reuse elements in a SWF file. For
example, a shape, button, sound, font, or bitmap can be stored in a file once and referenced many
times.

•

Compression—Shapes are compressed by using an efficient delta encoding scheme; often the first
coordinate of a line is assumed to be the last coordinate of the previous line. Distances are also often
expressed relative to the last position.

•

Default values—Some structures, like matrixes and color transforms, have common fields that are used
more often than others. For example, for a matrix, the most common field is the translation field.
Scaling and rotation are less common. Therefore, if the scaling field is not present, it is assumed to be
100%. If the rotation field is not present, it is assumed that there is no rotation. This use of default
values helps to minimize file sizes.

•

Change Encoding—As a rule, SWF files only store the changes between states. This is reflected in shape
data structures and in the place-move-remove model that the display list uses.

•

Shape Data Structure—The shape data structure uses a unique structure to minimize the size of shapes
and to render anti-aliased shapes efficiently on the screen.

Summary
A SWF file is made up of a header, followed by a number of tags. The two types of tags are definition tags and
control tags. Definition tags define the objects known as characters, which are stored in the dictionary. Control
tags manipulate characters, and control the flow of the file.

32

Chapter 3: The Display List
Displaying a frame of a SWF file is a three-stage process:
1. Objects are defined with definition tags such as DefineShape, DefineSprite, and so on. Each object is
given a unique ID called a character, and is stored in a repository called the dictionary.
2. Selected characters are copied from the dictionary and placed on the display list, which is the list of the
characters that will be displayed in the next frame.
3. Once complete, the contents of the display list are rendered to the screen with ShowFrame.
A depth value is assigned to each character on the display list. The depth determines the stacking order of the
character. Characters with lower depth values are displayed underneath characters with higher depth values. A
character with a depth value of 1 is displayed at the bottom of the stack. A character can appear more than once
in the display list, but at different depths. Only one character can be at any given depth.
In SWF 1 and 2, the display list was a flat list of the objects that are present on the screen at any given time. In
SWF 3 and later versions, the display list is a hierarchical list where an element on the display can have a list of
child elements. For more information, see DefineSprite.
The following six tags are used to control the display list:
•

PlaceObject

•

PlaceObject2 & PlaceObject3
specified depth.

•

RemoveObject Removes the specified character from the display list.

•

RemoveObject2 Removes the character at the specified depth.

•

ShowFrame

Adds a character to the display list.
Adds a character to the display list, or modifies the character at the

Renders the contents of the display list to the display.

Note: The older tags, PlaceObject and RemoveObject, are rarely used in SWF 3 and later versions.

33

The following diagram illustrates the display process. First, three objects are defined: a shape, a text object, and
a sprite. These objects are given character IDs and stored in the dictionary. Character 1 (the shape) is then
placed at depth 1, the bottom of the stack, and will be obscured by all other characters when the frame is
rendered. Character 2 (the text) is placed twice; once at depth 2, and once at depth 4, the top of the stack.
Character 3 (the sprite) is placed at depth 3.

Definition

Dictionary

Display List

DefineShape
Character ID = 1

Character ID = 1

Character ID = 2
Depth = 4

DefineText
Character ID = 1

Character ID = 2

Character ID = 3
Depth = 3

DefineSprite
Character ID = 1

Character ID = 3

Character ID = 2
Depth = 2
Character ID = 1
Depth = 1

34

Top

Bottom

Clipping layers
Flash Player supports a special kind of object in the display list called a clipping layer. A character placed as a
clipping layer is not displayed; rather it clips (or masks) the characters placed above it. The ClipDepth field in
PlaceObject2 specifies the top-most depth that the clipping layer masks.
For example, if a shape was placed at depth 1 with a ClipDepth of 4, all depths above 1, up to and including
depth 4, are masked by the shape placed at depth 1. Characters placed at depths above 4 are not masked.
Display List
Character ID =4
Depth = 5

Key
Top
Clipping Layer

Character ID = 3
Depth = 4

Character masked
by Clipping Layer

Character ID = 3
Depth = 3

Character not masked
by Clipping Layer

Character ID = 2
Depth = 2
Character ID = 1
Depth = 1
ClipDepth = 4

Bottom

Using the display list
The following procedure creates and displays an animation:
1. Define each character with a definition tag.
Each character is given a unique character ID, and added to the dictionary.
2. Add each character to the display list with a PlaceObject2 tag. Each PlaceObject2 tag specifies the
character to be displayed, plus the following attributes:

33

•

A depth value, which controls the stacking order of the character being placed. Characters with
lower depth values appear to be underneath characters with higher depth values. A depth value of 1
means the character is displayed at the bottom of the stack. Only one character can be at any given
depth.

•

A transformation matrix, which determines the position, scale, factor, and angle of rotation of the
character being placed. The same character can be placed more than once (at different depths) with
a different transformation matrix.

•

An optional color transform, which specifies the color effect applied to the character being placed.
Color effects include transparency and color shifts.

•

An optional name string, which identifies the character being placed for SetTarget actions. SetTarget
is used to perform actions inside sprite objects.

•

An optional ClipDepth value, which specifies the top-most depth that will be masked by the
character being placed.

•

An optional ratio value, which controls how a morph character is displayed when placed. A ratio of
zero displays the character at the start of the morph. A ratio of 65535 displays the character at the
end of the morph.

3. Use a ShowFrame tag to render the contents of the display list to the screen.
4. Use a PlaceObject2 tag to modify each character on the display List.
Each PlaceObject2 assigns a new transformation matrix to the character at a given depth. The character
ID is not specified because each depth can have only one character.
5. Use a ShowFrame tag to display the characters in their new positions.
Repeat steps 4 and 5 for each frame of the animation.
If a character does not change from frame to frame, you do not need to replace the unchanged
character after each frame.
6. Use a RemoveObject2 tag to Remove each character from the display list. Only the depth value is
required to identify the character being removed.

Display list tags
Display list tags are used to add character and character attributes to a display list.

PlaceObject
The PlaceObject tag adds a character to the display list. The CharacterId identifies the character to be added.
The Depth field specifies the stacking order of the character. The Matrix field species the position, scale, and
rotation of the character. If the size of the PlaceObject tag exceeds the end of the transformation matrix, it is
assumed that a ColorTransform field is appended to the record. The ColorTransform field specifies a color effect
(such as transparency) that is applied to the character. The same character can be added more than once to the
display list with a different depth and transformation matrix.
PlaceObject is rarely used in SWF 3 and later versions; it is superseded by PlaceObject2 and PlaceObject3.
The minimum file format version is SWF 1.

34

Field

Type

Comment

Header

RECORDHEADER

Tag type = 4

CharacterId

UI16

ID of character to place

Depth

UI16

Depth of character

Matrix

MATRIX

Transform matrix data

ColorTransform (optional)

CXFORM

Color transform data

PlaceObject2
The PlaceObject2 tag extends the functionality of the PlaceObject tag. The PlaceObject2 tag can both add a
character to the display list, and modify the attributes of a character that is already on the display list. The
PlaceObject2 tag changed slightly from SWF 4 to SWF 5. In SWF 5, clip actions were added.
The tag begins with a group of flags that indicate which fields are present in the tag. The optional fields are
CharacterId, Matrix, ColorTransform, Ratio, ClipDepth, Name, and ClipActions. The Depth field is the only field
that is always required.
The depth value determines the stacking order of the character. Characters with lower depth values are
displayed underneath characters with higher depth values. A depth value of 1 means the character is displayed
at the bottom of the stack. Any given depth can have only one character. This means a character that is already
on the display list can be identified by its depth alone (that is, a CharacterId is not required).
The PlaceFlagMove and PlaceFlagHasCharacter tags indicate whether a new character is being added to the
display list, or a character already on the display list is being modified. The meaning of the flags is as follows:
•

PlaceFlagMove = 0 and PlaceFlagHasCharacter = 1
A new character (with ID of CharacterId) is placed on the display list at the specified depth. Other fields
set the attributes of this new character.

•

PlaceFlagMove = 1 and PlaceFlagHasCharacter = 0
The character at the specified depth is modified. Other fields modify the attributes of this character.
Because any given depth can have only one character, no CharacterId is required.

•

PlaceFlagMove = 1 and PlaceFlagHasCharacter = 1
The character at the specified Depth is removed, and a new character (with ID of CharacterId) is placed
at that depth. Other fields set the attributes of this new character.

Frames replace the transformation matrix of the character at the desired depth. For example, a character that is
moved over a series of frames has PlaceFlagHasCharacter set in the first frame, and PlaceFlagMove set in

35

subsequent frames. The first frame places the new character at the desired depth, and sets the initial
transformation matrix. Subsequent
The optional fields in PlaceObject2 have the following meaning:
•

The CharacterId field specifies the character to be added to the display list. CharacterId is used only
when a new character is being added. If a character that is already on the display list is being modified,
the CharacterId field is absent.

•

The Matrix field specifies the position, scale and rotation of the character being added or modified.

•

The ColorTransform field specifies the color effect applied to the character being added or modified.

•

The Ratio field specifies a morph ratio for the character being added or modified. This field applies only
to characters defined with DefineMorphShape, and controls how far the morph has progressed. A ratio
of zero displays the character at the start of the morph. A ratio of 65535 displays the character at the
end of the morph. For values between zero and 65535 Flash Player interpolates between the start and
end shapes, and displays an in- between shape.

•

The ClipDepth field specifies the top-most depth that will be masked by the character being added. A
ClipDepth of zero indicates that this is not a clipping character.

•

The Name field specifies a name for the character being added or modified. This field is typically used
with sprite characters, and is used to identify the sprite for SetTarget actions. It allows the main file (or
other sprites) to perform actions inside the sprite (see Chapter 13: Sprites and Movie Clips).

•

The ClipActions field, which is valid only for placing sprite characters, defines one or more event
handlers to be invoked when certain events occur.

The minimum file format version is SWF 3.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 26

PlaceFlagHasClipActions

UB[1]

SWF 5 and later: has clip actions
(sprite characters only) Otherwise:
always 0

PlaceFlagHasClipDepth

UB[1]

Has clip depth

PlaceFlagHasName

UB[1]

Has name

PlaceFlagHasRatio

UB[1]

Has ratio

PlaceFlagHasColorTransform

UB[1]

Has color transform

36

PlaceFlagHasMatrix

UB[1]

Has matrix

PlaceFlagHasCharacter

UB[1]

Places a character

PlaceFlagMove

UB[1]

Defines a character to be moved

Depth

UI16

Depth of character

CharacterId

If PlaceFlagHasCharacter, UI16

ID of character to place

Matrix

If PlaceFlagHasMatrix, MATRIX

Transform matrix data

ColorTransform

If PlaceFlagHasColorTransform,
CXFORMWITHALPHA

Color transform data

Ratio

If PlaceFlagHasRatio, UI16

Name

If PlaceFlagHasName, STRING

Name of character

ClipDepth

If PlaceFlagHasClipDepth, UI16

Clip depth (see “Clipping layers”)

ClipActions

If PlaceFlagHasClipActions,
CLIPACTIONS

SWF 5 and later: Clip Actions Data

Clip actions are valid for placing sprite characters only. Clip actions define event handlers for a sprite character.
Field

Type

Comment

Reserved

UI16

Must be 0

AllEventFlags

CLIPEVENTFLAGS

All events used in these clip actions

ClipActionRecords

CLIPACTIONRECORD [one or more]

Individual event handlers

ClipActionEndFlag

If SWF version <= 5, UI16,
If SWF version >= 6, UI32

Must be 0

CLIPACTIONRECORD
Field

Type

Comment

EventFlags

CLIPEVENTFLAGS

Events to which this handler applies

ActionRecordSize

UI32

Offset in bytes from end of this field

37

to next CLIPACTIONRECORD (or
ClipActionEndFlag)
KeyCode

If EventFlags contain ClipEventKeyPress: UI8,
Otherwise absent

Key code to trap (see
“DefineButton2”)

Actions

ACTIONRECORD [one or more]

Actions to perform

PlaceObject3
The PlaceObject3 tag extends the functionality of the PlaceObject2 tag. PlaceObject3 adds the following new
features:

38

•

The PlaceFlagHasClassName field indicates that a class name will be specified, indicating the type
of object to place. Because we no longer use ImportAssets in ActionScript 3.0, there needed to be some
way to place a Timeline object using a class imported from another SWF, which does not have a 16-bit
character ID in the instantiating SWF. Supported in Flash Player 9.0.45.0 and later.

•

The PlaceFlagHasImage field indicates the creation of native Bitmap objects on the display list.
When PlaceFlagHasClassName and PlaceFlagHasImage are both defined, this indicates a Bitmap class to
be loaded from another SWF. Immediately following the flags is the class name (as above) for the
BitmapData class in the loaded SWF. A Bitmap object will be placed with the named BitmapData class as
it's internal data. When PlaceFlagHasCharacter and PlaceFlagHasImage are both defined, this indicates a
Bitmap from the current SWF. The BitmapData to be used as its internal data will be defined by the
following characterID. This only occurs when the BitmapData has a class associated with it. If there is no
class associated with the BitmapData, DefineShape should be used with a Bitmap fill. Supported in Flash
Player 9.0.45.0 and later.

•

The PlaceFlagHasCacheAsBitmap field specifies whether Flash Player should internally cache a
display object as a bitmap. Caching can speed up rendering when the object does not change frequently.

•

A number of different blend modes can be specified as an alternative to normal alpha compositing. The
following blend modes are supported:
o

Add

o

Layer Alpha

o

Lighten Darken

o

Overlay Difference

o

Multiply Erase

o

Screen Hardlight

•

o

Subtract

o

Invert

A number of bitmap filters can be applied to the display object. Adding filters implies that the display
object will be cached as a bitmap. The following bitmap filters are supported:
o Bevel
o

Drop shadow

o

Blur

o

Glow

o

Color matrix

o

Gradient bevel

o

Convolution

o

Gradient glow

The minimum file format version is SWF 8.
PlaceObject3
Field

Type

Comment

Header

RECORDHEADER

Tag type = 70

PlaceFlagHasClipActions

UB[1]

SWF 5 and later: has clip actions
(sprite characters only) Otherwise:
always 0

PlaceFlagHasClipDepth

UB[1]

Has clip depth

PlaceFlagHasName

UB[1]

Has name

PlaceFlagHasRatio

UB[1]

Has ratio

PlaceFlagHasColorTransform

UB[1]

Has color transform

PlaceFlagHasMatrix

UB[1]

Has matrix

PlaceFlagHasCharacter

UB[1]

Places a character

PlaceFlagMove

UB[1]

Defines a character to be moved

39

Reserved

UB[1]

Must be 0

PlaceFlagOpaqueBackground

UB[1]

Has opaque background. SWF 11 and
higher.

PlaceFlagHasVisible

UB[1]

Has visibility flag. SWF 11 and higher.

PlaceFlagHasImage

UB[1]

Has class name or character ID of
bitmap to place. If
PlaceFlagHasClassName, use
ClassName. If PlaceFlagHasCharacter,
use CharacterId

PlaceFlagHasClassName

UB[1]

Has class name of object to place

PlaceFlagHasCacheAsBitmap

UB[1]

Enables bitmap caching

PlaceFlagHasBlendMode

UB[1]

Has blend mode

PlaceFlagHasFilterList

UB[1]

Has filter list

Depth

UI16

Depth of character

ClassName

If PlaceFlagHasClassName or
(PlaceFlagHasImage and
PlaceFlagHasCharacter), String

Name of the class to place

CharacterId

If PlaceFlagHasCharacter, UI16

ID of character to place

Matrix

If PlaceFlagHasMatrix, MATRIX

Transform matrix data

ColorTransform

If PlaceFlagHasColorTransform,
CXFORMWITHALPHA

Color transform data

Ratio

If PlaceFlagHasRatio, UI16

Name

If PlaceFlagHasName, STRING

Name of character

ClipDepth

If PlaceFlagHasClipDepth, UI16

Clip depth (see Clipping layers)

SurfaceFilterList

If PlaceFlagHasFilterList, FILTERLIST

List of filters on this object

BlendMode

If PlaceFlagHasBlendMode, UI8

0 or 1 = normal;
2 = layer;
3 = multiply;
4 = screen;
5 = lighten;

40

6 = darken;
7 = difference;
8 = add;
9 = subtract;
10 = invert;
11 = alpha;
12 = erase;
13 = overlay;
14 = hardlight;
Values 15 to 255 are reserved.
BitmapCache

If PlaceFlagHasCacheAsBitmap, UI8

0 = Bitmap cache disabled; 1-255 =
Bitmap cache enabled

Visible

If PlaceFlagHasVisible, UI8

0 = Place invisible, 1 = Place visible

Background Color

If PlaceFlagHasVisible, RGBA

ClipActions

If PlaceFlagHasClipActions,
CLIPACTIONS

SWF 5 and later: Clip Actions Data

Field

Type

Comment

NumberOfFilters

UI8

Number of Filters

Filter

FILTER[NumberOfFilters]

List of filters

FILTERLIST

FILTER
Field

Type

Comment

FilterID

UI8

0 = Has DropShadowFilter
1 = Has BlurFilter
2 = Has GlowFilter
3 = Has BevelFilter
4 = Has GradientGlowFilter
5 = Has ConvolutionFilter
6 = Has ColorMatrixFilter
7 = Has GradientBevelFilter

41

DropShadowFilter

If FilterID = 0, DROPSHADOWFILTER

Drop Shadow filter

BlurFilter

If FilterID = 1, BLURFILTER

Blur filter

GlowFilter

If FilterID = 2, GLOWFILTER

Glow filter

BevelFilter

If FilterID = 3, BEVELFILTER

Bevel filter

GradientGlowFilter

If FilterID = 4, GRADIENTGLOWFILTER

Gradient Glow filter

ConvolutionFilter

If FilterID = 5, CONVOLUTIONFILTER

Convolution filter

ColorMatrixFilter

If FilterID = 6, COLORMATRIXFILTER

Color Matrix filter

GradientBevelFilter

If FilterID = 7, GRADIENTBEVELFILTER

Gradient Bevel filter

Color Matrix filter
A Color Matrix filter applies a color transformation on the pixels of a display list object. Given an input RGBA
pixel in a display list object, the color transformation is calculated in the following way:
The resulting RGBA values are saturated.
The matrix values are stored from left to right and each row from top to bottom. The last row is always assumed
to be (0,0,0,0,1) and does not need to be stored.
COLORMATRIXFILTER
Field

Type

Comment

Matrix

FLOAT[20]

Color matrix values

R'
G'
B'
A'
1

r0 r1 r2 r3 r4
g0 g1 g2 g3 g4
= b0 b1 b2 b3 b4
a0 a1 a2 a3 a4
0 0 0 0 1

R
G
B
A
1

Convolution filter
The Convolution filter is a two-dimensional discrete convolution. It is applied on each pixel of a display object. In
the following mathematical representation, F is the input pixel plane, G is the input matrix, and H is the output
pixel plane:
42

Ma tri x Y – 1 Ma tri x X – 1

H x y

=
j=0

43

i=0

M a t r i xX
Ma trix Y
F x + i – --------------------- y + j – ---------------------- + Bias G i j
2
2
---------------------------------------------------------------------------------------------------------------------------------------Divisor

The convolution is applied on each of the RGBA color components and then saturated, except when the
PreserveAlpha flag is set; in this case, the alpha channel value is not modified. The clamping flag specifies how
pixels outside of the input pixel plane are handled. If set to false, the DefaultColor value is used, and otherwise,
the pixel is clamped to the closest valid input pixel.
CONVOLUTIONFILTER
Field

Type

Comment

MatrixX

UI8

Horizontal matrix size

MatrixY

UI8

Vertical matrix size

Divisor

FLOAT

Divisor applied to the matrix values

Bias

FLOAT

Bias applied to the matrix values

Matrix

FLOAT[MatrixX * MatrixY]

Matrix values

DefaultColor

RGBA

Default color for pixels outside the image

Reserved

UB[6]

Must be 0

Clamp

UB[1]

Clamp mode

PreserveAlpha

UB[1]

Preserve the alpha

Blur filter
The blur filter is based on a sub-pixel precise median filter (also known as a box filter). The filter is applied on
each of the RGBA color channels.
The general mathematical representation of a simple non-sub-pixel precise median filter is as follows, and can
be easily extended to support sub-pixel precision.
This representation assumes that BlurX and BlurY are odd integers in order to get the same result as Flash Player.
The filter window is always centered on a pixel in Flash Player.

When the number of passes is set to three, it closely approximates a Gaussian Blur filter. A higher number of
passes is possible, but for performance reasons, Adobe does not recommend it.
44

BLURFILTER
Field

Type

Comment

BlurX

FIXED

Horizontal blur amount

BlurY

FIXED

Vertical blur amount

Passes

UB[5]

Number of blur passes

Reserved

UB[3]

Must be 0

Drop Shadow filter
The Drop Shadow filter is based on the same median filter as the blur filter, but the filter is applied only on the
alpha color channel to obtain a shadow pixel plane.
The angle parameter is in radians. With angle set to 0, the shadow shows on the right side of the object. The
distance is measured in pixels. The shadow pixel plane values are interpolated bilinearly if sub-pixel values are
used.
The strength of the shadow normalized is 1.0 in fixed point. The strength value is applied by multiplying each
value in the shadow pixel plane.
Various compositing options are available for the drop shadow to support both inner and outer shadows in
regular or knockout modes.
The resulting color value of each pixel is obtained by multiplying the color channel of the provided RGBA color
value by the associated value in the shadow pixel plane. The resulting pixel value is composited on the original
input pixel plane by using one of the specified compositing modes.
DROPSHADOWFILTER
Field

Type

Comment

DropShadowColor

RGBA

Color of the shadow

BlurX

FIXED

Horizontal blur amount

BlurY

FIXED

Vertical blur amount

Angle

FIXED

Radian angle of the drop shadow

Distance

FIXED

Distance of the drop shadow

Strength

FIXED8

Strength of the drop shadow

45

InnerShadow

UB[1]

Inner shadow mode

Knockout

UB[1]

Knockout mode

CompositeSource

UB[1]

Composite source. Always 1

Passes

UB[5]

Number of blur passes

Glow filter
The Glow filter works in the same way as the Drop Shadow filter, except that it does not have a distance and
angle parameter. Therefore, it can run slightly faster.
GLOWFILTER
Field

Type

Comment

GlowColor

RGBA

Color of the shadow

BlurX

FIXED

Horizontal blur amount

BlurY

FIXED

Vertical blur amount

Strength

FIXED8

Strength of the glow

InnerGlow

UB[1]

Inner glow mode

Knockout

UB[1]

Knockout mode

CompositeSource

UB[1]

Composite source. Always 1

Passes

UB[5]

Number of blur passes

Bevel filter
The Bevel filter creates a smooth bevel on display list objects.
BEVELFILTER
Field

Type

Comment

ShadowColor

RGBA

Color of the shadow

HighlightColor

RGBA

Color of the highlight

BlurX

FIXED

Horizontal blur amount

46

BlurY

FIXED

Vertical blur amount

Angle

FIXED

Radian angle of the drop shadow

Distance

FIXED

Distance of the drop shadow

Strength

FIXED8

Strength of the drop shadow

InnerShadow

UB[1]

Inner shadow mode

Knockout

UB[1]

Knockout mode

CompositeSource

UB[1]

Composite source. Always 1

OnTop

UB[1]

OnTop mode

Passes

UB[4]

Number of blur passes

Gradient Glow and Gradient Bevel filters
The Gradient Glow and Gradient Bevel filters are extensions of the normal Glow and Bevel Filters and allow a
gradient to be specified instead of a single color. Instead of multiplying a single color value by the shadow-pixel
plane value, the shadow-pixel plane value is mapped directly into the gradient ramp to obtain the resulting color
pixel value, which is then composited by using one of the specified compositing modes.
GRADIENTGLOWFILTER
Field

Type

Comment

NumColors

UI8

Number of colors in the gradient

GradientColors

RGBA[NumColors]

Gradient colors

GradientRatio

UI8[NumColors]

Gradient ratios

BlurX

FIXED

Horizontal blur amount

BlurY

FIXED

Vertical blur amount

Angle

FIXED

Radian angle of the gradient glow

Distance

FIXED

Distance of the gradient glow

Strength

FIXED8

Strength of the gradient glow

InnerShadow

UB[1]

Inner glow mode

47

Knockout

UB[1]

Knockout mode

CompositeSource

UB[1]

Composite source. Always 1

OnTop

UB[1]

OnTop mode

Passes

UB[4]

Number of blur passes

NumColors

UI8

Number of colors in the gradient

GradientColors

RGBA[NumColors]

Gradient colors

GradientRatio

UI8[NumColors]

Gradient ratios

BlurX

FIXED

Horizontal blur amount

BlurY

FIXED

Vertical blur amount

Angle

FIXED

Radian angle of the gradient bevel

Distance

FIXED

Distance of the gradient bevel

Strength

FIXED8

Strength of the gradient bevel

InnerShadow

UB[1]

Inner bevel mode

Knockout

UB[1]

Knockout mode

Field

Type

Comment

CompositeSource

UB[1]

Composite source. Always 1

OnTop

UB[1]

OnTop mode

Passes

UB[4]

Number of blur passes

GRADIENTBEVELFILTER

ClipEventFlags
The CLIPEVENTFLAGS sequence specifies one or more sprite events to which an event handler applies. In SWF 5
and earlier, CLIPEVENTFLAGS is 2 bytes; in SWF 6 and later, it is 4 bytes.
CLIPEVENTFLAGS

48

Field

Type

Comment

ClipEventKeyUp

UB[1]

Key up event

ClipEventKeyDown

UB[1]

Key down event

ClipEventMouseUp

UB[1]

Mouse up event

ClipEventMouseDown

UB[1]

Mouse down event

ClipEventMouseMove

UB[1]

Mouse move event

ClipEventUnload

UB[1]

Clip unload event

ClipEventEnterFrame

UB[1]

Frame event

ClipEventLoad

UB[1]

Clip load event

ClipEventDragOver

UB[1]

SWF 6 and later: mouse drag over event Otherwise:
always 0

ClipEventRollOut

UB[1]

SWF 6 and later: mouse rollout event. Otherwise:
always 0

ClipEventRollOver

UB[1]

SWF 6 and later: mouse rollover event. Otherwise:
always 0

ClipEventReleaseOutside

UB[1]

SWF 6 and later: mouse release outside event
Otherwise: always 0

ClipEventRelease

UB[1]

SWF 6 and later: mouse release inside event
Otherwise: always 0

ClipEventPress

UB[1]

SWF 6 and later: mouse press event. Otherwise:
always 0

ClipEventInitialize

UB[1]

SWF 6 and later: initialize event. Otherwise: always
0

ClipEventData

UB[1]

Data received event

Reserved

If SWF version >= 6,
UB[5]

Always 0

ClipEventConstruct

If SWF version >= 6,
UB[1]

SWF 7 and later: construct event Otherwise: always
0

49

ClipEventKeyPress

If SWF version >= 6,
UB[1]

Key press event

ClipEventDragOut

If SWF version >= 6,
UB[1]

Mouse drag out event

Reserved

If SWF version >= 6,
UB[8]

Always 0

The extra events added in SWF 6 correspond to the button movie clips in the Flash authoring tool, which are
sprites that can be scripted in the same way as buttons (see BUTTONCONDACTION). The DragOut through Press
events correspond to the button state transition events in button action conditions; the correspondence
between them is shown in the description of Button events (see “Events, state transitions, and actions”).
The KeyDown and KeyUp events are not specific to a particular key; handlers for these events are executed
whenever any key on the keyboard (with the possible exception of certain special keys) transitions to the down
state or up state, respectively. To find out what key made the transition, actions within a handler should call
methods of the ActionScript Key object.
The KeyPress event works differently from KeyDown and KeyUp. KeyPress is specific to a particular key or ASCII
character (which is specified in the CLIPACTIONRECORD). KeyPress events work in an identical way (see
BUTTONCONDACTION).

RemoveObject
The RemoveObject tag removes the specified character (at the specified depth) from the display list.
The minimum file format version is SWF 1.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 5

CharacterId

UI16

ID of character to remove

Depth

UI16

Depth of character

RemoveObject2
The RemoveObject2 tag removes the character at the specified depth from the display list. The minimum file
format version is SWF 3.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 28

50

Depth

UI16

Depth of character

ShowFrame
The ShowFrame tag instructs Flash Player to display the contents of the display list. The file is paused for the
duration of a single frame.
The minimum file format version is SWF 1.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 1

51

Chapter 4: Control Tags
Control tags manage some overall aspects of files, frames, and playback in SWF files.

SetBackgroundColor
The SetBackgroundColor tag sets the background color of the display. The minimum file format version is SWF 1.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 9

BackgroundColor

RGB

Color of the display background

FrameLabel
The FrameLabel tag gives the specified Name to the current frame. ActionGoToLabel uses this name to identify
the frame.
The minimum file format version is SWF 3.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 43

Name

STRING

Label for frame

In SWF files of version 6 or later, an extension to the FrameLabel tag called named anchors is available. A named
anchor is a special kind of frame label that, in addition to labeling a frame for seeking using ActionGoToLabel,
labels the frame for seeking using HTML anchor syntax. The browser plug-in versions of Adobe Flash Player, in
version 6 and later, will inspect the URL in the browser’s Location bar for an anchor specification (a trailing
phrase of the form
#anchorname). If an anchor specification is present in the Location bar, Flash Player will begin playback starting
at the frame that contains a FrameLabel tag that specifies a named anchor of the same name, if one exists;
otherwise playback will begin at Frame 1 as usual. In addition, when Flash Player arrives at a frame that contains
a named anchor, it will add an anchor specification with the given anchor name to the URL in the browser’s
Location bar. This ensures that when users create a bookmark at such a time, they can later return to the same
point in the SWF file, subject to the granularity at which named anchors are present within the file.
To create a named anchor, insert one additional non-null byte after the null terminator of the anchor name. This
is valid only for SWF 6 or later.
52

NamedAnchor
Field

Type

Comment

Header

RECORDHEADER

Tag type = 43

Name

Null-terminated STRING. (0 is NULL)

Label for frame.

Named Anchor flag

UI8

Always 1

Protect
The Protect tag marks a file as not importable for editing in an authoring environment. If the Protect tag
contains no data (tag length = 0), the SWF file cannot be imported. If this tag is present in the file, any authoring
tool should prevent the file from loading for editing.
If the Protect tag does contain data (tag length is not 0), the SWF file can be imported if the correct password is
specified. The data in the tag is a null-terminated string that specifies an MD5-encrypted password. Specifying a
password is only supported in SWF 5 or later.
The MD5 password encryption algorithm used was written by Poul-Henning Kamp and is freely distributable. It
resides in the FreeBSD tree at src/lib/libcrypt/crypt-md5.c. The EnableDebugger tag also uses MD5 password
encryption algorithm.
The minimum file format version is SWF 2.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 24

End
The End tag marks the end of a file. This must always be the last tag in a file. The End tag is also required to end
a sprite definition.
The minimum file format version is SWF 1.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 0

ExportAssets
The ExportAssets tag makes portions of a SWF file available for import by other SWF files (see “ImportAssets").
53

For example, ten SWF files that are all part of the same website can share an embedded custom font if one file
embeds the font and exports the font character. Each exported character is identified by a string. Any type of
character can be exported.
If the value of the character in ExportAssets was previously exported with a different identifier, Flash Player
associates the tag with the latter identifier. That is, if Flash Player has already read a given value for Tag1 and
the same Tag1 value is read later in the SWF file, the second Name1 value is used.
The minimum file format version is SWF 5.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 56

Count

UI16

Number of assets to export

Tag1

UI16

First character ID to export

Name1

STRING

Identifier for first exported character

TagN

UI16

Last character ID to export

NameN

STRING

Identifier for last exported character

ImportAssets
The ImportAssets tag imports characters from another SWF file. The importing SWF file references the exporting
SWF file by the URL where it can be found. Imported assets are added to the dictionary just like characters
defined within a SWF file.
The URL of the exporting SWF file can be absolute or relative. If it is relative, it will be resolved relative to the
location of the importing SWF file.
The ImportAssets tag must be earlier in the frame than any later tags that rely on the imported assets.
The ImportAssets tag was deprecated in SWF 8; Flash Player 8 or later ignores this tag. In SWF 8 or later, use the
ImportAssets2 tag instead.
The minimum file format version is SWF 5, and the maximum file format version is SWF 7.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 57

URL

STRING

URL where the source SWF file can be found

54

Count

UI16

Number of assets to import

Tag1

UI16

Character ID to use for first imported character in importing
SWF file (need not match character ID in exporting SWF file)

Name1

STRING

Identifier for first imported character (must match an
identifier in exporting SWF file)

TagN

UI16

Character ID to use for last imported character in importing
SWF file

NameN

STRING

Identifier for last imported character

EnableDebugger
The EnableDebugger tag enables debugging. The password in the EnableDebugger tag is encrypted by using the
MD5 algorithm, in the same way as the Protect tag.
The EnableDebugger tag was deprecated in SWF 6; Flash Player 6 or later ignores this tag because the format of
the debugging information required in the ActionScript debugger was changed in SWF 6. In SWF 6 or later, use
the EnableDebugger2 tag instead.
The minimum and maximum file format version is SWF 5.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 58

Password

Null-terminated STRING. (0 is NULL)

MD5-encrypted password

EnableDebugger2
The EnableDebugger2 tag enables debugging. The Password field is encrypted by using the MD5 algorithm, in
the same way as the Protect tag.
The minimum file format version is SWF 6.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 64

Reserved

UI16

Always 0

Password

Null-terminated STRING. (0 is NULL)

MD5-encrypted password

55

ScriptLimits
The ScriptLimits tag includes two fields that can be used to override the default settings for maximum recursion
depth and ActionScript time-out: MaxRecursionDepth and ScriptTimeoutSeconds.
The MaxRecursionDepth field sets the ActionScript maximum recursion limit. The default setting is 256 at the
time of this writing. This default can be changed to any value greater than zero (0).
The ScriptTimeoutSeconds field sets the maximum number of seconds the player should process ActionScript
before displaying a dialog box asking if the script should be stopped. The default value for ScriptTimeoutSeconds
varies by platform and is between 15 to 20 seconds. This default value is subject to change.
The minimum file format version is SWF 7.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 65

MaxRecursionDepth

UI16

Maximum recursion depth

ScriptTimeoutSeconds

UI16

Maximum ActionScript processing time before script stuck
dialog box displays

SetTabIndex
Flash Player maintains a concept of tab order of the interactive and textual objects displayed. Tab order is used
both for actual tabbing and, in SWF 6 and later, to determine the order in which objects are exposed to
accessibility aids (such as screen readers). The SWF 7 SetTabIndex tag sets the index of an object within the tab
order.
If no character is currently placed at the specified depth, this tag is ignored.
You can also use using the ActionScript tabIndex property to establish tab ordering, but this does not provide a
way to set a tab index for a static text object, because the player does not provide a scripting reflection of static
text objects. Fortunately, this is not a problem for the purpose of tabbing, because static text objects are never
actually tab stops. However, this is a problem for the purpose of accessibility ordering, because static text
objects are exposed to accessibility aids. When generating SWF content that is intended to be accessible and
contains static text objects, the SetTabIndex tag is more useful than the tabIndex property.
The minimum file format version is SWF 7.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 66

56

Depth

UI16

Depth of character

TabIndex

UI16

Tab order value

FileAttributes
The FileAttributes tag defines characteristics of the SWF file. This tag is required for SWF 8 and later and must be
the first tag in the SWF file. Additionally, the FileAttributes tag can optionally be included in all SWF file versions.
The HasMetadata flag identifies whether the SWF file contains the Metadata tag. Flash Player does not care
about this bit field or the related tag but it is useful for search engines.
The UseNetwork flag signifies whether Flash Player should grant the SWF file local or network file access if the
SWF file is loaded locally. The default behavior is to allow local SWF files to interact with local files only, and not
with the network. However, by setting the UseNetwork flag, the local SWF can forfeit its local file system access
in exchange for access to the network. Any version of SWF can use the UseNetwork flag to set the file access for
locally loaded SWF files that are running in Flash Player 8 or later.
The minimum file format version is SWF 8.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 69

Reserved

UB[1]

Must be 0

UseDirectBlit (see note
following table)

UB[1]

If 1, the SWF file uses hardware acceleration to blit
graphics to the screen, where such acceleration is
available. If 0, the SWF file will not use hardware
accelerated graphics facilities. Minimum file
version is 10.

UseGPU (see note
following table)

UB[1]

If 1, the SWF file uses GPU compositing features
when drawing graphics, where such acceleration is
available. If 0, the SWF file will not use hardware
accelerated graphics facilities. Minimum file
version is 10.

HasMetadata

UB[1]

If 1, the SWF file contains the Metadata tag. If 0,
the SWF file does not contain the Metadata tag.

ActionScript3

UB[1]

If 1, this SWF uses ActionScript 3.0. If 0, this SWF
uses ActionScript 1.0 or 2.0. Minimum file format
version is 9.

57

Reserved

UB[2]

Must be 0

UseNetwork

UB[1]

If 1, this SWF file is given network file access when
loaded locally. If 0, this SWF file is given local file
access when loaded locally.

Reserved

UB[24]

Must be 0

The UseDirectBlit and UseGPU flags are relevant only when a SWF file is playing in the standalone Flash Player.
When a SWF file plays in a web browser plug-in, UseDirectBlit is equivalent to specifying a wmode of “direct” in
the tags that embed the SWF inside the HTML page, while UseGPU is equivalent to a wmode of “gpu”.

ImportAssets2
The ImportAssets2 tag replaces the ImportAssets tag for SWF 8 and later. ImportAssets2 currently mirrors the
ImportAssets tag’s functionality.
The ImportAssets2 tag imports characters from another SWF file. The importing SWF file references the
exporting SWF file by the URL where it can be found. Imported assets are added to the dictionary just like
characters defined within a SWF file.
The URL of the exporting SWF file can be absolute or relative. If it is relative, it is resolved relative to the location
of the importing SWF file.
The ImportAssets2 tag must be earlier in the frame than any later tags that rely on the imported assets.
The minimum file format version is SWF 8.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 71

URL

STRING

URL where the source SWF file can be found

Reserved

UI8

Must be 1

Reserved

UI8

Must be 0

Count

UI16

Number of assets to import

Tag1

UI16

Character ID to use for first imported character in importing
SWF file (need not match character ID in exporting SWF file)

Name1

STRING

Identifier for first imported character (must match an identifier

58

in exporting SWF file)
TagN

UI16

Character ID to use for last imported character in importing
SWF file

NameN

STRING

Identifier for last imported character

SymbolClass
The SymbolClass tag creates associations between symbols in the SWF file and ActionScript 3.0 classes. It is the
ActionScript 3.0 equivalent of the ExportAssets tag. If the character ID is zero, the class is associated with the
main timeline of the SWF. This is how the root class of a SWF is designated. Classes listed in the SymbolClass tag
are available for creation by other SWF files (see StartSound2, DefineEditText (HasFontClass), and PlaceObject3
(PlaceFlagHasClassName and PlaceFlagHasImage). For example, ten SWF files that are all part of the same
website can share an embedded custom font if one file embeds and exports the font class.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 76

NumSymbols

UI16

Number of symbols that will be associated by this tag.

Tag1

U16

The 16-bit character tag ID for the symbol to associate

Name1

STRING

The fully-qualified name of the ActionScript 3.0 class with
which to associate this symbol. The class must have already
been declared by a DoABC tag.

...

...

...

TagN

U16

Tag ID for symbol N

NameN

STRING

Fully-qualified class name for symbol N

Metadata
The Metadata tag is an optional tag to describe the SWF file to an external process. The tag embeds XML
metadata in the SWF file so that, for example, a search engine can locate this tag, access a title for the SWF file,
and display that title in search results. Flash Player always ignores the Metadata tag.
If the Metadata tag is included in a SWF file, the FileAttributes tag must also be in the SWF file with its
HasMetadata flag set. Conversely, if the FileAttributes tag has the HasMetadata flag set, the Metadata tag must
be in the SWF file. The Metadata tag can only be in the SWF file one time.

59

The format of the metadata is RDF that is compliant with Adobe’s Extensible Metadata Platform (XMP™)
specification. For more information about RDF and XMP, see the following sources:
•

The RDF Primer at www.w3.org/TR/rdf-primer

•

The RDF Specification at www.w3.org/TR/1999/REC-rdf-syntax-19990222

•

The XMP home page at www.adobe.com/products/xmp

The following examples show two of many acceptable ways to represent the Metadata string in the SWF file.
The first example provides basic information about the SWF file, the title and description:
<rdf:RDF xmlns:rdf=’http://www.w3.org/1999/02/22-rdf-syntax-ns#’>
<rdf:Description rdf:about=’’ xmlns:dc=’http://purl.org/dc/1.1’>
<dc:title>Simple Title</dc:title>
<dc:description>Simple Description</dc:description>
</rdf:Description>
</rdf:RDF>

In the second example, the title is described for multiple languages:
<rdf:RDF xmlns:rdf=’http://www.w3.org/1999/02/22-rdf-syntax-ns#’>
<rdf:Description rdf:about=’’ xmlns:dc=’http://purl.org/dc/1.1’>
<dc:title>
<rdf:Alt>
<rdf:li xml:lang=’x-default’>Default Title</rdf:li>
<rdf:li xml:lang=’en-us’>US English Title</rdf:li>
<rdf:li xml:lang=’fr-fr’>Titre Français</rdf:li>
<rdf:li xml:lang=’it-it’>Titolo Italiano</rdf:li>
</rdf:Alt>
</dc:title>
<dc:description>Simple Description</dc:description>
</rdf:Description>
</rdf:RDF>

The Metadata string is stored in the SWF file with all unnecessary white space removed. The minimum file
format version is SWF 1.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 77

Metadata

STRING

XML Metadata

DefineScalingGrid
The DefineScalingGrid tag introduces the concept of 9-slice scaling, which allows component-style scaling to be
applied to a sprite or button character.
When the DefineScalingGrid tag associates a character with a 9-slice grid, Flash Player conceptually divides the
sprite or button into nine sections with a grid-like overlay. When the character is scaled, each of the nine areas is
60

scaled independently. To maintain the visual integrity of the character, corners are not scaled, while the
remaining areas of the image are scaled larger or smaller, as needed.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 78

CharacterId

UI16

ID of sprite or button character upon which the scaling grid
will be applied.

Splitter

RECT

Center region of 9-slice grid

The Splitter rectangle specifies the center portion of the nine regions of the scaling grid, and from this rectangle
Flash Player derives the 9-slice grid. The width and height of the rectangle must be at least one twip each (1/20
pixel), or Flash Player ignores the DefineScalingGrid tag.
When a sprite or button with a DefineScalingGrid association is scaled, the nine regions of the character scale
according to the following table:
No scale

Horizontal scale

Vertical scale

Horizontal and vertical Vertical scale
scale
Horizontal scale
No scale

No scale

No scale

9-slice scaling does not affect the children of, or any text within, the specified character. These objects
transform normally.
The sprite or button with a DefineScalingGrid association cannot be rotated or skewed, and doing so disables 9slice behavior. However, this limitation does not apply to parents or children of the 9-slice object, and parent
rotation or skew is applied to the 9-slice objects in the normal manner.
Flash Player stretches any fills in the character to fit the shape.
9-slice scaling does not affect the bounds or origin of any object.
If a 9-slice character is scaled below its original size, the five scaling regions are consumed until they become
very small. Once the minimum size is reached, Flash Player reverts to normal, non-9-slice scaling.
The minimum file format version is SWF 8.

61

DefineSceneAndFrameLabelData
The DefineSceneAndFrameLabelData tag contains scene and frame label data for a MovieClip. Scenes are
supported for the main timeline only, for all other movie clips a single scene is exported.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 86

SceneCount

EncodedU32

Number of scenes

Offset1

EncodedU32

Frame offset for scene 1

Name1

STRING

Name of scene 1

...

...

...

OffsetN

EncodedU32

Frame offset for scene N

NameN

STRING

Name of scene N

FrameLabelCount

EncodedU32

Number of frame labels

FrameNum1

EncodedU32

Frame number of frame label #1 (zero-based, global to
symbol)

FrameLabel1

STRING

Frame label string of frame label #1

...

...

...

FrameNumN

EncodedU32

Frame number of frame label #N (zero-based, global to
symbol)

FrameLabelN

STRING

Frame label string of frame label #N

62

Chapter 5: Actions
Actions are an essential part of an interactive SWF file. Actions allow a file to react to events such as mouse
movements or mouse clicks. The SWF 3 action model and earlier supports a simple action model. The SWF 4
action model supports a greatly enhanced action model that includes an expression evaluator, variables, and
conditional branching and looping. The SWF 5 action model adds a JavaScript-style object model, data types, and
functions.

SWF 3 action model
The SWF 3 action model consists of eleven instructions for Flash Player:
Instruction

See

Description

Play

ActionPlay

Start playing at the current frame

Stop

ActionStop

Stop playing at the current frame

NextFrame

ActionNextFrame

Go to the next frame

PreviousFrame

ActionPreviousFrame

Go to the previous frame

GotoFrame

ActionGotoFrame

Go to the specified frame

GotoLabel

ActionGoToLabel

Go to the frame with the specified label

WaitForFrame

ActionWaitForFrame

Wait for the specified frame

GetURL

ActionGetURL

Get the specified URL

StopSounds

ActionStopSounds

Stop all sounds playing

ToggleQuality

ActionToggleQuality

Toggle the display between high and low quality

SetTarget

ActionSetTarget

Change the context of subsequent actions to a
named object

An action (or list of actions) can be triggered by a button state transition, or by SWF 3 actions. The action is not
executed immediately, but is added to a list of actions to be processed. The list is executed on a ShowFrame tag,
or after the button state has changed. An action can cause other actions to be triggered, in which case, the
action is added to the list of actions to be processed. Actions are processed until the action list is empty.
By default, Timeline actions such as Stop (see ActionStop), Play (see ActionPlay), and GoToFrame (see
ActionGotoFrame) apply to files that contain them. However, the SetTarget action (see ActionSetTarget), which

63

is called Tell Target in the Adobe Flash user interface, can be used to send an action command to another file or
sprite (see DefineSprite).

SWF 3 actions
The actions in this section are available in SWF 3.

DoAction
DoAction instructs Flash Player to perform a list of actions when the current frame is complete. The actions are
performed when the ShowFrame tag is encountered, regardless of where in the frame the DoAction tag
appears.
Starting with SWF 9, if the ActionScript3 field of the FileAttributes tag is 1, the contents of the DoAction tag will
be ignored.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 12

Actions

ACTIONRECORD [zero or more]

List of actions to perform (see following table,
ActionRecord)

ActionEndFlag

UI8 = 0

Always set to 0

ACTIONRECORD
An ACTIONRECORD consists of an ACTIONRECORDHEADER followed by a possible data payload. The
ACTIONRECORDHEADER describes the action using an ActionCode. If the action also carries data, the
ActionCode’s high bit will be set which indicates that the ActionCode is followed by a 16-bit length and a data
payload. Note that many actions have no data payload and only consist of a single byte value.
An ACTIONRECORDHEADER has the following layout:
Field

Type

Comment

ActionCode

UI8

An action code

Length

If code >= 0x80, UI16

The number of bytes in the ACTIONRECORDHEADER, not counting
the ActionCode and Length fields.

ActionGotoFrame
ActionGotoFrame instructs Flash Player to go to the specified frame in the current file.

64

Field

Type

Comment

ActionGotoFrame

ACTIONRECORDHEADER

ActionCode = 0x81; Length is always 2

Frame

UI16

Frame index

ActionGetURL
ActionGetURL instructs Flash Player to get the URL that UrlString specifies. The URL can be of any type, including
an HTML file, an image or another SWF file. If the file is playing in a browser, the URL is displayed in the frame
that TargetString specifies. The "_level0" and "_level1" special target names are used to load another SWF file
into levels 0 and 1 respectively.
Field

Type

Comment

ActionGetURL

ACTIONRECORDHEADER

ActionCode = 0x83

UrlString

STRING

Target URL string

TargetString

STRING

Target string

ActionNextFrame
ActionNextFrame instructs Flash Player to go to the next frame in the current file.
Field

Type

Comment

ActionNextFrame

ACTIONRECORDHEADER

ActionCode = 0x04

ActionPreviousFrame
ActionPreviousFrame instructs Flash Player to go to the previous frame of the current file.
Field

Type

Comment

ActionPrevFrame

ACTIONRECORDHEADER

ActionCode = 0x05

ActionPlay
ActionPlay instructs Flash Player to start playing at the current frame.
Field

Type

Comment

ActionPlay

ACTIONRECORDHEADER

ActionCode = 0x06

65

ActionStop
ActionStop instructs Flash Player to stop playing the file at the current frame.
Field

Type

Comment

ActionStop

ACTIONRECORDHEADER

ActionCode = 0x07

ActionToggleQuality
ActionToggleQuality toggles the display between high and low quality.
Field

Type

ActionToggleQualty ACTIONRECORDHEADER

Comment
ActionCode = 0x08

ActionStopSounds
ActionStopSounds instructs Flash Player to stop playing all sounds.
Field

Type

Comment

ActionStopSounds

ACTIONRECORDHEADER

ActionCode = 0x09

ActionWaitForFrame
ActionWaitForFrame instructs Flash Player to wait until the specified frame; otherwise skips the specified
number of actions.
Field

Type

Comment

ActionWaitForFrame ACTIONRECORDHEADER

ActionCode = 0x8A; Length is always 3

Frame

UI16

Frame to wait for

SkipCount

UI8

Number of actions to skip if frame is not loaded

ActionSetTarget
ActionSetTarget instructs Flash Player to change the context of subsequent actions, so they apply to a named
object (TargetName) rather than the current file.
For example, the SetTarget action can be used to control the Timeline of a sprite object. The following sequence
of actions sends a sprite called "spinner" to the first frame in its Timeline:
1. SetTarget "spinner"
66

2. GotoFrame zero
3. SetTarget " " (empty string)
4. End of actions. (Action code = 0)
All actions following SetTarget “spinner” apply to the spinner object until SetTarget “”, which sets the action
context back to the current file. For a complete discussion of target names see DefineSprite.
Field

Type

Comment

ActionSetTarget

ACTIONRECORDHEADER

ActionCode = 0x8B

TargetName

STRING

Target of action target

ActionGoToLabel
ActionGoToLabel instructs Flash Player to go to the frame associated with the specified label. You can attach a
label to a frame with the FrameLabel tag.
Field

Type

Comment

ActionGoToLabel

ACTIONRECORDHEADER

ActionCode = 0x8C

Label

STRING

Frame label

SWF 4 action model
The SWF 4 file format supports a greatly enhanced action model that includes an expression evaluator,
variables, conditional branching and looping.
Flash Player 4 incorporates a stack machine that interprets and executes SWF 4 actions. The key SWF 4 action is
ActionPush. This action is used to push one or more parameters onto the stack. Unlike SWF 3 actions, SWF 4
actions do not have parameters embedded in the tag, rather they push parameters onto the stack, and pop
results off the stack.
The expression evaluator is also stack based. Arithmetic operators include ActionAdd, ActionSubtract,
ActionMultiply, and ActionDivide. The Flash authoring tool converts expressions to a series of stack operations.
For example, the expression 1+x*3 is represented as the following action sequence:
ActionPush "x"
ActionGetVariable
ActionPush "3"
ActionMultiply
ActionPush "1"
ActionAdd

The result of this expression is on the stack.
67

All values on the stack, including numeric values, are stored as strings. In the preceding example, the numeric
values 3 and 1 are pushed onto the stack as the strings "3" and "1".

The program counter
The current point of execution of Flash Player is called the program counter (PC). The value of the PC is defined
as the address of the action that follows the action currently being executed. Control flow actions such as
ActionJump change the value of the PC. These actions are similar to branch instructions in assembler, or the
goto instruction in other languages. For example, ActionJump tells Flash Player to jump to a new position in the
action sequence. The new PC is specified as an offset from the current PC. Both positive and negative offsets can
occur, so Flash Player can jump forward and backward in the action sequence.

SWF 4 actions
The following actions are available in SWF 4:
Type of action

Name of action

Arithmetic operators

ActionAdd ActionDivide ActionMultiply ActionSubtract

Numerical comparison

ActionEquals ActionLess

Logical operators

ActionAnd ActionNot ActionOr

String manipulation

ActionStringAdd ActionStringEquals ActionStringExtract ActionStringLength
ActionMBStringExtract ActionMBStringLength ActionStringLess

Stack operations

ActionPop ActionPush

Type conversion

ActionAsciiToChar ActionCharToAscii ActionToInteger ActionMBAsciiToChar
ActionMBCharToAscii

Control flow

ActionCall ActionIf ActionJump

Variables

ActionGetVariable ActionSetVariable

Movie control

ActionGetURL2 ActionGetProperty ActionGotoFrame2 ActionRemoveSprite
ActionSetProperty ActionSetTarget2 ActionStartDrag ActionWaitForFrame2
ActionCloneSprite ActionEndDrag

Utilities

ActionGetTime ActionRandomNumber ActionTrace

Stack operations
This section lists stack operations.

68

ActionPush
ActionPush pushes one or more values to the stack.
Field

Type

Comment

ActionPush

ACTIONRECORDHEADER

ActionCode = 0x96

Type

UI8

0 = string literal
1 = floating-point literal
The following types are available
in SWF, 5 and later:
2 = null
3 = undefined
4 = register
5 = Boolean
6 = double
7 = integer
8 = constant 8
9 = constant 16

String

If Type = 0, STRING

Null-terminated character string

Float

If Type = 1, FLOAT

32-bit IEEE single-precision littleendian floating-point value

RegisterNumber

If Type = 4, UI8

Register number

Boolean

If Type = 5, UI8

Boolean value

Double

If Type = 6, DOUBLE

64-bit IEEE double-precision littleendian double value

Integer

If Type = 7, UI32

32-bit little-endian integer

Constant8

If Type = 8, UI8

Constant pool index (for indexes <
256) (see ActionConstantPool)

Constant16

If Type = 9, UI16

Constant pool index (for indexes
>= 256) (see ActionConstantPool)

ActionPush pushes one or more values onto the stack. The Type field specifies the type of the value to be
pushed.

69

If Type = 1, the value to be pushed is specified as a 32-bit IEEE single-precision little-endian floating-point value.
PropertyIds are pushed as FLOATs. ActionGetProperty and ActionSetProperty use PropertyIds to access the
properties of named objects.
If Type = 4, the value to be pushed is a register number. Flash Player supports up to 4 registers. With the use of
ActionDefineFunction2, up to 256 registers can be used.
In the first field of ActionPush, the length in ACTIONRECORD defines the total number of Type and value bytes
that follow the ACTIONRECORD itself. More than one set of Type and value fields may follow the first field,
depending on the number of bytes that the length in ACTIONRECORD specifies.

ActionPop
ActionPop pops a value from the stack and discards it.
Field

Type

Comment

ActionPop

ACTIONRECORDHEADER

ActionCode = 0x17

ActionPop pops a value off the stack and discards the value.

Arithmetic operators
The following sections describe arithmetic operators.

ActionAdd
ActionAdd adds two numbers and pushes the result back to the stack.
Field

Type

Comment

ActionAdd

ACTIONRECORDHEADER

ActionCode = 0x0A

ActionAdd does the following:
1. Pops value A off the stack.
2. Pops value B off the stack.
3. Converts A and B to floating-point; non-numeric values evaluate to 0.
4. Adds the numbers A and B.
5. Pushes the result, A+B, to the stack.

70

ActionSubtract
ActionSubtract subtracts two numbers and pushes the result back to the stack.
Field

Type

Comment

ActionSubtract

ACTIONRECORDHEADER

ActionCode = 0x0B

ActionSubtract does the following:
1. Pops value A off the stack.
2. Pops value B off the stack.
3. Converts A and B to floating-point; non-numeric values evaluate to 0.
4. Subtracts A from B.
5. Pushes the result, B-A, to the stack.

ActionMultiply
ActionMultiply multiplies two numbers and pushes the result back to the stack.
Field

Type

Comment

ActionMultiply

ACTIONRECORDHEADER

ActionCode = 0x0C

ActionMultiply does the following:
1. Pops value A off the stack.
2. Pops value B off the stack.
3. Converts A and B to floating-point; non-numeric values evaluate to 0.
4. Multiplies A times B.
5. Pushes the result, A*B, to the stack.

71

ActionDivide
ActionDivide divides two numbers and pushes the result back to the stack.
Field

Type

Comment

ActionDivide

ACTIONRECORDHEADER

ActionCode = 0x0D

ActionDivide does the following:
1. Pops value A off the stack.
2. Pops value B off the stack.
3. Converts A and B to floating-point; non-numeric values evaluate to 0.
4. Divides B by A.
5. Pushes the result, B/A, to the stack.
6. If A is zero, the result NaN, Infinity, or -Infinity is pushed to the stack in SWF 5 and later. In SWF 4, the
result is the string #ERROR#.

Numerical comparison
ActionEquals
ActionEquals tests two numbers for equality.
Field

Type

Comment

ActionEquals

ACTIONRECORDHEADER

ActionCode = 0x0E

ActionEquals does the following:
1. Pops value A off the stack.
2. Pops value B off the stack.
3. Converts A and B to floating-point; non-numeric values evaluate to 0.
4. Compares the numbers for equality.
5. If the numbers are equal, true is pushed to the stack for SWF 5 and later.
6. For SWF 4, 1 is pushed to the stack.
72

7. Otherwise, false is pushed to the stack for SWF 5 and later. (For SWF 4, 0 is pushed to the stack.)

ActionLess
ActionLess tests if a number is less than another number
Field

Type

Comment

ActionLess

ACTIONRECORDHEADER

ActionCode = 0x0F

ActionLess does the following:
1. Pops value A off the stack.
2. Pops value B off the stack.
3. Converts A and B to floating-point; non-numeric values evaluate to 0.
4. If B < A, true is pushed to the stack for SWF 5 and later (1 is pushed for SWF 4); otherwise, false is
pushed to the stack for SWF 5 and later (0 is pushed for SWF 4).

Logical operators
ActionAnd
ActionAnd performs a logical AND of two numbers.
Field

Type

Comment

ActionAnd

ACTIONRECORDHEADER

ActionCode = 0x10

ActionAdd does the following:
1. Pops value A off the stack.
2. Pops value B off the stack.
3. Converts A and B to floating-point; non-numeric values evaluate to 0.
4. If both numbers are nonzero, true is pushed to the stack for SWF 5 and later (1 is pushed for SWF 4);
otherwise, false is pushed to the stack for SWF 5 and later (0 is pushed for SWF 4).

ActionOr
ActionOr performs a logical OR of two numbers.

73

Field

Type

Comment

ActionOr

ACTIONRECORDHEADER

ActionCode = 0x11

ActionOr does the following:
1. Pops value A off the stack.
2. Pops value B off the stack.
3. Converts A and B to floating-point; non-numeric values evaluate to 0.
4. If either of the numbers is nonzero, true is pushed to the stack for SWF 5 and later (1 is pushed for SWF
4); otherwise, false is pushed to the stack for SWF 5 and later (0 is pushed for SWF 4).

ActionNot
ActionNot performs a logical NOT of a number.
Note: In SWF 5 files, the ActionNot operator converts its argument to a Boolean value, and pushes a result of
type Boolean. In SWF 4 files, the argument and result are numbers.
Field

Type

Comment

ActionNot

ACTIONRECORDHEADER

ActionCode = 0x12

Result

Boolean

ActionNot does the following:
1. Pops a value off the stack.
2. Converts the value to floating point; non-numeric values evaluate to 0.
3. If the value is zero, true is pushed on the stack for SWF 5 and later (1 is pushed for SWF 4).
4. If the value is nonzero, false is pushed on the stack for SWF 5 and later (0 is pushed for SWF 4).

74

String manipulation
ActionStringEquals
ActionStringEquals tests two strings for equality.
Field

Type

Comment

ActionStringEquals

ACTIONRECORDHEADER

ActionCode = 0x13

ActionStringEquals does the following:
1. Pops value A off the stack.
2. Pops value B off the stack.
3. Compares A and B as strings.( The comparison is case-sensitive)
4. If the strings are equal, true is pushed to the stack for SWF 5 and later (SWF 4 pushes 1).
5. Otherwise, false is pushed to the stack for SWF 5 and later (SWF 4 pushes 0).

ActionStringLength
ActionStringLength computes the length of a string.
Field

Type

Comment

ActionStringLength

ACTIONRECORDHEADER

ActionCode = 0x14

ActionStringLength does the following:
1. Pops a string off the stack.
2. Calculates the length of the string and pushes it to the stack.

ActionStringAdd
ActionStringAdd concatenates two strings.
Field

Type

Comment

ActionStringAdd

ACTIONRECORDHEADER

ActionCode = 0x21

75

ActionStringAdd does the following:
1. Pops value A off the stack.
2. Pops value B off the stack.
3. Pushes the concatenation BA to the stack.

ActionStringExtract
ActionStringExtract extracts a substring from a string.
Field

Type

Comment

ActionStringExtract

ACTIONRECORDHEADER

ActionCode = 0x15

ActionStringExtract does the following:
1. Pops number count off the stack.
2. Pops number index off the stack.
3. Pops string string off the stack.
4. Pushes the substring of the string starting at the indexed character and count characters in length to the
stack.
5. If either index or count do not evaluate to integers, the result is the empty string.

ActionStringLess
ActionStringLess tests to see if a string is less than another string
Field

Type

Comment

ActionStringLess

ACTIONRECORDHEADER

ActionCode = 0x29

ActionStringLess does the following:
1. Pops value A off the stack.
2. Pops value B off the stack.
3. If B < A using a byte-by-byte comparison, true is pushed to the stack for SWF 5 and later (SWF 4 pushes
1); otherwise, false is pushed to the stack for SWF 5 and later (SWF 4 pushes 0).

76

ActionMBStringLength
ActionMBStringLength computes the length of a string and is multi-byte aware.
Field

Type

Comment

ActionMBStringLength

ACTIONRECORDHEADER

ActionCode = 0x31

ActionMBStringLength does the following:
1. Pops a string off the stack.
2. Calculates the length of the string in characters and pushes it to the stack.
This is a multi-byte aware version of ActionStringLength. On systems with double-byte support, a double-byte
character is counted as a single character.

ActionMBStringExtract
ActionMBStringExtract extracts a substring from a string and is multi-byte aware.
Field

Type

Comment

ActionMBStringExtract

ACTIONRECORDHEADER

ActionCode = 0x35

It does the following:
1. Pops the number count off the stack.
2. Pops the number index off the stack.
3. Pops the string string off the stack.
4. Pushes the substring of string starting at the index’th character and count characters in length to the
stack.
Note: If either index or count do not evaluate to integers, the result is the empty string.
This is a multi-byte aware version of ActionStringExtract. Index and count are treated as character indexes,
counting double-byte characters as single characters.

Type conversion
ActionToInteger
ActionToInteger converts a value to an integer.
77

Field

Type

Comment

ActionToInteger

ACTIONRECORDHEADER

ActionCode = 0x18

ActionToInteger does the following:
1. Pops a value off the stack.
2. Converts the value to a number.
3. Discards any digits after the decimal point, resulting in an integer.
4. Pushes the resulting integer to the stack.

ActionCharToAscii
ActionCharToAscii converts character code to ASCII.
Field

Type

Comment

ActionCharToAscii

ACTIONRECORDHEADER

ActionCode = 0x32

ActionCharToAscii does the following:
1. Pops a value off the stack.
2. Converts the first character of the value to a numeric ASCII character code.
3. Pushes the resulting character code to the stack.

ActionAsciiToChar
ActionAsciiToChar converts a value to an ASCII character code.
Field

Type

Comment

ActionAsciiToChar

ACTIONRECORDHEADER

ActionCode = 0x33

ActionAsciiToChar does the following:
1. Pops a value off the stack.
2. Converts the value from a number to the corresponding ASCII character.

78

3. Pushes the resulting character to the stack.

ActionMBCharToAscii
ActionMBCharToAscii converts character code to ASCII and is multi-byte aware.
Field

Type

Comment

ActionMBCharToAscii

ACTIONRECORDHEADER

ActionCode = 0x36

ActionMBCharToAscii does the following:
1. Pops a value off the stack.
2. Converts the first character of the value to a numeric character code.
If the first character of the value is a double-byte character, a 16-bit value is constructed with the first
byte as the high-order byte and the second byte as the low-order byte.
3. Pushes the resulting character code to the stack.

ActionMBAsciiToChar
ActionMBAsciiToChar converts ASCII to character code and is multi-byte aware.
Field

Type

Comment

ActionMBAsciiToChar

ACTIONRECORDHEADER

ActionCode = 0x37

ActionMBAsciiToChar does the following:
1. Pops a value off the stack.
2. Converts the value from a number to the corresponding character. If the character is a 16-bit value (>=
256), a double-byte character is constructed with the first byte containing the high-order byte, and the
second byte containing the low-order byte.
3. Pushes the resulting character to the stack.

79

Control flow
ActionJump
ActionJump creates an unconditional branch.
Field

Type

Comment

ActionJump

ACTIONRECORDHEADER

ActionCode = 0x99

BranchOffset

SI16

Offset

ActionJump adds BranchOffset bytes to the instruction pointer in the execution stream. The offset is a signed
quantity, enabling branches from –32,768 bytes to 32,767 bytes. An offset of 0 points to the action directly after
the ActionJump action.

ActionIf
ActionIf creates a conditional test and branch.
Field

Type

Comment

ActionIf

ACTIONRECORDHEADER

ActionCode = 0x9D

BranchOffset

SI16

Offset

ActionIf does the following:
1. Pops Condition, a number, off the stack.
2. Converts Condition to a Boolean value.
3. Tests if Condition is true. If Condition is true, BranchOffset bytes are added to the instruction pointer in
the execution stream.
Note: When playing a SWF 4 file, Condition is not converted to a Boolean value and is instead compared to 0,
not true.
The offset is a signed quantity, enabling branches from –32768 bytes to 32767 bytes. An offset of 0 points to the
action directly after the ActionIf action.

80

ActionCall
ActionCall calls a subroutine.
Field

Type

Comment

ActionCall

ACTIONRECORDHEADER

ActionCode = 0x9E

ActionCall does the following:
1. Pops a value off the stack. This value should be either a string that matches a frame label, or a number
that indicates a frame number. The value can be prefixed by a target string that identifies the movie clip
that contains the frame being called.
2. If the frame is successfully located, the actions in the target frame are executed. After the actions in the
target frame are executed, execution resumes at the instruction after the ActionCall instruction.
3. If the frame cannot be found, nothing happens.

Variables
ActionGetVariable
ActionGetVariable gets a variable’s value.
Field

Type

Comment

ActionGetVariable

ACTIONRECORDHEADER

ActionCode = 0x1C

ActionGetVariable does the following:
1. Pops a name off the stack, a string that names is the variable to get.
2. Pushes the value of the variable to the stack.
A variable in another execution context can be referenced by prefixing the variable name with the target path
and a colon. For example: /A/B:FOO references variable FOO in a movie clip with a target path of /A/B.

81

ActionSetVariable
ActionSetVariable sets a variable.
Field

Type

Comment

ActionSetVariable

ACTIONRECORDHEADER

ActionCode = 0x1D

ActionSetVariable does the following:
1. Pops the value off the stack.
2. Pops the name off the stack, a string which names the variable to set.
3. Sets the variable name in the current execution context to value.
A variable in another execution context can be referenced by prefixing the variable name with the target path
and a colon. For example: /A/B:FOO references the FOO variable in the movie clip with a target path of /A/B.

Movie control
ActionGetURL2
ActionGetURL2 gets a URL and is stack based.
Field

Type

Comment

ActionGetURL2

ACTIONRECORDHEADER

ActionCode = 0x9A
Length is always 1

SendVarsMethod

UB[2]

0 = None; 1 = GET
2 = POST

Reserved

UB[4]

Always 0

LoadTargetFlag

UB[1]

0 = Target is a browser window
1 = Target is a path to a sprite

LoadVariablesFlag

UB[1]

0 = No variables to load
1 = Load variables

ActionGetURL2 does the following:
1. Pops target off the stack.

82

•

A LoadTargetFlag value of 0 indicates that the target is a window. The target can be an empty string
to indicate the current window.

•

A LoadTargetFlag value of 1 indicates that the target is a path to a sprite. The target path can be in
slash or dot syntax.

2. Pops a URL off the stack; the URL specifies the URL to be retrieved.
3. SendVarsMethod specifies the method to use for the HTTP request.
•

A SendVarsMethod value of 0 indicates that this is not a form request, so the movie clip’s variables
should not be encoded and submitted.

•

A SendVarsMethod value of 1 specifies a HTTP GET request.

•

A SendVarsMethod value of 2 specifies a HTTP POST request.

4. If the SendVarsMethod value is 1 (GET) or 2 (POST), the variables in the current movie clip are submitted
to the URL by using the standard x-www-form-urlencoded encoding and the HTTP request method
specified by method.
If the LoadVariablesFlag is set, the server is expected to respond with a MIME type of application/x-www-formurlencoded and a body in the format var1=value1&var2=value2&...&varx=valuex. This response is used to
populate ActionScript variables rather than display a document. The variables populated can be in a timeline (if
LoadTargetFlag is 0) or in the specified sprite (if LoadTargetFlag is 1).
If the LoadTargetFlag is specified without the LoadVariablesFlag, the server is expected to respond with a MIME
type of application/x-shockwave-flash and a body that consists of a SWF file. This response is used to load a
subfile into the specified sprite rather than to display an HTML document.

ActionGotoFrame2
ActionGotoFrame2 goes to a frame and is stack based.
Field

Type

Comment

ActionGotoFrame2

ACTIONRECORDHEADER

ActionCode = 0x9F

Reserved

UB[6]

Always 0

SceneBiasFlag

UB[1]

Scene bias flag

Play flag

UB[1]

0 = Go to frame and stop
1 = Go to frame and play

SceneBias

If SceneBiasFlag = 1, UI16

Number to be added to
frame determined by stack

83

argument

ActionGotoFrame2 does the following:
1. Pops a frame off the stack.
•

If the frame is a number, n, the next frame of the movie to be displayed is the nth frame in the
current movie clip.

•

If the frame is a string, frame is treated as a frame label. If the specified label exists in the current
movie clip, the labeled frame will become the current frame. Otherwise, the action is ignored.

2. Either a frame or a number can be prefixed by a target path, for example, /MovieClip:3 or
/MovieClip:FrameLabel.
3. If the Play flag is set, the action goes to the specified frame and begins playing the enclosing movie clip.
Otherwise, the action goes to the specified frame and stops.

ActionSetTarget2
ActionSetTarget2 sets the current context and is stack based.
Field

Type

Comment

ActionSetTarget2

ACTIONRECORDHEADER

ActionCode = 0x20

ActionSetTarget2 pops the target off the stack and makes it the current active context.
This action behaves exactly like ActionSetTarget but is stack based to enable the target path to be the result of
expression evaluation.

ActionGetProperty
ActionGetProperty gets a file property.
Field

Type

Comment

ActionGetProperty

ACTIONRECORDHEADER

ActionCode = 0x22

ActionGetProperty does the following:
1. Pops index off the stack.

84

2. Pops target off the stack.
3. Retrieves the value of the property enumerated as index from the movie clip with target path target and
pushes the value to the stack.
The following table lists property index values. The _quality, _xmouse and _ymouse properties are available in
SWF 5 and later.
Property

Value

_X

0

_Y

1

_xscale

2

_yscale

3

_currentframe

4

_totalframes

5

_alpha

6

_visible

7

_width

8

_height

9

_rotation

10

_target

11

_framesloaded

12

_name

13

_droptarget

14

_url

15

_highquality

16

_focusrect

17

_soundbuftime

18

85

_quality

19

_xmouse

20

_ymouse

21

ActionSetProperty
ActionSetProperty sets a file property.
Field

Type

Comment

ActionSetProperty

ACTIONRECORDHEADER

ActionCode = 0x23

ActionSetProperty does the following:
1. Pops a value off the stack.
2. Pops an index off the stack.
3. Pops a target off the stack.
4. Sets the property enumerated as index in the movie clip with the target path target to the value value.

ActionCloneSprite
ActionCloneSprite clones a sprite.
Field

Type

Comment

ActionCloneSprite

ACTIONRECORDHEADER

ActionCode = 0x24

ActionCloneSprite does the following:
1. Pops a depth off the stack.
2. Pops a target off the stack.
3. Pops a source off the stack.
4. Duplicates the movie clip source, giving the new instance the name target, at z-order depth

86

ActionRemoveSprite
ActionRemoveSprite removes a clone sprite.
Field

Type

Comment

ActionRemoveSprite

ACTIONRECORDHEADER

ActionCode = 0x25

ActionRemoveSprite does the following:
1. Pops a target off the stack.
2. Removes the clone movie clip that the target path target identifies.

ActionStartDrag
ActionStartDrag starts dragging a movie clip.
Field

Type

Comment

ActionStartDrag

ACTIONRECORDHEADER

ActionCode = 0x27

ActionStartDrag does the following:
1. Pops a target off the stack; target identifies the movie clip to be dragged.
2. Pops lockcenter off the stack. If lockcenter evaluates to a nonzero value, the center of the dragged
movie clip is locked to the mouse position. Otherwise, the movie clip moves relative to the mouse
position when the drag started.
3. Pops constrain off the stack.
4. If constrain evaluates to a nonzero value:

87

•

Pops y2 off the stack.

•

Pops x2 off the stack.

•

Pops y1 off the stack.

•

Pops x1 off the stack.

ActionEndDrag
ActionEndDrag ends the drag operation in progress, if any.
Field

Type

Comment

ActionEndDrag

ACTIONRECORDHEADER

ActionCode = 0x28

ActionWaitForFrame2
ActionWaitForFrame2 waits for a frame to be loaded and is stack based.
Field

Type

Comment

ActionWaitForFrame2

ACTIONRECORDHEADER

ActionCode = 0x8D; Length is always 1

SkipCount

UI8

The number of actions to skip

ActionWaitForFrame2 does the following:
1. Pops a frame off the stack.
2. If the frame is loaded, skip the next n actions that follow the current action, where n is indicated by
SkipCount.
The frame is evaluated in the same way as ActionGotoFrame2.

Utilities
ActionTrace
ActionTrace sends a debugging output string.
Field

Type

Comment

ActionTrace

ACTIONRECORDHEADER

ActionCode = 0x26

ActionTrace does the following:
1. Pops a value off the stack.
2. In the Test Movie mode of the Adobe Flash editor, ActionTrace appends a value to the output window if
the debugging level is not set to None.
In Adobe Flash Player, nothing happens.
88

ActionGetTime
ActionGetTime reports the milliseconds since Adobe Flash Player started.
Field

Type

Comment

ActionGetTime

ACTIONRECORDHEADER

ActionCode = 0x34

ActionGetTime does the following:
1. Calculates as an integer the number of milliseconds since Flash Player was started.
2. Pushes the number to the stack.

ActionRandomNumber
ActionRandomNumber calculates a random number.
Field

Type

Comment

ActionRandomNumber

ACTIONRECORDHEADER

ActionCode = 0x30

ActionRandomNumber does the following:
1. Pops the maximum off the stack.
2. Calculates a random number as an integer in the range 0…(maximum-1).
3. Pushes the random number to the stack.

SWF 5 action model
SWF 5 is similar to SWF 4. New actions greatly expand ActionScript functionality. There are also new type
conversion, math and stack operator actions.

SWF 5 actions
Following is an overview of SWF 5 actions:
Type of action

Name of action

ScriptObject actions

ActionCallFunction
ActionCallMethod
ActionConstantPool

89

ActionDefineFunction
ActionDefineLocal
ActionDefineLocal2
ActionDelete
ActionDelete2
ActionEnumerate
ActionEquals2
ActionGetMember
ActionInitArray
ActionInitObject
ActionNewMethod
ActionNewObject
ActionSetMember
ActionTargetPath
ActionWith
Type actions

ActionToNumber
ActionToString
ActionTypeOf

Math actions

ActionAdd2
ActionLess2
ActionModulo

Stack operator actions

ActionBitAnd
ActionBitLShift
ActionBitOr
ActionBitRShift
ActionBitURShift
ActionBitXor ActionDecrement
ActionIncrement
ActionPush (Enhancements)
ActionPushDuplicate ActionReturn
ActionStackSwap
ActionStoreRegister

ScriptObject actions
ActionCallFunction
ActionCallFunction executes a function. The function can be an ActionScript built-in function (such as parseInt), a
user-defined ActionScript function, or a native function. For more information, see ActionNewObject.
90

Field

Type

Comment

ActionCallFunction

ACTIONRECORDHEADER

ActionCode = 0x3D

ActionCallFunction does the following:
1. Pops the function name (String) from the stack.
2. Pops numArgs (int) from the stack.
3. Pops the arguments off the stack.
4. Invokes the function, passing the arguments to it.
5. Pushes the return value of the function invocation to the stack.
If no appropriate return value is present (that is, the function does not have a return statement), a push
undefined message is generated by the compiler and is pushed to the stack. The undefined return value
should be popped off the stack.
For all of the call actions (ActionCallMethod, ActionNewMethod, ActionNewObject, and ActionCallFunction) and
initialization actions (ActionInitObject and ActionInitArray), the arguments of the function are pushed onto the
stack in reverse order, with the rightmost argument first and the leftmost argument last. The arguments are
subsequently popped off in order (first to last).

ActionCallMethod
ActionCallMethod pushes a method (function) call onto the stack, similar to ActionNewMethod.
Field

Type

Comment

ActionCallMethod

ACTIONRECORDHEADER

ActionCode = 0x52

If the named method exists, ActionCallMethod does the following:
1. Pops the name of the method from the stack. If the method name is blank or undefined, the object is
taken to be a function object that should be invoked, rather than the container object of a method. For
example, if CallMethod is invoked with object obj and method name blank, it's equivalent to using the
syntax:
obj();
If a method’s name is foo, it's equivalent to:
obj.foo();
91

2. Pops the ScriptObject, object, from the stack.
3. Pops the number of arguments, args, from the stack.
4. Pops the arguments off the stack.
5. Executes the method call with the specified arguments.
6. Pushes the return value of the method or function to the stack.
If no appropriate return value is present (the function does not have a return statement), a push
undefined is generated by the compiler and is pushed to the stack. The undefined return value should
be popped off the stack.
For all of the call actions (ActionCallMethod, ActionNewMethod, ActionNewObject, and ActionCallFunction) and
initialization actions (ActionInitObject and ActionInitArray), the arguments of the function are pushed onto the
stack in reverse order, with the rightmost argument first and the leftmost argument last. The arguments are
subsequently popped off in order (first to last).

ActionConstantPool
ActionConstantPool creates a new constant pool, and replaces the old constant pool if one already exists.
Field

Type

Comment

ActionConstantPool

ACTIONRECORDHEADER

ActionCode = 0x88

Count

UI16

Number of constants to follow

ConstantPool

STRING[Count]

String constants

ActionDefineFunction
Note: ActionDefineFunction is rarely used as of SWF 7 and later; it was superseded by ActionDefineFunction2.
ActionDefineFunction defines a function with a given name and body size.
Field

Type

Comment

ActionDefineFunction

ACTIONRECORDHEADER

ActionCode = 0x9B

FunctionName

STRING

Function name, empty if anonymous

NumParams

UI16

# of parameters

param 1

STRING

Parameter name 1

param 2

STRING

Parameter name 2

92

...
param N

STRING

Parameter name N

codeSize

UI16

# of bytes of code that follow

ActionDefineFunction parses (in order) FunctionName, NumParams, [param1, param2, …, param N] and then
code size.
ActionDefineFunction does the following:
1. Parses the name of the function (name) from the action tag.
2. Skips the parameters in the tag.
3. Parses the code size from the tag. After the DefineFunction tag, the next codeSize bytes of action data are
considered to be the body of the function.
4. Gets the code for the function.
ActionDefineFunction can be used in the following ways:
Usage 1 Pushes an anonymous function on the stack that does not persist. This function is a function literal that
is declared in an expression instead of a statement. An anonymous function can be used to define a function,
return its value, and assign it to a variable in one expression, as in the following ActionScript:
area = (function () {return Math.PI * radius *radius;})(5);

Usage 2 Sets a variable with a given FunctionName and a given function definition. This is the more
conventional function definition. For example, in ActionScript:
function Circle(radius) {
this.radius = radius;
this.area = Math.PI * radius * radius;
}

ActionDefineLocal
ActionDefineLocal defines a local variable and sets its value. If the variable already exists, the value is set to the
newly specified value.
Field

Type

Comment

ActionDefineLocal

ACTIONRECORDHEADER

ActionCode = 0x3C

93

ActionDefineLocal does the following:
1. Pops a value off the stack.
2. Pops a name off the stack.

ActionDefineLocal2
ActionDefineLocal2 defines a local variable without setting its value. If the variable already exists, nothing
happens. The initial value of the local variable is undefined.
Field

Type

Comment

ActionDefineLocal2

ACTIONRECORDHEADER

ActionCode = 0x41

ActionDefineLocal2 pops name off the stack.

ActionDelete
ActionDelete deletes a named property from a ScriptObject.
Field

Type

Comment

ActionDelete

ACTIONRECORDHEADER

ActionCode = 0x3A

ActionDelete does the following:
1. Pops the name of the property to delete off the stack.
2. Pops the object to delete the property from.

ActionDelete2
ActionDelete2 deletes a named property. Flash Player first looks for the property in the current scope, and if the
property cannot be found, continues to search in the encompassing scopes.
Field

Type

Comment

ActionDelete2

ACTIONRECORDHEADER

ActionCode = 0x3B

ActionDelete2 pops the name of the property to delete off the stack.

94

ActionEnumerate
ActionEnumerate obtains the names of all “slots” in use in an ActionScript object—that is, for an object obj, all
names X that could be retrieved with the syntax obj.X. ActionEnumerate is used to implement the for..in
statement in ActionScript.
Note: Certain special slot names are omitted; for a list of these, search for the term DontEnum in the ECMA-262
standard.
Field

Type

Comment

ActionEnumerate

ACTIONRECORDHEADER

ActionCode = 0x46

ActionEnumerate does the following:
1. Pops the name of the object variable (which can include slash-path or dot-path syntax) off of the stack.
2. Pushes a null value onto the stack to indicate the end of the slot names.
3. Pushes each slot name (a string) onto the stack.
The order in which slot names are pushed is undefined.

ActionEquals2
ActionEquals2 is similar to ActionEquals, but ActionEquals2 knows about types. The equality comparison
algorithm from ECMA-262 Section 11.9.3 is applied.
Field

Type

Comment

ActionEquals2

ACTIONRECORDHEADER

ActionCode = 0x49

ActionEquals2 does the following:
1. Pops arg1 off the stack.
2. Pops arg2 off the stack.
3. Pushes the return value to the stack.

95

ActionGetMember
ActionGetMember retrieves a named property from an object, and pushes the value of the property onto the
stack.
Field

Type

Comment

ActionGetMember

ACTIONRECORDHEADER

ActionCode = 0x4E

ActionGetMember does the following:
1. Pops the name of the member function.
2. Pops the ScriptObject object off of the stack.
3. Pushes the value of the property on to the stack.
For example, assume obj is an object, and it is assigned a property, foo, as follows:
obj.foo = 3;

Then, ActionGetMember with object set to obj and name set to foo pushes 3 onto the stack. If the specified
property does not exist, undefined is pushed to the stack.
The object parameter cannot actually be of type Object. If the object parameter is a primitive type such as
number, Boolean, or string, it is converted automatically to a temporary wrapper object of class Number,
Boolean, or String. Thus, methods of wrapper objects can be invoked on values of primitive types. For example,
the following correctly prints 5:
var x = "Hello";
trace (x.length);

In this case, the variable, x, contains the primitive string, "Hello". When x.length is retrieved, a temporary
wrapper object for x is created by using the String type, which has a length property.

ActionInitArray
ActionInitArray initializes an array in a ScriptObject and is similar to ActionInitObject. The newly created object is
pushed to the stack. The stack is the only existing reference to the newly created object. A subsequent
SetVariable or SetMember action can store the newly created object in a variable.
Field

Type

Comment

ActionInitArray

ACTIONRECORDHEADER

ActionCode = 0x42

96

ActionInitArray pops elems and then [arg1, arg2, …, argn] off the stack. ActionInitArray does the following:
1. Gets the number of arguments (or elements) from the stack.
2. If arguments are present, ActionInitArray initializes an array object with the right number of elements.
3. Initializes the array as a ScriptObject.
4. Sets the object type to Array.
5. Populates the array with initial elements by popping the values off of the stack.
For all of the call actions (ActionCallMethod, ActionNewMethod, ActionNewObject, and ActionCallFunction) and
initialization actions (ActionInitObject and ActionInitArray), the arguments of the function are pushed onto the
stack in reverse order, with the rightmost argument first and the leftmost argument last. The arguments are
subsequently popped off in order (first to last).

ActionInitObject
ActionInitObject initializes an object and is similar to ActionInitArray. The newly created object is pushed to the
stack. The stack is the only existing reference to the newly created object. A subsequent SetVariable or
SetMember action can store the newly created object in a variable.
Field

Type

Comment

ActionInitObject

ACTIONRECORDHEADER

ActionCode = 0x43

ActionInitObject pops elems off of the stack. Pops [value1, name1, …, valueN, nameN] off the stack.
ActionInitObject does the following:
1. Pops the number of initial properties from the stack.
2. Initializes the object as a ScriptObject.
3. Sets the object type to Object.
4. Pops each initial property off the stack.
For each initial property, the value of the property is popped off the stack, then the name of the
property is popped off the stack. The name of the property is converted to a string. The value can be of
any type.
For all of the call actions (ActionCallMethod, ActionNewMethod, ActionNewObject, and ActionCallFunction) and
initialization actions (ActionInitObject and ActionInitArray), the arguments of the function are pushed onto the
stack in reverse order, with the rightmost argument first and the leftmost argument last. The arguments are
subsequently popped off in order (first to last).
97

ActionNewMethod
ActionNewMethod invokes a constructor function to create a new object. A new object is constructed and
passed to the constructor function as the value of the this keyword. Arguments can be specified to the
constructor function. The return value from the constructor function is discarded. The newly constructed object
is pushed to the stack, similar to ActionCallMethod and ActionNewObject.
Field

Type

Comment

ActionNewMethod

ACTIONRECORDHEADER

ActionCode = 0x53

ActionNewMethod does the following:
1. Pops the name of the method from the stack.
2. Pops the ScriptObject from the stack. If the name of the method is blank, the ScriptObject is treated as a
function object that is invoked as the constructor function. If the method name is not blank, the named
method of the ScriptObject is invoked.
3. Pops the number of arguments from the stack.
4. Executes the method call.
5. Pushes the newly constructed object to the stack. If no appropriate return value occurs (for instance, the
function does not have a return statement), the compiler generates a push undefined and pushes it to
the stack. The undefined return value should be popped off the stack.
For all of the call actions (ActionCallMethod, ActionNewMethod, ActionNewObject, and ActionCallFunction) and
initialization actions (ActionInitObject and ActionInitArray), the arguments of the function are pushed onto the
stack in reverse order, with the rightmost argument first and the leftmost argument last. The arguments are
subsequently popped off in order (first to last).

ActionNewObject
ActionNewObject invokes a constructor function. A new object is created and passed to the constructor function
as the this keyword. In addition, arguments can optionally be specified to the constructor function on the stack.
The return value of the constructor function is discarded. The newly constructed object is pushed to the stack.
ActionNewObject is similar to ActionCallFunction and ActionNewMethod.
Field

Type

Comment

ActionNewObject

ACTIONRECORDHEADER

ActionCode = 0x40

98

ActionNewObject does the following:
1. Pops the object name (STRING) this from the stack.
2. Pops numArgs (int) from the stack.
3. Pops the arguments off the stack.
4. Invokes the named object as a constructor function, passing it the specified arguments and a newly
constructed object as the this keyword.
5. The return value of the constructor function is discarded.
6. The newly constructed object is pushed to the stack.
For all of the call actions (ActionCallMethod, ActionNewMethod, ActionNewObject, and ActionCallFunction) and
initialization actions (ActionInitObject and ActionInitArray), the arguments of the function are pushed onto the
stack in reverse order, with the rightmost argument first and the leftmost argument last. The arguments are
subsequently popped off in order (first to last).

ActionSetMember
ActionSetMember sets a property of an object. If the property does not already exist, it is created. Any existing
value in the property is overwritten.
Field

Type

Comment

ActionSetMember

ACTIONRECORDHEADER

ActionCode = 0x4F

ActionSetMember does the following:
1. Pops the new value off the stack.
2. Pops the object name off the stack.
3. Pops the object off of the stack.

ActionTargetPath
If the object in the stack is of type MovieClip, the object’s target path is pushed on the stack in dot notation. If
the object is not a MovieClip, the result is undefined rather than the movie clip target path.
Field

Type

Comment

ActionTargetPath

ACTIONRECORDHEADER

ActionCode = 0x45

99

ActionTargetPath does the following:
1. Pops the object off the stack.
2. Pushes the target path onto the stack.

ActionWith
Defines a With block of script.
Field

Type

Comment

ActionWith

ACTIONRECORDHEADER

ActionCode = 0x94

Size

UI16

# of bytes of code that follow

ActionWith does the following:
1. Pops the object involved with the With.
2. Parses the size (body length) of the With block from the ActionWith tag.
3. Checks to see if the depth of calls exceeds the maximum depth, which is 16 for SWF 6 and later, and 8
for SWF 5. If the With depth exceeds the maximum depth, the next Size bytes of data are skipped rather
than executed.
4. After the ActionWith tag, the next Size bytes of action codes are considered to be the body of the With
block.
5. Adds the With block to the scope chain.

Type actions
ActionToNumber
Converts the object on the top of the stack into a number, and pushes the number back to the stack.
For the Object type, the ActionScript valueOf() method is invoked to convert the object to a Number type for
ActionToNumber. Conversions between primitive types, such as from String to Number, are built-in.
Field

Type

Comment

ActionToNumber

ACTIONRECORDHEADER

ActionCode = 0x4A

100

ActionToNumber does the following:
1. Pops the object off of the stack.
2. Pushes the number on to the stack.

ActionToString
ActionToString converts the object on the top of the stack into a String, and pushes the string back to the stack.
For the Object type, the ActionScript toString() method is invoked to convert the object to the String type for
ActionToString.
Field

Type

Comment

ActionToString

ACTIONRECORDHEADER

ActionCode = 0x4B

ActionToString does the following:
1. Pops the object off of the stack.
2. Pushes the string on to the stack.

ActionTypeOf
ActionTypeOf pushes the object type to the stack, which is equivalent to the ActionScript TypeOf() method. The
possible types are:
number
boolean
string
object
movieclip
null
undefined
function

Field

Type

Comment

ActionTypeOf

ACTIONRECORDHEADER

ActionCode = 0x44

ActionTypeOf does the following:
1. Pops the value to determine the type of off the stack.
2. Pushes a string with the type of the object on to the stack.

101

Math actions
ActionAdd2
ActionAdd2 is similar to ActionAdd, but performs the addition differently, according to the data types of the
arguments. The addition operator algorithm in ECMA-262 Section 11.6.1 is used. If string concatenation is
applied, the concatenated string is arg2 followed by arg1.
Field

Type

Comment

ActionAdd2

ACTIONRECORDHEADER

ActionCode = 0x47

ActionAdd2 does the following:
1. Pops arg1 off of the stack.
2. Pops arg2 off of the stack.
3. Pushes the result back to the stack.

ActionLess2
ActionLess2 calculates whether arg1 is less than arg2 and pushes a Boolean return value to the stack. This action
is similar to ActionLess, but performs the comparison differently according to the data types of the arguments.
The abstract relational comparison algorithm in ECMA-262 Section 11.8.5 is used.
Field

Type

Comment

ActionLess2

ACTIONRECORDHEADER

ActionCode = 0x48

ActionLess2 does the following:
1. Pops arg1 off of the stack.
2. Pops arg2 off of the stack.
3. Compares arg2 < arg1.
4. Pushes the return value (a Boolean value) onto the stack.

102

ActionModulo
ActionModulo calculates x modulo y. If y is 0, then NaN (0x7FC00000) is pushed to the stack.
Field

Type

Comment

ActionModulo

ACTIONRECORDHEADER

ActionCode = 0x3F

ActionModulo does the following:
1. Pops x then y off of the stack.
2. Pushes the value x % y on to the stack.

Stack operator actions
ActionBitAnd
ActionBitAnd pops two numbers off of the stack, performs a bitwise AND, and pushes an S32 number to the
stack. The arguments are converted to 32-bit unsigned integers before performing the bitwise operation. The
result is a SIGNED 32-bit integer.
Field

Type

ActionBitAnd

ACTIONRECORDHEADER
ActionCode = 0x60

Comment

ActionBitAnd does the following:
1. Pops arg1 then arg2 off of the stack.
2. Pushes the result to the stack.

ActionBitLShift
ActionBitLShift pops the shift count arg and then value off of the stack. The value argument is converted to 32bit signed integer and only the least significant 5 bits are used as the shift count. The bits in the value arg are
shifted to the left by the shift count. ActionBitLShift pushes an S32 number to the stack.
Field

Type

Comment

ActionBitLShift

ACTIONRECORDHEADER

ActionCode = 0x63

103

ActionBitLShift does the following:
1. Pops shift count arg, then value off of the stack.
2. Pushes the result to the stack.

ActionBitOr
ActionBitOr pops two numbers off of the stack, performs a bitwise OR, and pushes an S32 number to the stack.
The arguments are converted to 32-bit unsigned integers before performing the bitwise operation. The result is
a SIGNED 32-bit integer.
Field

Type

Comment

ActionBitOr

ACTIONRECORDHEADER

ActionCode = 0x61

ActionBitOr does the following:
1. Pops arg1 then arg2 off of the stack.
2. Pushes the result to the stack.

ActionBitRShift
ActionBitRShift pops the shift count from the stack. Pops the value from the stack. The value argument is
converted to a 32-bit signed integer and only the least significant 5 bits are used as the shift count.
The bits in the arg value are shifted to the right by the shift count. ActionBitRShift pushes an S32 number to the
stack.
Field

Type

Comment

ActionBitRShift

ACTIONRECORDHEADER

ActionCode = 0x64

ActionBitRShift does the following:
1. Pops the shift count from the stack.
2. Pops the value to shift from the stack.
3. Pushes the result to the stack.

104

ActionBitURShift
ActionBitURShift pops the value and shift count arguments from the stack. The value argument is converted to
32-bit signed integer and only the least significant 5 bits are used as the shift count.
The bits in the arg value are shifted to the right by the shift count. ActionBitURShift pushes a UI32 number to the
stack.
Field

Type

Comment

ActionBitURShift

ACTIONRECORDHEADER

ActionCode = 0x65

ActionBitURShift does the following:
1. Pops the shift count from the stack.
2. Pops the value to shift from the stack.
3. Pushes the result to the stack.

ActionBitXor
ActionBitXor pops two numbers off of the stack, performs a bitwise XOR, and pushes an S32 number to the
stack.
The arguments are converted to 32-bit unsigned integers before performing the bitwise operation. The result is
a SIGNED 32-bit integer.
Field

Type

Comment

ActionBitXor

ACTIONRECORDHEADER

ActionCode = 0x62

ActionBitXor does the following:
1. Pops arg1 and arg2 off of the stack.
2. Pushes the result back to the stack.

105

ActionDecrement
ActionDecrement pops a value from the stack, converts it to number type, decrements it by 1, and pushes it
back to the stack.
Field

Type

Comment

ActionDecrement

ACTIONRECORDHEADER

ActionCode = 0x51

ActionDecrement does the following:
1. Pops the number off of the stack.
2. Pushes the result on to the stack.

ActionIncrement
ActionIncrement pops a value from the stack, converts it to number type, increments it by 1, and pushes it back
to the stack.
Field

Type

Comment

ActionIncrement

ACTIONRECORDHEADER

ActionCode = 0x50

ActionIncrement does the following:
1. Pops the number off of the stack.
2. Pushes the result on to the stack.

ActionPush (Enhancements)
With SWF 5, eight new types were added to ActionPush. For more on ActionPush, see the SWF 4 actions.

ActionPushDuplicate
ActionPushDuplicate pushes a duplicate of top of stack (the current return value) to the stack.
Field

Type

Comment

ActionPushDuplicate

ACTIONRECORDHEADER

ActionCode = 0x4C

106

ActionReturn
ActionReturn forces the return item to be pushed off the stack and returned. If a return is not appropriate, the
return item is discarded.
Field

Type

Comment

ActionReturn

ACTIONRECORDHEADER

ActionCode = 0x3E

ActionReturn pops a value off the stack.

ActionStackSwap
ActionStackSwap swaps the top two ScriptAtoms on the stack.
Field

Type

Comment

ActionStackSwap

ACTIONRECORDHEADER

ActionCode = 0x4D

ActionStackSwap does the following:
1. Pops Item1 and then Item2 off of the stack.
2. Pushes Item1 and then Item2 back to the stack.

ActionStoreRegister
ActionStoreRegister reads the next object from the stack (without popping it) and stores it in one of four
registers. If ActionDefineFunction2 is used, up to 256 registers are available.
Field

Type

Comment

ActionStoreRegister

ACTIONRECORDHEADER

ActionCode = 0x87

RegisterNumber

UI8

ActionStoreRegister parses register number from the StoreRegister tag.

107

SWF 6 action model
SWF 6 adds the DoInitAction action-definition tag, and a few new action bytecodes.

SWF 6 actions
The following actions are available in SWF 6:
•

DoInitAction

•

ActionInstanceOf

•

ActionEnumerate2

•

ActionStrictEquals

•

ActionGreater

•

ActionStringGreater

DoInitAction
The DoInitAction tag is similar to the DoAction tag: it defines a series of bytecodes to be executed. However, the
actions defined with DoInitAction are executed earlier than the usual DoAction actions, and are executed only
once.
In some situations, actions must be executed before the ActionScript representation of the first instance of a
particular sprite is created. The most common such action is calling Object.registerClass to associate an
ActionScript class with a sprite. Such a call is generally found within the #initclip pragma in the ActionScript
language. DoInitAction is used to implement the #initclip pragma.
A DoInitAction tag specifies a particular sprite to which its actions apply. A single frame can contain multiple
DoInitAction tags; their actions are executed in the order in which the tags appear. However, the SWF file can
contain only one DoInitAction tag for any particular sprite.
The specified actions are executed immediately before the normal actions of the frame in which the
DoInitAction tag appears. This only occurs the first time that this frame is encountered; playback reaches the
same frame again later, actions provided in DoInitAction are skipped.
Starting with SWF 9, if the ActionScript3 field of the FileAttributes tag is 1, the contents of the DoInitAction tag
will be ignored.
Note: Specifying actions at the beginning of a DoAction tag is not the same as specifying them in a DoInitAction
tag. Flash Player takes steps before the first action in a DoAction tag, most relevantly the creation of
ActionScript objects that represent sprites. The actions in DoInitAction occur before these implicit steps are
performed.
108

Field

Type

Comment

Header

RECORDHEADER

Tag type = 59

Sprite ID

UI16

Sprite to which these actions apply

Actions

ACTIONRECORD[zero or more]

List of actions to perform

ActionEndFlag

UI8

Always set to 0

ActionInstanceOf
ActionInstanceOf implements the ActionScript instanceof() operator. This is a Boolean operator that indicates
whether the left operand (typically an object) is an instance of the class represented by a constructor function
passed as the right operand.
Additionally, with SWF 7 or later, ActionInstanceOf also supports with interfaces. If the right operand
constructor is a reference to an interface object, and the left operand implements this interface,
ActionInstanceOf accurately reports that the left operand is an instance of the right operand.
Field

Type

Comment

ActionInstanceOf

ACTIONRECORDHEADER

ActionCode = 0x54

ActionInstanceOf does the following:
1. Pops constr then obj off of the stack.
2. Determines if obj is an instance of constr.
3. Pushes the return value (a Boolean value) onto the stack.

ActionEnumerate2
ActionEnumerate2 is similar to ActionEnumerate, but uses a stack argument of object type rather than using a
string to specify its name.
Field

Type

Comment

ActionEnumerate2

ACTIONRECORDHEADER

ActionCode = 0x55

ActionEnumerate2 does the following:
1. Pops obj off of the stack.
109

2. Pushes a null value onto the stack to indicate the end of the slot names.
3. Pushes each slot name (a string) from obj onto the stack.
Note: The order in which slot names are pushed is undefined.

ActionStrictEquals
ActionStrictEquals is similar to ActionEquals2, but the two arguments must be of the same type in order to be
considered equal. Implements the ‘===’ operator from the ActionScript language.
Field

Type

Comment

ActionStrictEquals

ACTIONRECORDHEADER

ActionCode = 0x66

ActionStrictEquals does the following:
1. Pops arg1 then arg2 off the stack.
2. Pushes the return value, a Boolean value, to the stack.

ActionGreater
ActionGreater is the exact opposite of ActionLess2. Originally there was no ActionGreater, because it can be
emulated by reversing the order of argument pushing, then performing an ActionLess followed by an ActionNot.
However, this argument reversal resulted in a reversal of the usual order of evaluation of arguments, which in a
few cases led to surprises.
Field

Type

Comment

ActionGreater

ACTIONRECORDHEADER

ActionCode = 0x67

ActionGreater does the following:
1. Pops arg1 and then arg2 off of the stack.
2. Compares if arg2 > arg1.
3. Pushes the return value, a Boolean value, onto the stack.

ActionStringGreater
ActionStringGreater is the exact opposite of ActionStringLess. This action code was added for the same reasons
as ActionGreater.

110

Field

Type

Comment

ActionStringGreater

ACTIONRECORDHEADER

ActionCode = 0x68

ActionStringGreater does the following:
1. Pops arg1 and then arg2 off of the stack.
2. Compares if arg2 > arg1, using byte-by-byte comparison.
3. Pushes the return value, a Boolean value, onto the stack.

SWF 7 action model
SWF 7 actions
The following actions are available in SWF 7:
•

ActionDefineFunction2

•

ActionExtends

•

ActionCastOp

•

ActionImplementsOp

•

ActionTry

•

ActionThrow

ActionDefineFunction2
ActionDefineFunction2 is similar to ActionDefineFunction, with additional features that can help speed up the
execution of function calls by preventing the creation of unused variables in the function’s activation object and
by enabling the replacement of local variables with a variable number of registers. With ActionDefineFunction2,
a function can allocate its own private set of up to 256 registers. Parameters or local variables can be replaced
with a register, which is loaded with the value instead of the value being stored in the function’s activation
object. (The activation object is an implicit local scope that contains named arguments and local variables. For
further description of the activation object, see the ECMA-262 standard.)
ActionDefineFunction2 also includes six flags to instruct Flash Player to preload variables, and three flags to
suppress variables. By setting PreloadParentFlag, PreloadRootFlag, PreloadSuperFlag, PreloadArgumentsFlag,
PreloadThisFlag, or PreloadGlobalFlag, common variables can be preloaded into registers before the function
executes (_parent, _root, super, arguments, this, or _global, respectively). With flags SuppressSuper,
111

SuppressArguments, and SuppressThis, common variables super, arguments, and this are not created. By using
suppress flags, Flash Player avoids pre- evaluating variables, thus saving time and improving performance.
No suppress flags are provided for _parent, _root, or _global because Flash Player always evaluates these
variables as needed; no time is ever wasted on pre-evaluating these variables.
Specifying both the preload flag and the suppress flag for any variable is not allowed.
The body of the function that ActionDefineFunction2 specifies should use ActionPush and ActionStoreRegister
for local variables that are assigned to registers. ActionGetVariable and ActionSetVariable cannot be used for
variables assigned to registers.
Flash Player 6 release 65 and later supports ActionDefineFunction2.
Field

Type

Comment

ActionDefineFunction2

ACTIONRECORDHEADER

ActionCode = 0x8E

FunctionName

STRING

Name of function, empty if
anonymous

NumParams

UI16

# of parameters

RegisterCount

UI8

Number of registers to allocate,
(from 0 to 254) up to 255 registers

PreloadParentFlag

UB[1]

0 = Don’t preload _parent into
register
1 = Preload _parent into register

PreloadRootFlag

UB[1]

0 = Don’t preload _root into
register
1 = Preload _root into register

SuppressSuperFlag

UB[1]

0 = Create super variable
1 = Don’t create super variable

PreloadSuperFlag

UB[1]

0 = Don’t preload super into
register
1 = Preload super into register

SuppressArgumentsFlag

112

UB[1]

0 = Create arguments variable

1 = Don’t create arguments variable
PreloadArgumentsFlag

UB[1]

0 = Don’t preload arguments into
register
1 = Preload arguments into register

SuppressThisFlag

UB[1]

0 = Create this variable
1 = Don’t create this variable

PreloadThisFlag

UB[1]

0 = Don’t preload this into register
1 = Preload this into register

Reserved

UB[7]

Always 0

PreloadGlobalFlag

UB[1]

0 = Don’t preload _global into
register
1 = Preload _global into register

Parameters

REGISTERPARAM[NumParams]

See REGISTERPARAM, following

codeSize

UI16

# of bytes of code that follow

REGISTERPARAM is defined as follows:
Field

Type

Comment

Register

UI8

For each parameter to the function, a register can be specified. If the
register specified is zero, the parameter is created as a variable named
ParamName in the activation object, which can be referenced with
ActionGetVariable and ActionSetVariable. If the register specified is
nonzero, the parameter is copied into the register, and it can be
referenced with ActionPush and ActionStoreRegister, and no variable is
created in the activation object.

ParamName

STRING

Parameter name

The function body following an ActionDefineFunction2 consists of further action codes, just as for
ActionDefineFunction.
Flash Player selects register numbers by first copying each argument into the register specified in the
113

corresponding REGISTERPARAM record. Next, the preloaded variables are copied into registers starting at 1, and
in the order this, arguments, super, _root, _parent, and _global, skipping any that are not to be preloaded. (The
SWF file must accurately specify which registers will be used by preloaded variables and ensure that no
parameter uses a register number that falls within this range, or else that parameter is overwritten by a
preloaded variable.)
The value of NumParams should equal the number of parameter registers. The value of RegisterCount should
equal NumParams plus the number of preloaded variables and the number of local variable registers desired.
For example, if NumParams is 2, RegisterCount is 6, PreloadThisFlag is 1, and PreloadRootFlag is 1, the
REGISTERPARAM records will probably specify registers 3 and 4. Register 1 will be this, register 2 will be _root,
registers 3 and 4 will be the first and second parameters, and registers 5 and 6 will be for local variables.

ActionExtends
ActionExtends implements the ActionScript extends keyword. ActionExtends creates an inheritance relationship
between two classes, called the subclass and the superclass.
SWF 7 adds ActionExtends to the file format to avoid spurious calls to the superclass constructor function (which
would occur when inheritance was established under ActionScript 1.0). Consider the following code:
Subclass.prototype = new Superclass();

Before the existence of ActionExtends, this code would result in a spurious call to the Superclass
superconstructor function. Now, ActionExtends is generated by the ActionScript compiler when the code class A
extends B is encountered, to set up the inheritance relationship between A and B.
Field

Type

Comment

ActionExtends

ACTIONRECORDHEADER

ActionCode = 0x69

ActionExtends does the following:
1. Pops the ScriptObject superclass constructor off the stack.
2. Pops the ScriptObject subclass constructor off the stack.
3. Creates a new ScriptObject.
4. Sets the new ScriptObject’s proto property to the superclass’ prototype property.
5. Sets the new ScriptObject’s constructor property to the superclass.
6. Sets the subclass’ prototype property to the new ScriptObject. These steps are the equivalent to the
following ActionScript:

114

Subclass.prototype = new Object();
Subclass.prototype. proto = Superclass.prototype;
Subclass.prototype. constructor = Superclass;

ActionCastOp
ActionCastOp implements the ActionScript cast operator, which allows the casting from one data type to
another. ActionCastOp pops an object off the stack and attempts to convert the object to an instance of the
class or to the interface represented by the constructor function.
Field

Type

Comment

ActionCastOp

ACTIONRECORDHEADER

ActionCode = 0x2B

ActionCastOp does the following:
1. Pops the ScriptObject to cast off the stack.
2. Pops the constructor function off the stack.
3. Determines if object is an instance of constructor (doing the same comparison as ActionInstanceOf).
4. If the object is an instance of constructor, the popped ScriptObject is pushed onto the stack.
If the object is not an instance of constructor, a null value is pushed onto the stack.

ActionImplementsOp
ActionImplementsOp implements the ActionScript implements keyword. The ActionImplementsOp action
specifies the interfaces that a class implements, for use by ActionCastOp. ActionImplementsOp can also specify
the interfaces that an interface implements, as interfaces can extend other interfaces.
Field

Type

Comment

ActionImplementsOp

ACTIONRECORDHEADER

ActionCode = 0x2C

ActionImplementsOp does the following:
1. Pops the constructor function off the stack. The constructor function represents the class that will
implement the interfaces. The constructor function must have a prototype property.
2. Pops the count of implemented interfaces off the stack.
3. For each interface count, pops a constructor function off of the stack. The constructor function
represents an interface.
115

4. Sets the constructor function’s list of interfaces to the array collected in the previous step, and sets the
count of interfaces to the count popped in step 2.

ActionTry
ActionTry defines handlers for exceptional conditions, implementing the ActionScript try, catch, and finally
keywords.
Field

Type

Comment

ActionTry

ACTIONRECORDHEADER

ActionCode = 0x8F

Reserved

UB[5]

Always zero

CatchInRegisterFlag

UB[1]

0 - Do not put caught object into register (instead,
store in named variable)
1 - Put caught object into register (do not store in
named variable)

FinallyBlockFlag

UB[1]

0 - No finally block
1 - Has finally block

CatchBlockFlag

UB[1]

0 - No catch block
1 - Has catch block

TrySize

UI16

Length of the try block

CatchSize

UI16

Length of the catch block

FinallySize

UI16

Length of the finally block

CatchName

If CatchInRegisterFlag = 0, STRING

Name of the catch variable

CatchRegister

If CatchInRegisterFlag = 1, UI8

Register to catch into

TryBody

UI8[TrySize]

Body of the try block

CatchBody

UI8[CatchSize]

Body of the catch block, if any

FinallyBody

UI8[FinallySize]

Body of the finally block, if any

The CatchSize and FinallySize fields always exist, whether or not the CatchBlockFlag or FinallyBlockFlag settings
are 1.
The try, catch, and finally blocks do not use end tags to mark the end of their respective blocks. Instead, the
length of a block is set by the TrySize, CatchSize, and FinallySize values.
116

ActionThrow
ActionThrow implements the ActionScript throw keyword. ActionThrow is used to signal, or throw, an
exceptional condition, which is handled by the exception handlers declared with ActionTry.
If any code within the try block throws an object, control passes to the catch block, if one exists, then to the
finally block, if one exists. The finally block always executes, regardless of whether an error was thrown.
If an exceptional condition occurs within a function and the function does not include a catch handler, the
function and any caller functions are exited until a catch block is found (executing all finally handlers at all
levels).
Any ActionScript data type can be thrown, though typically usage is to throw objects.
Field

Type

Comment

ActionThrow

ACTIONRECORDHEADER

ActionCode = 0x2A

ActionThrow pops the value to be thrown off the stack.

SWF 9 action model
SWF 9 added the DoABC action-definition tag. This tag contains an .abc bytecode block that is parsed by the
ActionScript 3.0 virtual machine. The DoABC tag is the main vehicle for delivering ActionScript 3.0 bytecode.

DoABC
The DoABC tag is similar to the DoAction tag: it defines a series of bytecodes to be executed. However, the
bytecodes contained within the DoABC tag run in the ActionScript 3.0 virtual machine.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 82

Flags

UI32

A 32-bit flags value, which may contain the following bits set:
kDoAbcLazyInitializeFlag = 1: Indicates that the ABC block should not
be executed immediately, but only parsed. A later finddef may cause
its scripts to execute.

Name

STRING

The name assigned to the bytecode

ABCData

BYTE[]

A block of .abc bytecode to be parsed by the ActionScript 3.0 virtual
machine, up to the end of the tag.

117

For details on the contents and format of the ABCData field, see the Adobe ActionScript Virtual Machine 2
(AVM2) Overview at www.adobe.com/go/avm2overview/.

SWF 10 action model
There are no changes to the action model in SWF 10.

118

Chapter 6: Shapes
The SWF shape architecture is designed to be compact, flexible and rendered very quickly to the screen. It is
similar to most vector formats in that shapes are defined by a list of edges called a path. A path may be closed,
where the start and end of the path meet to close the figure, or open, where the path forms an open-ended
stroke. A path may contain a mixture of straight edges, curved edges, and ‘pen up and move’ commands. The
latter allows multiple disconnected figures to be described by a single shape structure.
A fill style defines the appearance of an area enclosed by a path. Fill styles supported by the SWF file format
include a color, a gradient, or a bitmap image.
A line style defines the appearance of the outline of a path. The line style may be a stroke of any thickness and
color.
Most vector formats allow only one fill and line style per path. The SWF file format extends this concept by
allowing each edge to have its own line and fill style. This can have unpredictable results when fill styles change
in the middle of a path.
The Adobe Flash authoring tool also supports two fill styles per edge, one for each side of the edge: FillStyle0
and FillStyle1. FillStyle0 should always be used first and then FillStyle1 if the shape is filled on both sides of the
edge.

Shape overview
A shape is composed of the following elements:
•

CharacterId—A 16-bit value that uniquely identifies this shape as a ‘character’ in the dictionary. The
CharacterId can be referred to in control tags such as PlaceObject. Characters can be reused and
combined with other characters to make more complex shapes.

•

Bounding box—The rectangle that completely encloses the shape.

•

Fill style array—A list of all the fill styles used in a shape.

•

Line style array—A list of all the line styles used in a shape.

•

Shape record array—A list of shape records. Shape records can define straight or curved edges, style
changes, or move the drawing position.

Note: Line and fill styles are defined only once and may be used (and reused) by any of the edges in the shape.

119

Shape example
The following example appears to be a collection of shapes, but it can be described with a single DefineShape
tag.

The red circle, red square and green rounded-rectangle are closed paths. The curved line is an open path. The
red square consists of all straight edges, the red circle consists of all curved edges, while the rounded rectangle
has curved edges interspersed with straight edges.
There are two fill styles, solid green and solid red, and two line styles, 1-pixel black, and 2- pixel black. The red
circle and red square share the same fill and line styles. The rounded rectangle and curved line share the same
line style.
Here’s how to describe this example with the SWF file format.
Define the fill styles:
1. First, the fill styles are defined with a FILLSTYLEARRAY. The two unique fill styles are solid red and solid
green.
2. This is followed by a LINESTYLEARRAY that includes the two unique line styles: 1-pixel black, and 2-pixel
black.
3. This is followed by an array of shape records (see Shape records).
All shape records share a similar structure but can have varied meaning. A shape record can define straight or
curved edge, a style change, or it can move the current drawing position.
Define the curved line:
1. The first shape record selects the 2-pixel-wide line style, and moves the drawing position to the top of
the curved line by setting the StateMoveTo flag.
2. The next shape record is a curved edge, which ends to the bottom of the line. The path is not closed.
Define the red square:
120

1. The next shape record selects the 1-pixel line style and the red fill style. It also moves the drawing position to
the upper-left corner of the red rectangle.
2. The following four shape records are straight edges. The last edge must end at the upper- left corner. Flash
Player requires that closed figures be joined explicitly. That is, the first and last points must be coincident.
Define the red circle:
1. The next shape record does not change any style settings, but moves the drawing position to the top of
the red circle.
2. The following eight shape records are curved edges that define the circle. Again, the path must finish
where it started.
Define the green rounded-rectangle:
1. The next shape record selects the 2-pixel-wide line style, and the green fill. It also moves the drawing
position to the upper left of the rounded-rectangle.
2. The following twelve shape records are a mixture of straight shape records (the sides) interspersed with
curved shape records (the rounded corners). The path finishes where it began.

Shape structures
Fill styles
The SWF file format supports three basic types of fills for a shape.

121

•

Solid fill A simple RGB or RGBA color that fills a portion of a shape. An alpha value of 255 means a
completely opaque fill. An alpha value of zero means a completely transparent fill. Any alpha between 0
and 255 will be partially transparent.

•

Gradient Fill
A gradient fill can be either a linear or a radial gradient. For an in-depth description of
how gradients are defined, see Gradients.

•

Bitmap fill Bitmap fills refer to a bitmap characterId. There are two styles: clipped and tiled. A clipped
bitmap fill repeats the color on the edge of a bitmap if the fill extends beyond the edge of the bitmap. A
tiled fill repeats the bitmap if the fill extends beyond the edge of the bitmap.

FILLSTYLEARRAY
A fill style array enumerates a number of fill styles. The format of a fill style array is described in the following
table:
Field

Type

Comment

FillStyleCount

UI8

Count of fill styles.

FillStyleCountExtended

If FillStyleCount = 0xFF, UI16

Extended count of fill styles.
Supported only for Shape2 and
Shape3.

FillStyles

FILLSTYLE[FillStyleCount]

Array of fill styles.

FILLSTYLE
The format of a fill style value within the file is described in the following table:
Field

Type

Comment

FillStyleType

UI8

Type of fill style:
0x00 = solid fill
0x10 = linear gradient fill
0x12 = radial gradient fill
0x13 = focal radial gradient fill (SWF
8 file format and later only)
0x40 = repeating bitmap fill
0x41 = clipped bitmap fill
0x42 = non-smoothed repeating
bitmap
0x43 = non-smoothed clipped
bitmap

Color

If type = 0x00, RGBA (if Shape3);
RGB (if Shape1 or Shape2)

Solid fill color with opacity
information.

GradientMatrix

If type = 0x10, 0x12, or 0x13,
MATRIX

Matrix for gradient fill.

122

Gradient

If type = 0x10 or 0x12, GRADIENT If
type = 0x13, FOCALGRADIENT (SWF
8 and later only)

Gradient fill.

BitmapId

If type = 0x40, 0x41, 0x42 or 0x43,
UI16

ID of bitmap character for fill.

BitmapMatrix

If type = 0x40, 0x41, 0x42 or 0x43,
MATRIX

Matrix for bitmap fill.

Line styles
A line style array enumerates a number of line styles.

LINESTYLEARRAY
The format of a line style array is described in the following table:
Field

Type

Comment

LineStyleCount

UI8

Count of line styles.

LineStyleCountExtended

If LineStyleCount = 0xFF, UI16

Extended count of line styles.

LineStyles

If Shape1, Shape2, or Shape3,
LINESTYLE[count]. If Shape4,
LINESTYLE2[count]

Array of line styles.

LINESTYLE
A line style represents a width and color of a line. The format of a line style value within the file is described in
the following table:
Field

Type

Comment

Width

UI16

Width of line in twips.

Color

RGB (Shape1 or Shape2) RGBA
(Shape3)

Color value including alpha channel
information for Shape3.

Note 1: Before the introduction of LINESTYLE2 in SWF 8, all lines in the SWF file format have rounded joins and
round caps. Different join styles and end styles can be simulated with a very narrow shape that looks identical to
the desired stroke.

123

Note 2: The SWF file format has no native support for dashed or dotted line styles. A dashed line can be
simulated by breaking up the path into a series of short lines.

LINESTYLE2
LINESTYLE2 builds upon the capabilities of the LINESTYLE record by allowing the use of new types of joins and
caps as well as scaling options and the ability to fill a stroke. In order to use LINESTYLE2, the shape must be
defined with DefineShape4—not DefineShape, DefineShape2, or DefineShape3.
While the LINESTYLE record permits only rounded joins and round caps, LINESTYLE2 also supports miter and
bevel joins, and square caps and no caps. The following diagram illustrates the complete array of joins and caps:

Miter Join

Round Join

Bevel Join

None Cap

Round Cap

Square Cap

When using LINESTYLE2 for a miter join, a MiterLimitFactor must be specified and is used to calculate the
maximum miter length:
Maximum miter length = LINESTYLE2 MiterLimitFactor * LINESTYLE2 Width

If the miter join exceeds the maximum miter length, Flash Player will cut off the miter. Note that
MiterLimitFactor is an 8.8 fixed-point value.
LINESTYLE2 also includes the option for pixel hinting to correct blurry vertical or horizontal lines.
Field

Type

Comment

Width

UI16

Width of line in twips.

StartCapStyle

UB[2]

Start cap style:; 0 = Round cap; 1 =
No cap; 2 = Square cap

JoinStyle

UB[2]

Join style: 0 = Round join; 1 = Bevel
join; 2 = Miter join

HasFillFlag

UB[1]

If 1, fill is defined in FillType. If 0,
uses Color field.

NoHScaleFlag

UB[1]

If 1, stroke thickness will not scale if

124

the object is scaled horizontally.
NoVScaleFlag

UB[1]

If 1, stroke thickness will not scale if
the object is scaled vertically.

PixelHintingFlag

UB[1]

If 1, all anchors will be aligned to
full pixels.

Reserved

UB[5]

Must be 0.

NoClose

UB[1]

If 1, stroke will not be closed if the
stroke’s last point matches its first
point. Flash Player will apply caps
instead of a join.

EndCapStyle

UB[2]

End cap style: 0 = Round cap; 1 =
No cap; 2 = Square cap

MiterLimitFactor

If JoinStyle = 2, UI16

Miter limit factor is an 8.8 fixedpoint value.

Color

If HasFillFlag = 0, RGBA

Color value including alpha
channel.

FillType

If HasFillFlag = 1, FILLSTYLE

Fill style for this stroke.

Shape structures
The SHAPE structure defines a shape without fill style or line style information.

SHAPE
SHAPE is used by the DefineFont tag, to define character glyphs.
Field

Type

Comment

NumFillBits

UB[4]

Number of fill index bits.

NumLineBits

UB[4]

Number of line index bits.

ShapeRecords

SHAPERECORD[one or more]

Shape records (see following).

SHAPEWITHSTYLE
The SHAPEWITHSTYLE structure extends the SHAPE structure by including fill style and line style information.
125

SHAPEWITHSTYLE is used by the DefineShape tag.
Field

Type

Comment

FillStyles

FILLSTYLEARRAY

Array of fill styles.

LineStyles

LINESTYLEARRAY

Array of line styles.

NumFillBits

UB[4]

Number of fill index bits.

NumLineBits

UB[4]

Number of line index bits.

ShapeRecords

SHAPERECORD[one or more]

Shape records (see following).

Note: The LINESTYLELARRAY and FILLSTYLEARRAY begin at index 1, not index 0.
The following diagram illustrates the SHAPEWITHSTYLE structure.
Shape Tag
Fill Styles
Line Styles

Change Fills
Edges

Change Fills
Edges

First, the Fill styles and Line styles are defined. These are defined only once and are referred to by array index.
The blue area represents the array of Shape records. The first shape record selects a fill from the fill style array,
and moves the drawing position to the start of the shape. This is followed by a series of edge records that define
the shape. The next record changes the fill style, and the subsequent edge records are filled using this new style.
This tag is a completely autonomous object. The style change records only refer to fill and line styles that have
been defined in this tag.

Shape records
There are four types of shape records:
•
126

End shape record

•

Style change record

•

Straight edge record

•

Curved edge record

Each shape record begins with a TypeFlag. If the TypeFlag is zero, the shape record is a non- edge record, and a
further five bits of flag information follow.

EndShapeRecord
The end shape record simply indicates the end of the shape record array. It is a non-edge record with all five
flags equal to zero.
Field

Type

Comment

TypeFlag

UB[1]

Non-edge record flag. Always 0.

EndOfShape

UB[5]

End of shape flag. Always 0.

StyleChangeRecord
The style change record is also a non-edge record. It can be used to do the following:
1. Select a fill or line style for drawing.
2. Move the current drawing position (without drawing).
3. Replace the current fill and line style arrays with a new set of styles.
Because fill and line styles often change at the start of a new path, it is useful to perform more than one action
in a single record. For example, say a DefineShape tag defines a red circle and a blue square. After the circle is
closed, it is necessary to move the drawing position, and replace the red fill with the blue fill. The style change
record can achieve this with a single shape record.
Field

Type

Comment

TypeFlag

UB[1]

Non-edge record flag. Always 0.

StateNewStyles

UB[1]

New styles flag. Used by
DefineShape2 and DefineShape3
only.

StateLineStyle

UB[1]

Line style change flag.

StateFillStyle1

UB[1]

Fill style 1 change flag.

127

StateFillStyle0

UB[1]

Fill style 0 change flag.

StateMoveTo

UB[1]

Move to flag.

MoveBits

If StateMoveTo, UB[5]

Move bit count.

MoveDeltaX

If StateMoveTo, SB[MoveBits]

Delta X value.

MoveDeltaY

If StateMoveTo, SB[MoveBits]

Delta Y value.

FillStyle0

If StateFillStyle0, UB[FillBits]

Fill 0 Style.

FillStyle1

If StateFillStyle1, UB[FillBits]

Fill 1 Style.

LineStyle

If StateLineStyle, UB[LineBits]

Line Style.

FillStyles

If StateNewStyles, FILLSTYLEARRAY

Array of new fill styles.

LineStyles

If StateNewStyles, LINESTYLEARRAY

Array of new line styles.

NumFillBits

If StateNewStyles, UB[4]

Number of fill index bits for new
styles.

NumLineBits

If StateNewStyles, UB[4]

Number of line index bits for new
styles.

MoveDeltaX and MoveDeltaY are relative to the shape origin.
The style arrays begin at index 1, not index 0. FillStyle = 1 refers to the first style in the fill style array, FillStyle = 2
refers to the second style in the fill style array, and so on. A fill style index of zero means the path is not filled,
and a line style index of zero means the path has no stroke. Initially the fill and line style indices are set to zero—
no fill or stroke.

FillStyle0 and FillStyle1
The Adobe Flash authoring tool supports two fill styles per edge, one for each side of the edge: FillStyle0 and
FillStyle1. For shapes that don’t self-intersect or overlap, FillStyle0 should be used. For overlapping shapes the
situation is more complex.
For example, if a shape consists of two overlapping squares, and only FillStyle0 is defined, Flash Player renders a
‘hole’ where the paths overlap. This area can be filled using FillStyle1. In this situation, the rule is that for any
directed vector, FillStyle0 is the color to the left of the vector, and FillStyle1 is the color to the right of the vector
(as shown in the following diagram).

128

Note: FillStyle0 and FillStyle1 should not be confused with FILLSTYLEARRAY indices. FillStyle0 and FillStyle1 are
variables that contain indices into the FILLSTYLEARRAY.

Edge records
Edge records have a TypeFlag of 1. There are two types of edge records: straight and curved. The StraightFlag
determines the type.

StraightEdgeRecord
The StraightEdgeRecord stores the edge as an X-Y delta. The delta is added to the current drawing position, and
this becomes the new drawing position. The edge is rendered between the old and new drawing positions.
Straight edge records support three types of lines:
1. General lines.
2. Horizontal lines.
3. Vertical lines.
General lines store both X and Y deltas, the horizontal and vertical lines store only the X delta and Y delta
respectively.

129

Field

Type

Comment

TypeFlag

UB[1]

This is an edge record. Always 1.

StraightFlag

UB[1]

Straight edge. Always 1.

NumBits

UB[4]

Number of bits per value (2 less
than the actual number).

GeneralLineFlag

UB[1]

General Line equals 1. Vert/Horz
Line equals 0.

VertLineFlag

If GeneralLineFlag = 0, SB[1]

Vertical Line equals 1. Horizontal
Line equals 0.

DeltaX

If GeneralLineFlag = 1 or if
VertLineFlag = 0, SB[NumBits+2]

X delta

DeltaY

If GeneralLineFlag = 1 or if
VertLineFlag = 1, SB[NumBits+2]

Y delta.

CurvedEdgeRecord
The SWF file format differs from most vector file formats by using Quadratic Bezier curves rather than Cubic
Bezier curves. PostScript™ uses Cubic Bezier curves, as do most drawing applications.The SWF file format uses
Quadratic Bezier curves because they can be stored more compactly, and can be rendered more efficiently.
The following diagram shows a Quadratic Bezier curve and a Cubic Bezier curve.

A Quadratic Bezier curve has 3 points: 2 on-curve anchor points, and 1 off-curve control point. A Cubic Bezier
curve has 4 points: 2 on-curve anchor points, and 2 off-curve control points.
The curved-edge record stores the edge as two X-Y deltas. The three points that define the
Quadratic Bezier are calculated like this:
1. The first anchor point is the current drawing position.
2. The control point is the current drawing position + ControlDelta.
3. The last anchor point is the current drawing position + ControlDelta + AnchorDelta. The last anchor point
becomes the current drawing position.
130

Field

Type

Comment

TypeFlag

UB[1]

This is an edge record. Always 1.

StraightFlag

UB[1]

Curved edge. Always 0.

NumBits

UB[4]

Number of bits per value (2 less
than the actual number).

ControlDeltaX

SB[NumBits+2]

X control point change.

ControlDeltaY

SB[NumBits+2]

Y control point change.

AnchorDeltaX

SB[NumBits+2]

X anchor point change.

AnchorDeltaY

SB[NumBits+2]

Y anchor point change.

Converting between quadratic and cubic Bezier curves
Replace the single off-curve control point of the quadratic Bezier curve with two new off-curve control points for
the cubic Bezier curve. Place each new off-curve control point along the line between one of the on-curve
anchor points and the original off-curve control point. The new off-curve control points should be 2/3 of the way
from the on-curve anchor point to the original off-curve control point. The diagram of quadratic and cubic Bezier
curves above illustrates this substitution.
A cubic Bezier curve can be approximated only with a quadratic Bezier curve, because you are going from a
third-order curve to a second-order curve. This involves recursive subdivision of the curve, until the cubic curve
and the quadratic equivalent are matched within some arbitrary tolerance.
For a discussion of how to approximate cubic Bezier curves with quadratic Bezier curves see the following:
•

Converting Bezier Curves to Quadratic Splines at stevehollasch.com/cgindex/curves/cbezquadspline.html

•

TrueType Reference Manual, Converting Outlines to the TrueType Format at
developer.apple.com/fonts/TTRefMan/RM08/appendixE.html

Shape tags
DefineShape
The DefineShape tag defines a shape for later use by control tags such as PlaceObject. The ShapeId uniquely
identifies this shape as ‘character’ in the Dictionary. The ShapeBounds field is the rectangle that completely
encloses the shape. The SHAPEWITHSTYLE structure includes all the paths, fill styles and line styles that make up
the shape.

131

The minimum file format version is SWF 1.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 2.

ShapeId

UI16

ID for this character

ShapeBounds

RECT

Bounds of the shape.

Shapes

SHAPEWITHSTYLE

Shape information.

DefineShape2
DefineShape2 extends the capabilities of DefineShape with the ability to support more than 255 styles in the
style list and multiple style lists in a single shape. The minimum file format version is SWF 2.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 22.

ShapeId

UI16

ID for this character.

ShapeBounds

RECT

Bounds of the shape.

Shapes

SHAPEWITHSTYLE

Shape information.

DefineShape3
DefineShape3 extends the capabilities of DefineShape2 by extending all of the RGB color fields to support RGBA
with opacity information.
The minimum file format version is SWF 3.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 32.

ShapeId

UI16

ID for this character.

ShapeBounds

RECT

Bounds of the shape.

Shapes

SHAPEWITHSTYLE

Shape information.

DefineShape4
DefineShape4 extends the capabilities of DefineShape3 by using a new line style record in the shape. LINESTYLE2
132

allows new types of joins and caps as well as scaling options and the ability to fill a stroke.
DefineShape4 specifies not only the shape bounds but also the edge bounds of the shape. While the shape
bounds are calculated along the outside of the strokes, the edge bounds are taken from the outside of the
edges, as shown in the following diagram. The EdgeBounds field assists Flash Player in accurately determining
certain layouts.

Strokes
Edges
Shape bounds
Edge bounds

In addition, DefineShape4 includes new hinting flags UsesNonScalingStrokes and UsesScalingStrokes. These flags
assist Flash Player in creating the best possible area for invalidation.
The minimum file format version is SWF 8.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 83.

ShapeId

UI16

ID for this character.

ShapeBounds

RECT

Bounds of the shape.

EdgeBounds

RECT

Bounds of the shape, excluding
strokes.

Reserved

UB[5]

Must be 0.

UsesFillWindingRule

UB[1]

If 1, use fill winding rule. Minimum
file format version is SWF 10

UsesNonScalingStrokes

UB[1]

If 1, the shape contains at least one
non-scaling stroke.

UsesScalingStrokes

UB[1]

If 1, the shape contains at least one
scaling stroke.

Shapes

SHAPEWITHSTYLE

Shape information.

133

Chapter 7: Gradients
Gradients are a special type of shape fill for SWF shapes. They create ramps of colors that interpolate between
two or more fixed colors.
Here is an overview of the SWF gradient model:
•

There are two styles of gradient: Linear and Radial. In addition, with the SWF 8 file format, a new radial
gradient type is added to allow the focal point to be set.

•

Each gradient has its own transformation matrix, and can be transformed independently of its parent
shape.

•

A gradient can have up to eight control points in SWF 7 file format and previous versions, or up to
fifteen control points in SWF 8 and later. Colors are interpolated between the control points to create
the color ramp.

•

Each control point is defined by a ratio and an RGBA color. The ratio determines the position of the
control point in the gradient; the RGBA value determines its color.

Following are some examples of SWF gradients (from left to right):
•

A simple white-to-black linear gradient.

•

A simple white-to-black radial gradient.

•

A “rainbow” gradient consisting of seven control points; red, yellow, green, cyan, blue, purple, and red.

•

A three-point gradient, where the end points are opaque and the center point is transparent. The result
is a gradient in the alpha-channel that allows the diamond shape in the background to show through.

Gradient transformations
All gradients are defined in a standard space called the gradient square. The gradient square is centered at (0,0),
and extends from (-16384,-16384) to (16384,16384).
Each gradient is mapped from the gradient square to the display surface using a standard transformation matrix.
This matrix is stored in the FILLSTYLE structure.

134

Example: In the following diagram a linear gradient is mapped onto a circle 4096 units in diameter, and centered
at (2048,2048).

The 2x3 MATRIX required for this mapping is:
|
0.125
0.000
|
0.000
0.125
| 2048.000 2048.000

|
|
|

The gradient is scaled to one-eighth of its original size (32768 / 4096 = 8), and translated to (2048, 2048).

Gradient control points
The position of a control point in the gradient is determined by a ratio value between 0 and 255. For a linear
gradient, a ratio of zero maps to the left side of the gradient square, and a ratio of 255 maps to the right side.
For a radial gradient, a ratio of zero maps to the center point of the gradient square, and a ratio of 255 maps to
the largest circle that fits inside the gradient square.
The color of a control point is determined by an RGBA value. An alpha value of zero means the gradient is
completely transparent at this point. An alpha value of 255 means the gradient is completely opaque at this
point.
Control points are sorted by ratio, with the smallest ratio first.

Gradient structures
The gradient structures are part of the FILLSTYLE structure.

GRADIENT
SWF 8 and later supports up to 15 gradient control points, spread modes and a new interpolation type.
Note that for the DefineShape, DefineShape2 or DefineShape3 tags, the SpreadMode and InterpolationMode
fields must be 0, and the NumGradients field cannot exceed 8.
Field

135

Type

Comment

SpreadMode

UB[2]

0 = Pad mode; 1 = Reflect mode; 2
= Repeat mode; 3 = Reserved

InterpolationMode

UB[2]

0 = Normal RGB mode
interpolation; 1 = Linear RGB mode
interpolation; 2 and 3 = Reserved

NumGradients

UB[4]

1 to 15

GradientRecords

GRADRECORD[nGrads]

Gradient records (see following)

FOCALGRADIENT
A FOCALGRADIENT must be declared in DefineShape4—not DefineShape, DefineShape2 or DefineShape3.
The value range is from -1.0 to 1.0, where -1.0 means the focal point is close to the left border of the radial
gradient circle, 0.0 means that the focal point is in the center of the radial gradient circle, and 1.0 means that
the focal point is close to the right border of the radial gradient circle.
Field

Type

Comment

SpreadMode

UB[2]

0 = Pad mode; 1 = Reflect mode; 2 = Repeat mode; 3 =
Reserved

InterpolationMode

UB[2]

0 = Normal RGB mode interpolation; 1 = Linear RGB mode
interpolation; 2 and 3 = Reserved

NumGradients

UB[4]

1 to 15

GradientRecords

GRADRECORD[nGrads]

Gradient records (see following)

FocalPoint

FIXED8

Focal point location

GRADRECORD
The GRADRECORD defines a gradient control point:
Field

Type

Comment

Ratio

UI8

Ratio value

Color

RGB (Shape1 or Shape2) RGBA
(Shape3)

Color of gradient

136

Chapter 8: Bitmaps
The SWF file format specification supports a variety of bitmap formats. All bitmaps are compressed to reduce
file size. Lossy compression, best for imprecise images such as photographs, is provided by JPEG bitmaps;
lossless compression, best for precise images such as diagrams, icons, or screen captures, is provided by ZLIB
bitmaps. Both types of bitmaps can optionally contain alpha channel (opacity) information.
The JPEG format, officially defined as ITU T.81 or ISO/IEC 10918-1, is an open standard developed by the
Independent Joint Photographic Experts Group. The JPEG format is not described in this document. For general
information on the JPEG format, see JPEG at www.jpeg.org/. For a specification of the JPEG format, see the
International Telecommunication Union at www.itu.int/ and search for recommendation T.81. The JPEG data in
SWF files is encoded using the JPEG Interchange Format specified in Annex B. Flash Player also understands the
popular JFIF format, an extension of the JPEG Interchange Format.
In all cases where arrays of non-JPEG pixel data are stored in bitmap tags, the pixels appear in row-major order,
reading like English text, proceeding left to right within rows and top to bottom overall.

DefineBits
This tag defines a bitmap character with JPEG compression. It contains only the JPEG compressed image data
(from the Frame Header onward). A separate JPEGTables tag contains the JPEG encoding data used to encode
this image (the Tables/Misc segment).
Note: Only one JPEGTables tag is allowed in a SWF file, and thus all bitmaps defined with DefineBits must share
common encoding tables.
The data in this tag begins with the JPEG SOI marker 0xFF, 0xD8 and ends with the EOI marker 0xFF, 0xD9.
Before version 8 of the SWF file format, SWF files could contain an erroneous header of 0xFF, 0xD9, 0xFF, 0xD8
before the JPEG SOI marker.
The minimum file format version for this tag is SWF 1.
Field

Type

Comment

Header

RECORDHEADER (long)

Tag type = 6

CharacterID

UI16

ID for this character

JPEGData

UI8[image data size]

JPEG compressed image

JPEGTables
This tag defines the JPEG encoding table (the Tables/Misc segment) for all JPEG images defined using the
137

DefineBits tag. There may only be one JPEGTables tag in a SWF file.
The data in this tag begins with the JPEG SOI marker 0xFF, 0xD8 and ends with the EOI marker 0xFF, 0xD9.
Before version 8 of the SWF file format, SWF files could contain an erroneous header of 0xFF, 0xD9, 0xFF, 0xD8
before the JPEG SOI marker.
The minimum file format version for this tag is SWF 1.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 8

JPEGData

UI8[encoding data size]

JPEG encoding table

DefineBitsJPEG2
This tag defines a bitmap character with JPEG compression. It differs from DefineBits in that it contains both the
JPEG encoding table and the JPEG image data. This tag allows multiple JPEG images with differing encoding
tables to be defined within a single SWF file.
The data in this tag begins with the JPEG SOI marker 0xFF, 0xD8 and ends with the EOI marker 0xFF, 0xD9.
Before version 8 of the SWF file format, SWF files could contain an erroneous header of 0xFF, 0xD9, 0xFF, 0xD8
before the JPEG SOI marker.
In addition to specifying JPEG data, DefineBitsJPEG2 can also contain PNG image data and non-animated GIF89a
image data.
•

If ImageData begins with the eight bytes 0x89 0x50 0x4E 0x47 0x0D 0x0A 0x1A 0x0A, the ImageData
contains PNG data.

•

If ImageData begins with the six bytes 0x47 0x49 0x46 0x38 0x39 0x61, the ImageData contains GIF89a
data.

The minimum file format version for this tag is SWF 2. The minimum file format version for embedding PNG of
GIF89a data is SWF 8.
Field

Type

Comment

Header

RECORDHEADER (long)

Tag type = 21

CharacterID

UI16

ID for this character

ImageData

UI8[data size]

Compressed image data in either
JPEG, PNG, or GIF89a format

138

DefineBitsJPEG3
This tag defines a bitmap character with JPEG compression. This tag extends DefineBitsJPEG2, adding alpha
channel (opacity) data. Opacity/transparency information is not a standard feature in JPEG images, so the alpha
channel information is encoded separately from the JPEG data, and compressed using the ZLIB standard for
compression. The data format used by the ZLIB library is described by Request for Comments (RFCs) documents
1950 to 1952.
The data in this tag begins with the JPEG SOI marker 0xFF, 0xD8 and ends with the EOI marker 0xFF, 0xD9.
Before version 8 of the SWF file format, SWF files could contain an erroneous header of 0xFF, 0xD9, 0xFF, 0xD8
before the JPEG SOI marker.
In addition to specifying JPEG data, DefineBitsJPEG2 can also contain PNG image data and non-animated GIF89a
image data.
•

If ImageData begins with the eight bytes 0x89 0x50 0x4E 0x47 0x0D 0x0A 0x1A 0x0A, the ImageData
contains PNG data.

•

If ImageData begins with the six bytes 0x47 0x49 0x46 0x38 0x39 0x61, the ImageData contains GIF89a
data.

If ImageData contains PNG or GIF89a data, the optional BitmapAlphaData is not supported.
The minimum file format version for this tag is SWF 3. The minimum file format version for embedding PNG of
GIF89a data is SWF 8.
Field

Type

Comment

Header

RECORDHEADER (long)

Tag type = 35.

CharacterID

UI16

ID for this character.

AlphaDataOffset

UI32

Count of bytes in ImageData.

ImageData

UI8[data size]

Compressed image data in either JPEG, PNG, or GIF89a
format

BitmapAlphaData

UI8[alpha data size]

ZLIB compressed array of alpha data. Only supported when
tag contains JPEG data. One byte per pixel. Total size after
decompression must equal (width * height) of JPEG image.

DefineBitsLossless
Defines a lossless bitmap character that contains RGB bitmap data compressed with ZLIB. The data format used
by the ZLIB library is described by Request for Comments (RFCs) documents 1950 to 1952.
139

Two kinds of bitmaps are supported. Colormapped images define a colormap of up to 256 colors, each
represented by a 24-bit RGB value, and then use 8-bit pixel values to index into the colormap. Direct images
store actual pixel color values using 15 bits (32,768 colors) or 24 bits (about 17 million colors).
The minimum file format version for this tag is SWF 2.
Field

Type

Comment

Header

RECORDHEADER (long)

Tag type = 20

CharacterID

UI16

ID for this character

BitmapFormat

UI8

Format of compressed data:
3 = 8-bit colormapped image
4 = 15-bit RGB image
5 = 24-bit RGB image

BitmapWidth

UI16

Width of bitmap image

BitmapHeight

UI16

Height of bitmap image

BitmapColorTableSize

If BitmapFormat = 3
UI8; Otherwise absent

This value is one less than the actual
number of colors in the color table,
allowing for up to 256 colors.

If BitmapFormat = 3, COLORMAPDATA

ZLIB compressed bitmap data

ZlibBitmapData

If BitmapFormat = 4 or 5, BITMAPDATA
The COLORMAPDATA and BITMAPDATA structures contain image data. These structures are each compressed as
a single block of data. Their layouts before compression follow.
Note: Row widths in the pixel data fields of these structures must be rounded up to the next 32-bit word
boundary. For example, an 8-bit image that is 253 pixels wide must be padded out to 256 bytes per line. To
determine the appropriate padding, make sure to take into account the actual size of the individual pixel
structures; 15-bit pixels occupy 2 bytes and 24-bit pixels occupy 4 bytes (see PIX15 and PIX24).
COLORMAPDATA
Field

Type

Comment

ColorTableRGB

RGB[color table size]

Defines the mapping from color
indices to RGB values. Number of
RGB values is BitmapColorTableSize

140

+ 1.
UI8[image data size]

Array of color indices. Number of
entries is BitmapWidth *
BitmapHeight, subject to padding
(see note preceding this table).

Field

Type

Comment

BitmapPixelData

If BitmapFormat = 4, PIX15[image
data size] If BitmapFormat = 5,
PIX24[image data size]

Array of pixel colors. Number of
entries is BitmapWidth *
BitmapHeight, subject to padding
(see note above).

Field

Type

Comment

Pix15Reserved

UB[1]

Always 0

Pix15Red

UB[5]

Red value

Pix15Green

UB[5]

Green value

Pix15Blue

UB[5]

Blue value

Field

Type

Comment

Pix24Reserved

UI8

Always 0

Pix24Red

UI8

Red value

Pix24Green

UI8

Green value

Pix24Blue

UI8

Blue value

ColormapPixelData

BITMAPDATA

PIX15

PIX24

141

DefineBitsLossless2
DefineBitsLossless2 extends DefineBitsLossless with support for opacity (alpha values). The colormap colors in
colormapped images are defined using RGBA values, and direct images store 32-bit ARGB colors for each pixel.
The intermediate 15-bit color depth is not available in DefineBitsLossless2.
The minimum file format version for this tag is SWF 3.
Field

Type

Comment

Header

RECORDHEADER (long)

Tag type = 36

CharacterID

UI16

ID for this character

BitmapFormat

UI8

Format of compressed data
3 = 8-bit colormapped image
5 = 32-bit ARGB image

BitmapWidth

UI16

Width of bitmap image

BitmapHeight

UI16

Height of bitmap image

BitmapColorTableSize

If BitmapFormat = 3, UI8; Otherwise absent

This value is one less than the actual
number of colors in the color table,
allowing for up to 256 colors.

ZlibBitmapData

If BitmapFormat = 3, ALPHACOLORMAPDATA

ZLIB compressed bitmap data

If BitmapFormat = 4 or 5, ALPHABITMAPDATA

The COLORMAPDATA and BITMAPDATA structures contain image data. These structures are each compressed as
a single block of data. Their layouts before compression follow.
Note: Row widths in the pixel data field of ALPHACOLORMAPDATA must be rounded up to the next 32-bit word
boundary. For example, an 8-bit image that is 253 pixels wide must be padded out to 256 bytes per line. Row
widths in ALPHABITMAPDATA are always 32-bit aligned because the ARGB structure is 4 bytes.
ALPHACOLORMAPDATA
Field

Type

Comment

ColorTableRGB

RGBA[color table size]

Defines the mapping from color indices to RGBA values.
Number of RGBA values is BitmapColorTableSize + 1.

142

UI8[image data size]

Array of color indices. Number of entries is BitmapWidth
* BitmapHeight, subject to padding (see note preceding
this table).

Field

hType

Comment

BitmapPixelData

ARGB[image data size]

Array of pixel colors. Number of entries is BitmapWidth *
BitmapHeight. The RGB data must already be multiplied
bythe alpha channel value.

ColormapPixelData

ALPHABITMAPDATA

DefineBitsJPEG4
This tag defines a bitmap character with JPEG compression. This tag extends DefineBitsJPEG3, adding a
deblocking parameter. While this tag also supports PNG and GIF89a data, the deblocking filter is not applied to
such data.
The minimum file format version for this tag is SWF 10.
Field

Type

Comment

Header

RECORDHEADER (long)

Tag type = 90.

CharacterID

UI16

ID for this character.

AlphaDataOffset

UI32

Count of bytes in ImageData.

DeblockParam

UI16

Parameter to be fed into the deblocking filter. The
parameter describes a relative strength of the
deblocking filter from 0-100% expressed in a normalized
8.8 fixed point format.

ImageData

UI8[data size]

Compressed image data in either JPEG, PNG, or GIF89a
format.

BitmapAlphaData

UI8[alpha data size]

ZLIB compressed array of alpha data. Only supported
when tag contains JPEG data. One byte per pixel. Total
size after decompression must equal (width * height) of
JPEG image.

143

Chapter 9: Shape Morphing
Shape morphing is the metamorphosis of one shape into another over time. The SWF file format specification
supports a flexible morphing model, which allows a number of shape attributes to vary during the morph. The
SWF file format defines only the start and end states of the morph. Adobe Flash Player is responsible for
interpolating between the endpoints and generating the ‘in-between’ states.
The following shape attributes can be varied during the morph:
•

The position of each edge in the shape.

•

The color and thickness of the outline.

•

The fill color of the shape (if filled with a color).

•

The bitmap transform (if filled with a bitmap).

•

The gradient transform (if filled with a gradient).

•

The color and position of each point in the gradient (if filled with a gradient).

The following restrictions apply to morphing:
•

The start and end shapes must have the same number of edges.

•

The start and end shapes must have the same type of fill (that is, solid, gradient or bitmap).

•

The style change records must be the same for the start and end shapes.

•

If filled with a bitmap, both shapes must be filled with the same bitmap.

•

If filled with a gradient, both gradients must have the same number of color points.

The following illustration shows a morph from a blue rectangle to a red quadrilateral over five frames. The green
outlines represent the ‘in-between’ shapes of the morph sequence. Both shapes have the same number of
points, and the same type of fill, namely a solid fill. Besides changing shape, the shape also gradually changes
color from blue to red.

144

There are two tags involved in defining and playing a morph sequence:
•

DefineMorphShape

•

PlaceObject2

DefineMorphShape defines the start and end states of the morph. A morph object does not use previously
defined shapes; it is considered a special type of shape with only one character ID. DefineMorphShape contains
a list of edges for both the start and end shapes. It also defines the fill and line styles, as they are at the start and
end of the morph sequence.
The PlaceObject 2 tag displays the morph object at some point in time during the morph sequence. The ratio
field controls how far the morph has progressed. A ratio of zero produces a shape identical to the start
condition. A ratio of 65535 produces a shape identical to the end condition.

DefineMorphShape
The DefineMorphShape tag defines the start and end states of a morph sequence. A morph object should be
displayed with the PlaceObject2 tag, where the ratio field specifies how far the morph has progressed.
The minimum file format version is SWF 3.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 46

CharacterId

UI16

ID for this character

StartBounds

RECT

Bounds of the start shape

EndBounds

RECT

Bounds of the end shape

Offset

UI32

Indicates offset to EndEdges

MorphFillStyles

MORPHFILLSTYLEARRAY

Fill style information is stored in the same manner as for a
standard shape; however, each fill consists of interleaved
information based on a single style type to accommodate
morphing.

MorphLineStyles

MORPHLINESTYLEARRAY

Line style information is stored in the same manner as for a
standard shape; however, each line consists of interleaved
information based on a single style type to accommodate
morphing.

StartEdges

SHAPE

Contains the set of edges and the style bits that indicate style
changes (for example, MoveTo, FillStyle, and LineStyle). Number

145

of edges must equal the number of edges in EndEdges.
EndEdges

SHAPE

Contains only the set of edges, with no style information.
Number of edges must equal the number of edges in StartEdges.

•

StartBounds This defines the bounding-box of the shape at the start of the morph.

•

EndBounds - This defines the bounding-box at the end of the morph.

•

MorphFillStyles This contains an array of interleaved fill styles for the start and end shapes. The fill style
for the start shape is followed by the corresponding fill style for the end shape.

•

MorphLineStyles - This contains an array of interleaved line styles.

•

StartEdges - This array specifies the edges for the start shape, and the style change records for both
shapes. Because the StyleChangeRecords must be the same for the start and end shapes, they are
defined only in the StartEdges array.

•

EndEdges - This array specifies the edges for the end shape, and contains no style change records. The
number of edges specified in StartEdges must equal the number of edges in EndEdges.

Strictly speaking, MoveTo records fall into the category of StyleChangeRecords; however, they should be
included in both the StartEdges and EndEdges arrays.
It is possible for an edge to change type over the course of a morph sequence. A straight edge can become a
curved edge and vice versa. In this case, think of both edges as curved. A straight edge can be easily represented
as a curve, by placing the off-curve (control) point at the midpoint of the straight edge, and the on-curve
(anchor) point at the end of the straight edge. The calculation is as follows:
CurveControlDelta.x = StraightDelta.x / 2;
CurveControlDelta.y = StraightDelta.y / 2;
CurveAnchorDelta.x = StraightDelta.x / 2;
CurveAnchorDelta.y = StraightDelta.y / 2;

DefineMorphShape2
The DefineMorphShape2 tag extends the capabilities of DefineMorphShape by using a new morph line style
record in the morph shape. MORPHLINESTYLE2 allows the use of new types of joins and caps as well as scaling
options and the ability to fill the strokes of the morph shape.
DefineMorphShape2 specifies not only the shape bounds but also the edge bounds of the shape. While the
shape bounds are calculated along the outside of the strokes, the edge bounds are taken from the outside of the
edges. For an example of shape bounds versus edge bounds, see the diagram in DefineShape4. The new
StartEdgeBounds and EndEdgeBounds fields assist Flash Player in accurately determining certain layouts.

146

In addition, DefineMorphShape2 includes new hinting information, UsesNonScalingStrokes and
UsesScalingStrokes. These flags assist Flash Player in creating the best possible area for invalidation.
The minimum file format version is SWF 8.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 84

CharacterId

UI16

ID for this character

StartBounds

RECT

Bounds of the start shape

EndBounds

RECT

Bounds of the end shape

StartEdgeBounds

RECT

Bounds of the start shape, excluding strokes

EndEdgeBounds

RECT

Bounds of the end shape, excluding strokes

Reserved

UB[6]

Must be 0

UsesNonScalingStrokes

UB[1]

If 1, the shape contains at least one non-scaling stroke.

UsesScalingStrokes

UB[1]

If 1, the shape contains at least one scaling stroke.

Offset

UI32

Indicates offset to EndEdges

MorphFillStyles

MORPHFILLSTYLEARRAY

Fill style information is stored in the same manner as
for a standard shape; however, each fill consists of
interleaved information based on a single style type to
accommodate morphing.

MorphLineStyles

MORPHLINESTYLEARRAY

Line style information is stored in the same manner as
for a standard shape; however, each line consists of
interleaved information based on a single style type to
accommodate morphing.

StartEdges

SHAPE

Contains the set of edges and the style bits that
indicate style changes (for example, MoveTo, FillStyle,
and LineStyle). Number of edges must equal the
number of edges in EndEdges.

EndEdges

SHAPE

Contains only the set of edges, with no style
information. Number of edges must equal the number
of edges in StartEdges.

147

Morph fill styles
MORPHFILLSTYLEARRAY
A morph fill style array enumerates a number of fill styles.
Field

Type

Comment

FillStyleCount

Count = UI8

Count of fill styles.

FillStyleCountExtended

If Count = 0xFF UI16

Extended count of fill styles.

FillStyles

MORPHFILLSTYLE[count]

Array of fill styles.

MORPHFILLSTYLE
A fill style represents how a closed shape is filled in.
Field

Type

Comment

FillStyleType

UI8

Type of fill style:
0x00 = solid fill
0x10 = linear gradient fill
0x12 = radial gradient fill
0x13 = focal radial gradient fill (SWF 8 file format
and later only)
0x40 = repeating bitmap
0x41 = clipped bitmap fill
0x42 = non-smoothed repeating bitmap
0x43 = non-smoothed clipped bitmap

StartColor

If type = 0x00, RGBA

Solid fill color with opacity information for start
shape.

EndColor

If type = 0x00, RGBA

Solid fill color with opacity information for end
shape.

StartGradientMatrix

If type = 0x10 or 0x12, MATRIX

Matrix for gradient fill for start shape.

148

EndGradientMatrix

If type = 0x10 or 0x12, MATRIX

Matrix for gradient fill for end shape.

Gradient

If type = 0x10 or 0x12,
MORPHGRADIENT

Gradient fill.

BitmapId

If type = 0x40, 0x41, 0x42 or
0x43, UI16

ID of bitmap character for fill.

StartBitmapMatrix

If type = 0x40, 0x41, 0x42 or
0x43, MATRIX

Matrix for bitmap fill for start shape.

EndBitmapMatrix

If type = 0x40, 0x41, 0x42 or
0x43, MATRIX

Matrix for bitmap fill for end shape.

Morph gradient values
Morph gradient values control gradient information for a fill style.

MORPHGRADIENT
The format of gradient information is described in the following table:
Field

Type

Comment

NumGradients

UI8

1 to 8.

GradientRecords

MORPHGRADRECORD [NumGradients]

Gradient records (see following).

MORPHGRADRECORD
The gradient record format is described in the following table:
Field

Type

Comment

StartRatio

UI8

Ratio value for start shape.

StartColor

RGBA

Color of gradient for start shape.

EndRatio

UI8

Ratio value for end shape.

EndColor

RGBA

Color of gradient for end shape.

149

Morph line styles
A morph line style array enumerates a number of fill styles.

MORPHLINESTYLEARRAY
The format of a line style array is described in the following table.
Field

Type

Comment

LineStyleCount

UI8

Count of line styles.

LineStyleCountExtended

If count = 0xFF UI16

Extended count of line styles.

LineStyles

MORPHLINESTYLE[count], (if MorphShape1)
MORPHLINESTYLE2[count], (if MorphShape2)

Array of line styles.

A line style represents a width and color of a line.

MORPHLINESTYLE
The format of a line style value within the file is described in the following table.
Field

Type

Comment

StartWidth

UI16

Width of line in start shape in twips.

EndWidth

UI16

Width of line in end shape in twips.

StartColor

RGBA

Color value including alpha channel information for start shape.

EndColor

RGBA

Color value including alpha channel information for end shape.

MORPHLINESTYLE2
MORPHLINESTYLE2 builds upon the capabilities of the MORPHLINESTYLE record by allowing the use of new
types of joins and caps as well as scaling options and the ability to fill morph strokes. In order to use
MORPHLINESTYLE2, the shape must be defined with DefineMorphShape2—not DefineMorphShape.
While the MORPHLINESTYLE record permits only rounded joins and round caps, MORPHLINESTYLE2 also
supports miter and bevel joins, and square caps and no caps. For an illustration of the available joins and caps,
see the diagram in the LINESTYLE2 description.
When using MORPHLINESTYLE for a miter join, a MiterLimitFactor must be specified and is used along with
150

StartWidth or EndWidth to calculate the maximum miter length:
Max miter length = MORPHLINESTYLE2 MiterLimitFactor * MORPHLINESTYLE2

Width
If the miter join exceeds the maximum miter length, Flash Player will cut off the miter. Note that
MiterLimitFactor is an 8.8 fixed-point value.
MORPHLINESTYLE2 also includes the option for pixel hinting in order to correct blurry vertical or horizontal lines.
Field

Type

Comment

StartWidth

UI16

Width of line in start shape in twips.

EndWidth

UI16

Width of line in end shape in twips.

StartCapStyle

UB[2]

Start-cap style:; 0 = Round cap; 1 = No cap; 2 = Square cap

JoinStyle

UB[2]

Join style:; 0 = Round join; 1 = Bevel join; 2 = Miter join

HasFillFlag

UB[1]

If 1, fill is defined in FillType. If 0, uses StartColor and EndColor fields.

NoHScaleFlag

UB[1]

If 1, stroke thickness will not scale if the object is scaled horizontally.

NoVScaleFlag

UB[1]

If 1, stroke thickness will not scale if the object is scaled vertically.

PixelHintingFlag

UB[1]

If 1, all anchors will be aligned to full pixels.

Reserved

UB[5]

Must be 0.

NoClose

UB[1]

If 1, stroke will not be closed if the stroke’s last point matches its first
point. Flash Player will apply caps instead of a join.

EndCapStyle

UB[2]

End-cap style: 0 = Round cap; 1 = No cap; 2 = Square cap

MiterLimitFactor

If JoinStyle = 2,
UI16

Miter limit factor as an 8.8 fixed-point value.

StartColor

If HasFillFlag = 0,
RGBA

Color value including alpha channel information for start shape.

EndColor

If HasFillFlag = 0,
RGBA

Color value including alpha channel information for end shape.

FillType

If HasFillFlag = 1,
MORPHFILLSTYLE

Fill style.

151

Chapter 10: Fonts and Text
The SWF file format specification supports a variety of text-drawing primitives. In SWF 6 or later files, all text is
represented using Unicode encodings, eliminating dependencies on playback locale for text and strings. As of
version 10, the Flash Player also supports right-to- left scripts and support for Hebrew, Arabic, Thai, and other
complex scripts.

Glyph text and device text
The SWF file format supports two kinds of text: glyph text and device text. Glyph text works by embedding
character shapes in the SWF file, while device text uses the text rendering capabilities of the playback platform.
Glyph text looks identical on all playback platforms. It can be drawn with either the standard anti-aliasing used
by all shapes on a Flash Player Stage, or, in SWF 8 file format and later, rendered with the advanced text
rendering engine. The usage of glyph text creates larger SWF files than for device text, especially in files that use
many different characters from a large character set.
Device text is anti-aliased by the operating system that hosts Flash Player, and its appearance varies depending
on the playback platform. Fonts for device text can be specified in two ways: directly, as a font name that will be
sought verbatim on the playback platform; or indirectly, using one of a small number of special font names that
are mapped to highly available fonts that differ in name from platform to platform, but are chosen to be as
similar in appearance as possible across platforms.
Glyph text characters are defined using the DefineFont, DefineFont2, or DefineFont3 tag. Device text fonts are
defined using the DefineFont and DefineFontInfo tags together, or the DefineFont2 tag. DefineFont2 tags for
device text fonts do not need to include any character glyphs if they will only be used for dynamic text (see next
section), although it is a good idea to include them if there is any doubt about the specified font being available
at playback time on any platform. It is possible to use a given DefineFont or DefineFont2 tag as a glyph font for
certain text blocks, and as a device font for others, as long as both glyphs and character codes are provided.

Static text and dynamic text
Text can be defined as static text or, in SWF 4 file format or later, dynamic text. Dynamic text can be changed
programmatically at runtime, and, optionally, can be made editable for Adobe Flash Player users as well.
Dynamic text can emulate almost all features of static text; exact positioning of individual characters is the only
advantage of static text, aside from implementation effort and version compatibility. Dynamic text also has
many formatting capabilities that static text does not have. These rich formatting capabilities are expressed as a
subset of HTML text-markup tags.
Static text is defined using the DefineText tag. Dynamic text is defined using the DefineEditText tag. Both of
these tags make reference to DefineFont or DefineFont2 tags to obtain their character sources. DefineEditText
152

tags require DefineFont2 tags rather than DefineFont tags; DefineText tags can use either DefineFont or
DefineFont2 tags.
The DefineEditText tag provides a flag that indicates whether to use glyph text or device text. However, the
DefineText tag does not. This means that, for static text, SWF file format provides no means to indicate whether
to use glyph text or device text. This situation is resolved by runtime flags. Normally, all static text is rendered as
glyph text. When a Flash Player plug-in is embedded in an HTML page, an HTML tag option called devicefont will
cause Flash Player to render all static text as device text, if possible; as usual, glyph text is used as a fallback. The
ability of the DefineEditText tag to specify glyph text or device text is another reason to consider dynamic text
superior to static text.

Glyph text
Glyph definitions
Glyphs are defined once in a standard coordinate space called the EM square. The same set of glyphs are used
for every point size of a given font. To render a glyph at different point sizes, Flash Player scales the glyph from
EM coordinates to point-size coordinates.

•

Glyph fonts—without using the advanced text rendering engine —do not include any hinting
information for improving the quality of small font sizes. However, anti-aliasing dramatically improves
the legibility of scaled-down text. Glyph text remains legible down to about 12 points (viewed at 100%).
At 12 points and lower, advanced anti-aliasing is recommended for readable glyph text. This gives
superior text quality at small point sizes and includes extra font meta-information for improved
rendering.

•

TrueType fonts can be readily converted to SWF glyphs. A simple algorithm can replace the Quadratic Bsplines (used by TrueType fonts) with Quadratic Bezier curves (used by SWF glyphs).

Example:

153

To the left is the glyph for the TrueType letter 'b' of Monotype Arial. It is made up of curved and straight edges.
Squares indicate on-curve points, and crosses indicate off-curve points. The black circle is the reference point for
the glyph. The blue outline indicates the bounding box of the glyph.

The EM square
The EM square is an imaginary square that is used to size and align glyphs. The EM square is generally large
enough to completely contain all glyphs, including accented glyphs. It includes the font’s ascent, descent, and
some extra spacing to prevent lines of text from colliding.

SWF glyphs are always defined on an EM square of 1024 by 1024 units. Glyphs from other sources (such as
TrueType fonts) may be defined on a different EM square. To use these glyphs in SWF file format, they should be
scaled to fit an EM square of 1024.

Converting TrueType fonts to SWF glyphs
TrueType glyphs are defined using Quadratic B-Splines, which can be easily converted to the
Quadratic Bezier curves used by SWF glyphs.
A TrueType B-spline is composed of one on-curve point, followed by many off-curve points, followed by another
on-curve point. The midpoint between any two off-curve points is guaranteed to be on the curve. A SWF Bezier
curve is composed of one on-curve point, followed by one off-curve point, followed by another on-curve point.
The conversion from TrueType to SWF curves involves inserting a new on-curve point at the midpoint of two
successive off-curve points.
Example:
Following is a four point B-Spline. P0 and P3 are on-curve points. P1 and P2 are successive off-curve points.

This curve can be represented as two Quadratic Bezier curves by inserting a new point M, at the midpoint of
P1,P2. The result is two Quadratic Bezier curves; P0,P1,M and M,P2,P3.
154

The complete procedure for converting TrueType glyphs to SWF glyphs is as follows:
1. Negate the y-coordinate. (In TrueType glyphs, the y-axis points up; in SWF glyphs, the y- axis points
down.)
2. Scale the x and y co-ordinates from the EM square of the TrueType font, to the EM square of the SWF
glyph (always 1024).
3. Insert an on-curve (anchor) point at the midpoint of each pair of off-curve points.

Kerning and advance values
Kerning defines the horizontal distance between two glyphs. Some font systems store kerning information along
with each font definition. SWF file format, in contrast, stores kerning information with every glyph instance
(every character in a glyph text block). This is referred to as an advance value.

In the preceding example, the A glyph overlaps the V glyph. In this case, the advance is narrower than the width
of the A glyph.

Advanced text rendering engine
Glyph text can be rendered using the normal Flash Player renderer or, in SWF 8 and later, with the advanced
text rendering engine.
The advanced text rendering engine is a high-quality text renderer supported inside the Flash Player renderer.
The advanced system has the following advantages over using the normal renderer for text:
•

Readable, even at small point sizes.

•

Maintains the aesthetic look and feel of a font, even at small point sizes.

•

Supports pixel snapping for ultra-clear text (when left-aligned dynamic text is used).

•

Improved performance over glyph text, typically.

•

LCD sub-pixel rendering when Flash Player detects an LCD screen.

A limitation of the advanced text rendering engine, however, is that it does not animate well as compared to
glyph text.

155

The advanced text rendering engine uses Continuous Stroke Modulation (CSM) parameters to tune its
performance. CSM is the continuous modulation of both stroke weight and edge sharpness. CSM uses two
rendering parameters: inside and outside cutoff. Optimal values for these parameters are highly subjective and
can depend on user preferences, lighting conditions, display properties, typeface, foreground and background
colors, and point size. However, under most circumstances, high-quality type can be achieved with a small set of
interpolated values.
The function that creates the edges for advanced anti-aliasing has an outside cutoff (below which the edge isn’t
drawn) and an inside cutoff (above which the edge is opaque). Between these two cutoff values is a linear
function ranging from zero at the outside cutoff to the maximum value at the inside cutoff.
Adjusting the outside and inside cutoff values affects stroke weight and edge sharpness. The spacing between
these two parameters is comparable to twice the filter radius of classic
anti-aliasing methods: a narrow spacing provides a sharper edge while a wider spacing provides a softer, more
filtered edge. When the spacing is zero, the resulting density image is a bi-level bitmap. When the spacing is very
wide, the resulting density image has a watercolor- like edge. Typically, users prefer sharp, high-contrast edges
at small point sizes and softer edges for animated text and larger point sizes.
The outside cutoff typically has a negative value, the inside cutoff typically has a positive value, and their
midpoint typically lies near zero. Adjusting these parameters to shift the midpoint towards negative infinity will
increase the stroke weight; shifting the midpoint towards positive infinity will decrease the stroke weight. Note
that the outside cutoff should always be less than or equal to the inside cutoff.
Flash Player creates a table of CSM parameters as a function of text size and text color for each advanced antialiased font in use. This default table typically provides a suitable set of CSM settings across a wide range of
point sizes. However, you can specify a user-defined table to replace the default table by using the ActionScript
function setAdvancedAntialiasingTable().
The CSM parameters are intended to make fonts more readable and not to create effects. Extreme values of
CSM result in rendering artifacts. To apply effects to text, it is much better to use reasonable CSM values and
then apply filters or blend effects.

DefineFont and DefineText
Of the four text types supported in SWF file format (static glyph, static device, dynamic glyph, and dynamic
device), the most complex is static glyph text. The other types use simpler variations on the rules used for
defining static glyph text.
Static glyph text is defined using two tags:

156

•

The DefineFont tag defines a set of glyphs.

•

The DefineText tag defines the text string that is displayed in the font.

The DefineFont tag defines all the glyphs used by subsequent DefineText tags. DefineFont includes an array of
SHAPERECORDs, which describe the outlines of the glyphs. These shape records are the same records used by
DefineShape to define non-text shapes. To keep file size to a minimum, only the glyphs actually used are
included in the DefineFont tag.
The DefineText tag stores the actual text string(s) to be displayed, represented as a series of glyph indices. It also
includes the bounding box of the text object, a transformation matrix, and style attributes such as color and size.
DefineText contains an array of TEXTRECORDs. A TEXTRECORD selects the current font, color, and point size, as
well as the x and y position of the next character in the text. These styles apply to all characters that follow, until
another TEXTRECORD changes the styles. A TEXTRECORD also contains an array of indices into the glyph table of
the current font. Characters are not referred to by their character codes, as in traditional programming, but
rather by an index into the glyph table. The glyph data also includes the advance value for each character in the
text string.
Note: A DefineFont tag must always come before any DefineText tags that refer to it.

Static glyph text example
Consider the example of displaying the static glyph text bob in the Arial font, with a point size of 24.
First, define the glyphs with a DefineFont tag. The glyph table, of type SHAPE, has two SHAPERECORDs. At index
0 is the shape of a lowercase b (see diagram). At index 1 is the shape of a lowercase o. (The second b in bob is a
duplicate, and is not required). DefineFont also includes a unique ID so it can be selected by the DefineText tag.
Next, define the text itself with a DefineText tag. The TEXTRECORD sets the position of the first character, selects
the Arial font (using the font’s character ID), and sets the point size to 24, so the font is scaled to the correct
size. (Remember that glyphs are defined in EM coordinates—the actual point size is part of the DefineText tag).
It also contains an array of GLYPHENTRYs. Each glyph entry contains an index into the font’s shape array. In this
example, the first glyph entry has index 0 (which corresponds to the b shape), the second entry has index 1 (the
o), and the third entry has index 0 (b again). Each GLYPHENTRY also includes an advance value for accurately
positioning the glyph.
The following diagram illustrates how the DefineText tag interacts with the DefineFont tag:
DefineFont
SHAPE Array:

DefineText
TEXTRECORD

Glyph 'b'

Glyph index to 'b'

Glyph 'o'

Glyph index to 'o'
Glyph inex to 'b'

157

Font tags
DefineFont
The DefineFont tag defines the shape outlines of each glyph used in a particular font. Only the glyphs that are
used by subsequent DefineText tags are actually defined.
DefineFont tags cannot be used for dynamic text. Dynamic text requires the DefineFont2 tag. The minimum file
format version is SWF 1.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 10

FontID

UI16

ID for this font character

OffsetTable

UI16[nGlyphs]

Array of shape offsets

GlyphShapeTable

SHAPE[nGlyphs]

Array of shapes

The font ID uniquely identifies the font. It can be used by subsequent DefineText tags to select the font. Like all
SWF character IDs, font IDs must be unique among all character IDs in a SWF file.
If you provide a DefineFontInfo tag to go along with a DefineFont tag, be aware that the order of the glyphs in
the DefineFont tag must match the order of the character codes in the DefineFontInfo tag, which must be sorted
by code point order.
The OffsetTable and GlyphShapeTable are used together. These tables have the same number of entries, and
there is a one-to-one ordering match between the order of the offsets and the order of the shapes. The
OffsetTable points to locations in the GlyphShapeTable. Each offset entry stores the difference (in bytes)
between the start of the offset table and the location of the corresponding shape. Because the GlyphShapeTable
immediately follows the OffsetTable, the number of entries in each table (the number of glyphs in the font) can
be inferred by dividing the first entry in the OffsetTable by two.
The first STYLECHANGERECORD of each SHAPE in the GlyphShapeTable does not use the LineStyle and LineStyles
fields. In addition, the first STYLECHANGERECORD of each shape must have both fields StateFillStyle0 and
FillStyle0 set to 1.

DefineFontInfo
The DefineFontInfo tag defines a mapping from a glyph font (defined with DefineFont) to a device font. It
provides a font name and style to pass to the playback platform’s text engine, and a table of character codes
158

that identifies the character represented by each glyph in the corresponding DefineFont tag, allowing the glyph
indices of a DefineText tag to be converted to character strings.
The presence of a DefineFontInfo tag does not force a glyph font to become a device font; it merely makes the
option available. The actual choice between glyph and device usage is made according to the value of devicefont
(see the introduction) or the value of UseOutlines in a DefineEditText tag. If a device font is unavailable on a
playback platform, Flash Player will fall back to glyph text.
The minimum file format version is SWF 1.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 13.

FontID

UI16

Font ID this information is for.

FontNameLen

UI8

Length of font name.

FontName

UI8[FontNameLen]

Name of the font (see following).

FontFlagsReserved

UB[2]

Reserved bit fields.

FontFlagsSmallText

UB[1]

SWF 7 file format or later: Font is small.
Character glyphs are aligned on pixel
boundaries for dynamic and input text.

FontFlagsShiftJIS

UB[1]

ShiftJIS character codes.

FontFlagsANSI

UB[1]

ANSI character codes.

FontFlagsItalic

UB[1]

Font is italic.

FontFlagsBold

UB[1]

Font is bold.

FontFlagsWideCodes

UB[1]

If 1, CodeTable is UI16 array; otherwise,
CodeTable is UI8 array.

CodeTable

If FontFlagsWideCodes,
UI16[nGlyphs], Otherwise,
UI8[nGlyphs]

Glyph to code table, sorted in ascending
order.

The entries in the CodeTable must be sorted in ascending order by code point, by the value they provide. The
order of the entries in the CodeTable must also match the order of the glyphs in the DefineFont tag to which this
DefineFontInfo tag applies. This places a requirement on the ordering of glyphs in the corresponding DefineFont
tag.
159

SWF 6 or later files require Unicode text encoding. One aspect of this requirement is that all character code
tables are specified using UCS-2 (UCS-2 is generally the first 64k code points of UTF-16). This encoding uses a
fixed 2 bytes for each character. This means that when a DefineFontInfo tag appears in a SWF 6 or later file,
FontFlagsWideCodes must be set, FontFlagsShiftJIS and FontFlagsANSI must be cleared, and CodeTable must
consist of UI16 entries (as always, in little-endian byte order) encoded in UCS-2.
Another Unicode requirement that applies to SWF 6 or later files is that font names must always be encoded
using UTF-8. In SWF 5 or earlier files, font names are encoded in a platform-specific way, in the codepage of the
system on which they were authored. The playback platform will interpret them using its current codepage, with
potentially inconsistent results. If the playback platform is an ANSI system, font names will be interpreted as
ANSI strings. If the playback platform is a Japanese shift-JIS system, font names will be interpreted as shift-JIS
strings. Many other values for the playback platform’s codepage, which governs this decision, are possible. This
playback locale dependency is undesirable, which is why SWF 6 file format moved toward a standard encoding
for font names. Note that font name strings in the DefineFontInfo tag are not null-terminated; instead their
length is specified by the FontNameLen field. FontNameLen specifies the number of bytes in FontName, which is
not necessarily equal to the number of characters, since some encodings may use more than one byte per
character.
Font names are normally used verbatim, passed directly to the playback platform’s font system in order to
locate a font. However, there are several special indirect font names that are mapped to different actual font
names depending on the playback platform. These indirect mappings are hard-coded into each platform-specific
port of Flash Player, and the fonts for each platform are chosen from among system default fonts or other fonts
that are very likely to be available. As a secondary consideration, the indirect mappings are chosen so as to
maximize the similarity of indirect fonts across platforms.
The following tables describe the indirect font names that are supported.

Western indirect fonts
Font name

Example

_sans

Hello world

_serif

Hello world

_typewriter

Hello world

Japanese indirect fonts
Font name:
English Name:

160

Gothic

UTF-8 Byte String (hex):

5F E3 82

B4 E3 82

B7 E3 83

83

E3 82

AF

Example appearance:

Font name:
English Name:

Tohaba (Gothic Mono)

UTF-8 Byte String (hex):

5F E7 AD 89

E5 B9 85

Example appearance:

Font name:
English Name:

Mincho

UTF-8 Byte String (hex):

5F E6 98 8E E6 9C 9D

Example appearance:

DefineFontInfo2
When generating SWF 6 or later, it is recommended that you use the new DefineFontInfo2 tag rather than
DefineFontInfo. DefineFontInfo2 is identical to DefineFontInfo, except that it adds a field for a language code. If
you use the older DefineFontInfo, the language code will be assumed to be zero, which results in behavior that is
dependent on the locale in which Flash Player is running.
The minimum file format version is SWF 6.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 62.

FontID

UI16

Font ID this information is for.

FontNameLen

UI8

Length of font name.

FontName

UI8[FontNameLen]

Name of the font.

FontFlagsReserved

UB[2]

Reserved bit fields.

FontFlagsSmallText

UB[1]

SWF 7 or later: Font is small. Character glyphs are aligned
on pixel boundaries for dynamic and input text.

161

FontFlagsShiftJIS

UB[1]

Always 0.

FontFlagsANSI

UB[1]

Always 0.

FontFlagsItalic

UB[1]

Font is italic.

FontFlagsBold

UB[1]

Font is bold.

FontFlagsWideCodes

UB[1]

Always 1.

LanguageCode

LANGCODE

Language ID.

CodeTable

UI16[nGlyphs]

Glyph to code table in UCS-2, sorted in ascending order.

DefineFont2
The DefineFont2 tag extends the functionality of DefineFont. Enhancements include the following:
•

32-bit entries in the OffsetTable, for fonts with more than 64K glyphs.

•

Mapping to device fonts, by incorporating all the functionality of DefineFontInfo.

•

Font metrics for improved layout of dynamic glyph text.

DefineFont2 tags are the only font definitions that can be used for dynamic text.
The minimum file format version is SWF 3.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 48.

FontID

UI16

ID for this font character.

FontFlagsHasLayout

UB[1]

Has font metrics/layout information.

FontFlagsShiftJIS

UB[1]

ShiftJIS encoding.

FontFlagsSmallText

UB[1]

SWF 7 or later: Font is small. Character
glyphs are aligned on pixel boundaries for
dynamic and input text.

FontFlagsANSI

UB[1]

ANSI encoding.

FontFlagsWideOffsets

UB[1]

If 1, uses 32 bit offsets.

FontFlagsWideCodes

UB[1]

If 1, font uses 16-bit codes; otherwise font

162

uses 8 bit codes.
FontFlagsItalic

UB[1]

Italic Font.

FontFlagsBold

UB[1]

Bold Font.

LanguageCode

LANGCODE

SWF 5 or earlier: always 0; SWF 6 or later:
language code

FontNameLen

UI8

Length of name.

FontName

UI8[FontNameLen]

Name of font (see DefineFontInfo).

NumGlyphs

UI16

Count of glyphs in font. May be zero for
device fonts.

OffsetTable

If FontFlagsWideOffsets,
UI32[NumGlyphs] Otherwise
UI16[NumGlyphs]

Same as in DefineFont.

CodeTableOffset

If FontFlagsWideOffsets, UI32
Otherwise UI16

Byte count from start of OffsetTable to start
of CodeTable.

GlyphShapeTable

SHAPE[NumGlyphs]

Same as in DefineFont.

CodeTable

If FontFlagsWideCodes,
UI16[NumGlyphs] Otherwise
UI8[NumGlyphs]

Sorted in ascending order. Always UCS-2 in
SWF 6 or later.

FontAscent

If FontFlagsHasLayout, UI16

Font ascender height.

FontDescent

If FontFlagsHasLayout, UI16

Font descender height.

FontLeading

If FontFlagsHasLayout, SI16

Font leading height (see following).

FontAdvanceTable

If FontFlagsHasLayout,
SI16[NumGlyphs]

Advance value to be used for each glyph in
dynamic glyph text.

FontBoundsTable

If FontFlagsHasLayout,
RECT[NumGlyphs]

Not used in Flash Player through version 7
(but must be present).

KerningCount

If FontFlagsHasLayout, UI16

Not used in Flash Player through version 7
(always set to 0 to save space).

FontKerningTable

If FontFlagsHasLayout,
KERNINGRECORD[KerningCount]

Not used in Flash Player through version 7
(omit with KerningCount of 0).

163

In SWF 6 or later files, DefineFont2 has the same Unicode requirements as DefineFontInfo. Similarly to the
DefineFontInfo tag, the CodeTable (and thus also the OffsetTable, GlyphShapeTable, and FontAdvanceTable)
must be sorted in code point order.
If a DefineFont2 tag will be used only for dynamic device text, and no glyph-rendering fallback is desired, set
NumGlyphs to zero, and omit all tables having NumGlyphs entries. This will substantially reduce the size of the
DefineFont2 tag. DefineFont2 tags without glyphs cannot support static text, which uses glyph indices to select
characters, and also cannot support glyph text, which requires glyph shape definitions.
Layout information (ascent, descent, leading, advance table, bounds table, kerning table) is useful only for
dynamic glyph text. This information takes the place of the per-character placement information that is used in
static glyph text. The layout information in the DefineFont2 tag is fairly standard font-metrics information that
can typically be extracted directly from a standard font definition, such as a TrueType font.
Note: Leading is a vertical line-spacing metric. It is the distance (in EM-square coordinates) between the bottom
of the descender of one line and the top of the ascender of the next line.
As with DefineFont, in DefineFont2 the first STYLECHANGERECORD of each SHAPE in the GlyphShapeTable does
not use the LineStyle and LineStyles fields. In addition, the first STYLECHANGERECORD of each shape must have
both fields StateFillStyle0 and FillStyle0 set to 1.
The DefineFont2 tag reserves space for a font bounds table and kerning table. This information is not used in
Flash Player through version 7. However, this information must be present in order to constitute a well-formed
DefineFont2 tag. Supply minimal (low-bit) RECTs for FontBoundsTable, and always set KerningCount to zero,
which allows FontKerningTable to be omitted.

DefineFont3
The DefineFont3 tag is introduced along with the DefineFontAlignZones tag in SWF 8. The DefineFontAlignZones
tag is optional but recommended for SWF files using advanced anti- aliasing, and it modifies the DefineFont3
tag.
The DefineFont3 tag extends the functionality of DefineFont2 by expressing the SHAPE coordinates in the
GlyphShapeTable at 20 times the resolution. All the EMSquare coordinates are multiplied by 20 at export,
allowing fractional resolution to 1/20 of a unit. This allows for more precisely defined glyphs and results in
better visual quality.
The minimum file format version is SWF 8.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 75.

164

FontID

UI16

ID for this font character.

FontFlagsHasLayout

UB[1]

Has font metrics/layout information.

FontFlagsShiftJIS

UB[1]

ShiftJIS encoding.

FontFlagsSmallText

UB[1]

SWF 7 or later: Font is small. Character glyphs
are aligned on pixel boundaries for dynamic
and input text.

FontFlagsANSI

UB[1]

ANSI encoding.

FontFlagsWideOffsets

UB[1]

If 1, uses 32 bit offsets.

FontFlagsWideCodes

UB[1]

Must be 1.

FontFlagsItalic

UB[1]

Italic Font.

FontFlagsBold

UB[1]

Bold Font.

LanguageCode

LANGCODE

SWF 5 or earlier: always 0; SWF 6 or later:
language code

FontNameLen

UI8

Length of name.

FontName

UI8[FontNameLen]

Name of font (see DefineFontInfo).

NumGlyphs

UI16

Count of glyphs in font. May be zero for
device fonts.

OffsetTable

If FontFlagsWideOffsets,
UI32[NumGlyphs] Otherwise
UI16[NumGlyphs]

Same as in DefineFont.

CodeTableOffset

If FontFlagsWideOffsets, UI32
Otherwise UI16

Byte count from start of OffsetTable to start
of CodeTable.

GlyphShapeTable

SHAPE[NumGlyphs]

Same as in DefineFont.

CodeTable

UI16[NumGlyphs]

Sorted in ascending order. Always UCS-2 in
SWF 6 or later.

FontAscent

If FontFlagsHasLayout, UI16

Font ascender height.

FontDescent

If FontFlagsHasLayout, UI16

Font descender height.

FontLeading

If FontFlagsHasLayout, SI16

Font leading height (see following).

165

FontAdvanceTable

If FontFlagsHasLayout,
SI16[NumGlyphs]

Advance value to be used for each glyph in
dynamic glyph text.

FontBoundsTable

If FontFlagsHasLayout,
RECT[NumGlyphs]

Not used in Flash Player through version 7
(but must be present).

KerningCount

If FontFlagsHasLayout, UI16

Not used in Flash Player through version 7
(always set to 0 to save space).

FontKerningTable

If FontFlagsHasLayout,
KERNINGRECORD [KerningCount]

Not used in Flash Player through version 7
(omit with KerningCount of 0).

DefineFontAlignZones
The DefineFont3 tag can be modified by a DefineFontAlignZones tag. The advanced text rendering engine uses
alignment zones to establish the borders of a glyph for pixel snapping. Alignment zones are critical for highquality display of fonts.
The alignment zone defines a bounding box for strong vertical and horizontal components of a glyph. The box is
described by a left coordinate, thickness, baseline coordinate, and height. Small thicknesses or heights are often
set to 0.
For example, consider the letter I. The letter I has a strong horizontal at its baseline and the top of the letter. The
letter I also has strong verticals that occur at the edges of the stem—not the short top bar or serif. These strong
verticals and horizontals of the center block of the letter define the alignment zones.

The minimum file format version is SWF 8.
DefineFontAlignZones
Field

Type

Comment

Header

RECORDHEADER

Tag type = 73.

FontID

UI16

ID of font to use, specified by DefineFont3.

166

CSMTableHint

UB[2]

Font thickness hint. Refers to the thickness of the
typical stroke used in the font. 0 = thin; 1 = medium;
2 = thick; Flash Player maintains a selection of CSM
tables for many fonts. However, if the font is not
found in Flash Player's internal table, this hint is used
to choose an appropriate table.

Reserved

UB[6]

Must be 0.

ZoneTable

ZONERECORD[GlyphCount]

Alignment zone information for each glyph.

ZONERECORD
Field

Type

Comment

NumZoneData

UI8

Number of ZoneData entries. Always 2.

ZoneData

ZONEDATA[NumZoneData]

Compressed alignment zone information.

Reserved

UB[6]

Must be 0.

ZoneMaskY

UB[1]

Set if there are Y alignment zones.

ZoneMaskX

UB[1]

Set if there are X alignment zones.

ZONEDATA
Field

Type

Comment

AlignmentCoordinate

FLOAT16

X (left) or Y (baseline) coordinate of the alignment
zone.

Range

FLOAT16

Width or height of the alignment zone.

Kerning record
A Kerning Record defines the distance between two glyphs in EM square coordinates. Certain pairs of glyphs
appear more aesthetically pleasing if they are moved closer together, or farther apart. The FontKerningCode1
and FontKerningCode2 fields are the character codes for the left and right characters. The
FontKerningAdjustment field is a signed integer that defines a value to be added to the advance value of the left
character.

167

Field

Type

Comment

FontKerningCode1

If FontFlagsWideCodes, UI16
Otherwise UI8

Character code of the left character.

FontKerningCode2

If FontFlagsWideCodes, UI16
Otherwise UI8

Character code of the right character.

FontKerningAdjustment

SI16

Adjustment relative to left character’s advance
value.

DefineFontName
The DefineFontName tag contains the name and copyright information for a font embedded in the SWF file.
The minimum file format version is SWF 9.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 88

FontID

UI16

ID for this font to which this refers

FontName

STRING

Name of the font. For fonts starting
as Type 1, this is the PostScript
FullName. For fonts starting in sfnt
formats such as TrueType and
OpenType, this is name ID 4,
platform ID 1, language ID 0 (Full
name, Mac OS, English).

FontCopyright

STRING

Arbitrary string of copyright
information

Static text tags
DefineText
The DefineText tag defines a block of static text. It describes the font, size, color, and exact position of every
character in the text object.
The minimum file format version is SWF 1.
168

Field

Type

Comment

Header

RECORDHEADER

Tag type = 11.

CharacterID

UI16

ID for this text character.

TextBounds

RECT

Bounds of the text.

TextMatrix

MATRIX

Transformation matrix for the text.

GlyphBits

UI8

Bits in each glyph index.

AdvanceBits

UI8

Bits in each advance value.

TextRecords

TEXTRECORD[zero or more]

Text records.

EndOfRecordsFlag

UI8

Must be 0.

The TextBounds field is the rectangle that completely encloses all the characters in this text block.
The GlyphBits and AdvanceBits fields define the number of bits used for the GlyphIndex and GlyphAdvance
fields, respectively, in each GLYPHENTRY record.

Text records
A TEXTRECORD sets text styles for subsequent characters. It can be used to select a font, change the text color,
change the point size, insert a line break, or set the x and y position of the next character in the text. The new
text styles apply until another TEXTRECORD changes the styles.
The TEXTRECORD also defines the actual characters in a text object. Characters are referred to by an index into
the current font’s glyph table, not by a character code. Each TEXTRECORD contains a group of characters that all
share the same text style, and are on the same line of text.
Field

Type

Comment

TextRecordType

UB[1]

Always 1.

StyleFlagsReserved

UB[3]

Always 0.

StyleFlagsHasFont

UB[1]

1 if text font specified.

StyleFlagsHasColor

UB[1]

1 if text color specified.

StyleFlagsHasYOffset

UB[1]

1 if y offset specified.

169

StyleFlagsHasXOffset

UB[1]

1 if x offset specified.

FontID

If StyleFlagsHasFont, UI16

Font ID for following text.

TextColor

If StyleFlagsHasColor, RGB If this record is
part of a DefineText2 tag, RGBA

Font color for following text.

XOffset

If StyleFlagsHasXOffset, SI16

x offset for following text.

YOffset

If StyleFlagsHasYOffset, SI16

y offset for following text.

TextHeight

If hasFont, UI16

Font height for following text.

GlyphCount

UI8

Number of glyphs in record.

GlyphEntries

GLYPHENTRY[GlyphCount]

Glyph entry (see following).

The FontID field is used to select a previously defined font. This ID uniquely identifies a DefineFont or
DefineFont2 tag from earlier in the SWF file.
The TextHeight field defines the height of the font in twips. For example, a pixel height of 50 is equivalent to a
TextHeight of 1000. (50 * 20 = 1000).
The XOffset field defines the offset from the left of the TextBounds rectangle to the reference point of the glyph
(the point within the EM square from which the first curve segment departs). Typically, the reference point is on
the baseline, near the left side of the glyph (see the example for Glyph text). The XOffset is generally used to
create indented text or non-left- justified text. If there is no XOffset specified, the offset is assumed to be zero.
The YOffset field defines the offset from the top of the TextBounds rectangle to the reference point of the glyph.
The TextYOffset is generally used to insert line breaks, moving the text position to the start of a new line.
The GlyphCount field defines the number of characters in this string, and the size of the GLYPHENTRY table.

Glyph entry
The GLYPHENTRY structure describes a single character in a line of text. It is composed of an index into the
current font’s glyph table, and an advance value. The advance value is the horizontal distance between the
reference point of this character and the reference point of the following character.
Field

Type

Comment

GlyphIndex

UB[GlyphBits]

Glyph index into current font.

GlyphAdvance

SB[AdvanceBits]

x advance value for glyph.

170

DefineText2
The DefineText2 tag is almost identical to the DefineText tag. The only difference is that Type 1 text records
contained within a DefineText2 tag use an RGBA value (rather than an RGB value) to define TextColor. This
allows partially or completely transparent characters.
Text defined with DefineText2 is always rendered with glyphs. Device text can never include transparency.
The minimum file format version is SWF 3.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 33.

CharacterID

UI16

ID for this text character.

TextBounds

RECT

Bounds of the text.

TextMatrix

MATRIX

Transformation matrix.

GlyphBits

UI8

Bits in each glyph index.

AdvanceBits

UI8

Bits in each advance value.

TextRecords

TEXTRECORD[zero or more]

Text records.

EndOfRecordsFlag

UI8

Must be 0.

Dynamic text tags
DefineEditText
The DefineEditText tag defines a dynamic text object, or text field.
A text field is associated with an ActionScript variable name where the contents of the text field are stored. The
SWF file can read and write the contents of the variable, which is always kept in sync with the text being
displayed. If the ReadOnly flag is not set, users may change the value of a text field interactively.
Fonts used by DefineEditText must be defined using DefineFont2, not DefineFont. The minimum file format
version is SWF 4.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 37.

CharacterID

UI16

ID for this dynamic text character.

171

Bounds

RECT

Rectangle that completely encloses the text field.

HasText

UB[1]

0 = text field has no default text. 1 = text field initially
displays the string specified by InitialText.

WordWrap

UB[1]

0 = text will not wrap and will scroll sideways. 1 = text will
wrap automatically when the end of line is reached.

Multiline

UB[1]

0 = text field is one line only. 1 = text field is multi-line and
scrollable.

Password

UB[1]

0 = characters are displayed as typed. 1 = all characters are
displayed as an asterisk.

ReadOnly

UB[1]

0 = text editing is enabled. 1 = text editing is disabled.

HasTextColor

UB[1]

0 = use default color. 1 = use specified color (TextColor).

HasMaxLength

UB[1]

0 = length of text is unlimited. 1 = maximum length of string
is specified by MaxLength.

HasFont

UB[1]

0 = use default font. 1 = use specified font (FontID) and
height (FontHeight). (Can’t be true if HasFontClass is true).

HasFontClass

UB[1]

0 = no fontClass, 1 = fontClass and Height specified for this
text. (can't be true if HasFont is true). Supported in Flash
Player 9.0.45.0 and later.

AutoSize

UB[1]

0 = fixed size. 1 = sizes to content (SWF 6 or later only).

HasLayout

UB[1]

Layout information provided.

NoSelect

UB[1]

Enables or disables interactive text selection.

Border

UB[1]

Causes a border to be drawn around the text field.

WasStatic

UB[1]

0 = Authored as dynamic text; 1 = Authored as static text

HTML

UB[1]

0 = plaintext content. 1 = HTML content (see following).

UseOutlines

UB[1]

0 = use device font. 1 = use glyph font.

FontID

If HasFont, UI16

ID of font to use.

FontClass

If HasFontClass, STRING

Class name of font to be loaded from another SWF and
used for this text.

172

FontHeight

If HasFont, UI16

Height of font in twips.

TextColor

If HasTextColor, RGBA

Color of text.

MaxLength

If HasMaxLength, UI16

Text is restricted to this length.

Align

If HasLayout, UI8

0 = Left; 1 = Right; 2 = Center; 3 = Justify

LeftMargin

If HasLayout, UI16

Left margin in twips.

RightMargin

If HasLayout, UI16

Right margin in twips.

Indent

If HasLayout, UI16

Indent in twips.

Leading

If HasLayout, SI16

Leading in twips (vertical distance between bottom of
descender of one line and top of ascender of the next).

VariableName

STRING

Name of the variable where the contents of the text field
are stored. May be qualified with dot syntax or slash syntax
for non-global variables.

InitialText

If HasText STRING

Text that is initially displayed.

If the HTML flag is set, the contents of InitialText are interpreted as a limited subset of the HTML tag language,
with a few additions not normally present in HTML. The following tags are supported:
Tag

Description

<p> ... </p>

Defines a paragraph. The attribute align may be present, with value left,
right, or center.

<br>

Inserts a line break.
... </a>

Defines a hyperlink. The attribute href must be present. The attribute target
is optional, and specifies a window name.

<font> ... </font>

Defines a span of text that uses a given font. The following attributes are
available: • face, which specifies a font name that must match a font name
supplied in a DefineFont2 tag; • size, which is specified in twips, and may
include a leading ‘+’ or ‘-’ for relative sizes; • color, which is specified as a
#RRGGBB hex triplet

<b> ... </b>

Defines a span of bold text.

<i> ... </i>

Defines a span of italic text.

<a>

173

<u> ... </u>

Defines a span of underlined text.

<li> ... </li>

Defines a bulleted paragraph. The <ul> tag is not necessary and is not
supported. Numbered lists are not supported.

<textformat> ...
</textformat>

Defines a span of text with certain formatting options. The following
attributes are available:;
• leftmargin, which specifies the left margin in twips;
• rightmargin, which specifies the right margin in twips;
• indent, which specifies the left indent in twips;
• blockindent, which specifies a block indent in twips;
• leading, which specifies the leading in twips;
• tabstops, which specifies a comma-separated list of tab stops, each
specified in twips
Inserts a tab character, which advances to the next tab stop as defined with
the <textformat> tag.

<tab>

CSMTextSettings
In addition to the advanced text rendering tags discussed earlier in this chapter, the rendering engine also
supports a tag for modifying text fields. The CSMTextSettings tag modifies a previously streamed DefineText,
DefineText2, or DefineEditText tag. The CSMTextSettings tag turns advanced anti-aliasing on or off for a text
field, and can also be used to define quality and options.
The minimum file format version is SWF 8.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 74.

TextID

UI16

ID for the DefineText, DefineText2, or DefineEditText to which
this tag applies.

UseFlashType

UB[2]

0 = use normal renderer. 1 = use advanced text rendering
engine.

GridFit

UB[3]

0 = Do not use grid fitting. AlignmentZones and LCD sub-pixel
information will not be used. 1 = Pixel grid fit. Only supported
for left-aligned dynamic text. This setting provides the ultimate

174

in advanced anti-aliased text readability, with crisp letters
aligned to pixels. 2 = Sub-pixel grid fit. Align letters to the 1/3
pixel used by LCD monitors. Can also improve quality for CRT
output.
Reserved

UB[3]

Must be 0.

Thickness

F32

The thickness attribute for the associated text field. Set to 0.0
to use the default (anti-aliasing table) value.

Sharpness

F32

The sharpness attribute for the associated text field. Set to 0.0
to use the default (anti-aliasing table) value.

Reserved

UI8

Must be 0.

The Thickness and Sharpness fields are interpretations of the CSM parameters, applied to a particular text field.
The thickness and sharpness setting will override the default CSM setting for that text field.
The CSM parameters, at render time, are computed from the thickness and sharpness:
outsideCutoff = (0.5f * sharpness - thickness) * fontSize;
insideCutoff = (-0.5f * sharpness - thickness) * fontSize;

Using the font size in the cutoff calculations results in linear scaling of CSM parameters, and linear scaling tends
to be a poor approximation when significant scaling is applied. When a text field will scale, it is usually better to
use the default table or provide your own anti- aliasing table.

DefineFont4
DefineFont4 supports only the new Flash Text Engine. The storage of font data for embedded fonts is in CFF
format.
The minimum file format version is SWF 10.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 91

FontID

UI16

ID for this font character.

FontFlagsReserved

UB[5]

Reserved bit fields.

FontFlagsHasFontData

UB[1]

Font is embedded. Font tag includes SFNT font data block.

FontFlagsItalic

UB[1]

Italic font

175

FontFlagsBold

UB[1]

Bold font

FontName

STRING

Name of the font.

FontData

FONTDATA[0 or 1]

When present, this is an OpenType CFF font, as defined in the
OpenType specification at
www.microsoft.com/typography/otspec. The following tables
must be present: ‘CFF ’, ‘cmap’, ‘head’, ‘maxp’, ‘OS/2’, ‘post’,
and either (a) ‘hhea’ and ‘hmtx’, or (b), ‘vhea’, ‘vmtx’, and
‘VORG’. The ‘cmap’ table must include one of the following
kinds of Unicode ‘cmap’ subtables: (0, 4), (0, 3), (3, 10), (3, 1), or
(3, 0) [notation: (platform ID, platform- specific encoding ID)].
Tables such as ‘GSUB’, ‘GPOS’, ‘GDEF’, and ‘BASE’ may also be
present. Only present for embedded fonts.

176

Chapter 11: Sounds
The SWF file format specification defines a small and efficient sound model. SWF supports several audio coding
formats and can store the audio using a range of sample rates in both bstereo and mono. Adobe Flash Player
supports rate conversion and multichannel mixing of these sounds. The number of simultaneous channels
supported depends on the CPU resources available to the Flash Player, but is typically three to eight channels.
There are two types of sounds in SWF file format:
•

Event sounds

•

Streaming sounds

Event sounds are played in response to some event such as a mouse click, or when Flash Player reaches a certain
frame. Event sounds must be defined (downloaded) before they are used. They can be reused for multiple
events if desired. Event sounds may also have a sound “style” that modifies how the basic sound is played.
Streaming sounds are downloaded and played in tight synchronization with the timeline. In this mode, sound
packets are stored with each frame.
Note: The exact sample rates used are as follows. These are standard sample rates based on CD audio, which is
sampled at 44100 Hz. The four sample rates are one-eighth, one- quarter, one-half, and exactly the 44100 Hz
sampling rate.
5.5 kHz = 5512 Hz
11 kHz = 11025 Hz
22 kHz = 22050 Hz
44 kHz = 44100 Hz

Audio coding formats
The Flash Player can store audio using a variety of coding formats. Each of these will be described more
thoroughly in later sections of these chapters. This section lists the coding formats supported, the format
number that the Flash Player uses to reference that format, and the first SWF version that recognizes the format
number.
Coding format

Audio format number

Minimum SWF version

Uncompressed, native-endian

0

1

ADPCM

1

1

MP3

2

4

177

Uncompressed, little-endian

3

4

Nellymoser 16 kHz

4

10

Nellymoser 8 kHz

5

10

Nellymoser

6

6

Speex

11

10

Event sounds
There are several control tags and records required to play an event sound:
•

The DefineSound tag provides the audio samples that make up an event sound.

•

The SOUNDINFO record defines the styles that are applied to the event sound. Styles include fade-in,
fade-out, synchronization and looping flags, and envelope control.

•

The StartSound tag instructs the Flash Player to begin playing the sound.

•

The StartSound2 tag instructs the Flash Player to begin playing a sound class from another SWF.

DefineSound
The DefineSound tag defines an event sound. It includes the audio coding format, sampling rate, size of each
sample (8 or 16 bit), a stereo/mono flag, and an array of audio samples. Note that not all of these parameters
will be honored depending on the audio coding format.
The minimum file format version is SWF 1.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 14

SoundId

UI16

ID for this sound.

SoundFormat

UB[4]

Format of SoundData. See “Audio coding formats”.

SoundRate

UB[2]

The sampling rate. This is ignored for Nellymoser and Speex
codecs. 5.5kHz is not allowed for MP3. 0 = 5.5 kHz; 1 = 11 kHz; 2
= 22 kHz; 3 = 44 kHz

SoundSize

UB[1]

Size of each sample. This parameter only pertains to
uncompressed formats. This is ignored for compressed formats
which always decode to 16 bits internally. 0 = snd8Bit; 1 =

178

snd16Bit
SoundType

UB[1]

Mono or stereo sound. This is ignored for Nellymoser and
Speex. 0 = sndMono; 1 = sndStereo

SoundSampleCount

UI32

Number of samples. Not affected by mono/stereo setting; for
stereo sounds this is the number of sample pairs.

SoundData

UI8[size of sound data]

The sound data; varies by format.

The SoundId field uniquely identifies the sound so it can be played by StartSound. Format 0 (uncompressed) and
Format 3 (uncompressed little-endian) are similar. Both encode uncompressed audio samples. For 8-bit samples,
the two formats are identical. For 16- bit samples, the two formats differ in byte ordering. Using format 0, 16-bit
samples are encoded and decoded according to the native byte ordering of the platform on which the encoder
and Flash Player, respectively, are running. Using format 3, 16-bit samples are always encoded in little-endian
order (least significant byte first), and are byte-swapped if necessary in Flash Player before playback. Format 0 is
clearly disadvantageous because it introduces a playback platform dependency. For 16-bit samples, format 3 is
highly preferable to format 0 for SWF 4 or later.
The contents of SoundData vary depending on the value of the SoundFormat field in the SoundStreamHead tag:
•

If SoundFormat is 0 or 3, SoundData contains raw, uncompressed samples.

•

If SoundFormat is 1, SoundData contains an ADPCM sound data record.

•

If SoundFormat is 2, SoundData contains an MP3 sound data record.

•

If SoundFormat is 4, 5, or 6, SoundData contains Nellymoser data (see “Nellymoser compression” ).

•

If SoundFormat is 11, SoundData contains Speex data (see “Speex compression”).

StartSound
StartSound is a control tag that either starts (or stops) playing a sound defined by DefineSound. The SoundId
field identifies which sound is to be played. The SoundInfo field defines how the sound is played. Stop a sound
by setting the SyncStop flag in the SOUNDINFO record.
The minimum file format version is SWF 1.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 15.

SoundId

UI16

ID of sound character to play.

179

SoundInfo

SOUNDINFO

Sound style information.

StartSound2
StartSound is a control tag that either starts (or stops) playing a sound defined by DefineSound. The SoundId
field identifies which sound is to be played. The SoundInfo field defines how the sound is played. Stop a sound
by setting the SyncStop flag in the SOUNDINFO record.
The minimum file format version is SWF 9. Supported in Flash Player 9.0.45.0 and later.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 89.

SoundClassName

STRING

Name of the sound class to play.

SoundInfo

SOUNDINFO

Sound style information.

Sound styles
SOUNDINFO
The SOUNDINFO record modifies how an event sound is played. An event sound is defined with the DefineSound
tag. Sound characteristics that can be modified include:
•

Whether the sound loops (repeats) and how many times it loops.

•

Where sound playback begins and ends.

•

A sound envelope for time-based volume control.

Field

Type

Comment

Reserved

UB[2]

Always 0.

SyncStop

UB[1]

Stop the sound now.

SyncNoMultiple

UB[1]

Don’t start the sound if already playing.

HasEnvelope

UB[1]

Has envelope information.

HasLoops

UB[1]

Has loop information.

HasOutPoint

UB[1]

Has out-point information.

180

HasInPoint

UB[1]

Has in-point information.

InPoint

If HasInPoint, UI32

Number of samples to skip at beginning of sound.

OutPoint

If HasOutPoint, UI32

Position in samples of last sample to play.

LoopCount

If HasLoops, UI16

Sound loop count.

EnvPoints

If HasEnvelope, UI8

Sound Envelope point count.

EnvelopeRecords

If HasEnvelope,
Sound Envelope records.
SOUNDENVELOPE[EnvPoints]

SOUNDENVELOPE
The SOUNDENVELOPE structure is defined as follows:
Field

Type

Comment

Pos44

UI32

Position of envelope point as a number of 44 kHz samples. Multiply accordingly if
using a sampling rate less than 44 kHz.

LeftLevel

UI16

Volume level for left channel. Minimum is 0, maximum is 32768.

RightLevel

UI16

Volume level for right channel. Minimum is 0, maximum is 32768.

For mono sounds, set the LeftLevel and RightLevel fields to the same value. If the values differ, they will be
averaged.

Streaming sound
The SWF file format supports a streaming sound mode where sound data is played and downloaded in tight
synchronization with the timeline. In this mode, sound packets are stored with each frame.
When streaming sound is present, and the playback CPU is too slow to maintain the desired SWF frame rate,
Flash Player skips frames of animation in order to maintain sound synchronization and avoid dropping sound
samples. (Actions from the skipped frames are still executed.)
The main timeline of a SWF file can only have a single streaming sound playing at a time, but each sprite can
have its own streaming sound (see Sprites and Movie Clips).

SoundStreamHead
If a timeline contains streaming sound data, there must be a SoundStreamHead or SoundStreamHead2 tag
181

before the first sound data block (see “SoundStreamBlock”). The SoundStreamHead tag defines the data format
of the sound data, the recommended playback format, and the average number of samples per
SoundStreamBlock.
The minimum file format version is SWF 1.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 18.

Reserved

UB[4]

Always zero.

PlaybackSoundRate

UB[2]

Playback sampling rate: 0 = 5.5 kHz; 1 = 11 kHz; 2 =
22 kHz; 3 = 44 kHz

PlaybackSoundSize

UB[1]

Playback sample size. Always 1 (16 bit).

PlaybackSoundType

UB[1]

Number of playback channels: mono or stereo. 0 =
sndMono; 1 = sndStereo

StreamSoundCompression

UB[4]

Format of streaming sound data. 1 = ADPCM; SWF
4 and later only: 2 = MP3

StreamSoundRate

UB[2]

The sampling rate of the streaming sound data: 0 =
5.5 kHz; 1 = 11 kHz; 2 = 22 kHz; 3 = 44 kHz

StreamSoundSize

UB[1]

The sample size of the streaming sound data.
Always 1 (16 bit).

StreamSoundType

UB[1]

Number of channels in the streaming sound data.
0 = sndMono; 1 = sndStereo

StreamSoundSampleCount

UI16

Average number of samples in each
SoundStreamBlock. Not affected by mono/stereo
setting; for stereo sounds this is the number of
sample pairs.

LatencySeek

If
See “MP3 sound data”. The value here should
StreamSoundCompression= match the SeekSamples field in the first
2, SI16, Otherwise absent
SoundStreamBlock for this stream.

The PlaybackSoundRate, PlaybackSoundSize, and PlaybackSoundType fields are advisory only; Flash Player may
ignore them.

182

SoundStreamHead2
The SoundStreamHead2 tag is identical to the SoundStreamHead tag, except it allows different values for
StreamSoundCompression and StreamSoundSize (SWF 3 file format).
Field

Type

Comment

Header

RECORDHEADER

Tag type = 45

Reserved

UB[4]

Always zero.

PlaybackSoundRate

UB[2]

Playback sampling rate: 0 = 5.5 kHz; 1 = 11 kHz; 2
= 22 kHz; 3 = 44 kHz

PlaybackSoundSize

UB[1]

Playback sample size: 0 = 8-bit; 1 = 16-bit

PlaybackSoundType

UB[1]

Number of playback channels: 0 = sndMono; 1 =
sndStereo

StreamSoundCompression

UB[4]

Format of SoundData. See “Audio coding
formats” .

StreamSoundRate

UB[2]

The sampling rate of the streaming sound data:
5.5 kHz is not allowed for MP3. 0 = 5.5 kHz; 1 =
11 kHz; 2 = 22 kHz; 3 = 44 kHz

StreamSoundSize

UB[1]

Size of each sample. Always 16 bit for
compressed formats. May be 8 or 16 bit for
uncompressed formats: 0 = 8-bit; 1 = 16-bit

StreamSoundType

UB[1]

Number of channels in the streaming sound data:
0 = sndMono; 1 = sndStereo

StreamSoundSampleCount

UI16

Average number of samples in each
SoundStreamBlock. Not affected by mono/stereo
setting; for stereo sounds this is the number of
sample pairs.

LatencySeek

If
See MP3 sound data. The value here should
StreamSoundCompression= match the SeekSamples field in the first
2, SI16, Otherwise absent
SoundStreamBlock for this stream.

The PlaybackSoundRate, PlaybackSoundSize, and PlaybackSoundType fields are advisory only; Flash Player may
ignore them.
183

SoundStreamBlock
The SoundStreamBlock tag defines sound data that is interleaved with frame data so that sounds can be played
as the SWF file is streamed over a network connection. The SoundStreamBlock tag must be preceded by a
SoundStreamHead or SoundStreamHead2 tag. There may only be one SoundStreamBlock tag per SWF frame.
The minimum file format version is SWF 1.
Field

Type

Comment

Header

RECORDHEADER (long)

Tag type = 19.

StreamSoundData

UI8[size of compressed data]

Compressed sound data.

The contents of StreamSoundData vary depending on the value of the StreamSoundCompression field in the
SoundStreamHead tag:
•

If StreamSoundCompression is 0 or 3, StreamSoundData contains raw, uncompressed samples.

•

If StreamSoundCompression is 1, StreamSoundData contains an ADPCM sound data record.

•

If StreamSoundCompression is 2, StreamSoundData contains an MP3 sound data record.

•

If StreamSoundCompression is 6, StreamSoundData contains a NELLYMOSERDATA record.

MP3STREAMSOUNDDATA
Field

Type

Comment

SampleCount

UI16

Number of samples represented by this block. Not
affected by mono/stereo setting; for stereo sounds this
is the number of sample pairs.

Mp3SoundData

MP3SOUNDDATA

MP3 frames with SeekSamples values.

Frame subdivision for streaming sound
The best streaming sound playback is obtained by providing a SoundStreamBlock tag in every SWF frame, and
including the same number of sound samples in each SoundStreamBlock. The ideal number of samples per SWF
frame is easily determined: divide the sampling rate by the SWF frame rate. If this results in a non-integer
number, write an occasional SoundStreamBlock with one more or one fewer samples, so that the average
number of samples per frame remains as close as possible to the ideal number.
For uncompressed audio, it is possible to include an arbitrary number of samples in a SoundStreamBlock, so an
184

ideal number of samples can be included in each SWF frame. For MP3 sound, the situation is different. MP3 data
is itself organized into frames, and an MP3 frame contains a fixed number of samples (576 or 1152, depending
on the sampling rate). SoundStreamBlocks containing MP3 data must contain whole MP3 frames rather than
fragments, so a SoundStreamBlock with MP3 data always contains a number of samples that is a multiple of 576
or 1152.
There are two requirements for keeping MP3 streaming sound in sync with SWF playback:
•

Distribute MP3 frames appropriately among SWF frames.

•

Provide appropriate SeekSamples values in SoundStreamBlock tags.

These techniques are described in the rest of this section.
For streaming ADPCM sound, the logic for distributing ADPCM packets among SWF frames is identical to
distributing MP3 frames among SWF frames. However, for ADPCM sound, there is no concept of SeekSamples or
latency. For this and other reasons, MP3 is a preferable format for SWF 4 or later files.
To determine the ideal number of MP3 frames for each SWF frame, divide the ideal number of samples per SWF
frame by the number of samples per MP3 frame. This will usually result in a non-integer number. Achieve the
ideal average by interleaving SoundStreamBlocks with different numbers of MP3 frames. For example, at a SWF
frame rate of 12 and a sampling rate of 11 kHz, there are 576 samples per MP3 frame; the ideal number of MP3
frames per SWF frame is (11025 / 12) / 576, roughly 1.6; this can be achieved by writing SoundStreamBlocks
with one or two MP3 frames. While writing SoundStreamBlocks, keep track of the difference between the ideal
number of total samples and the total number of samples written so far. Put as many MP3 frames in each
SoundStreamBlock as is possible without exceeding the ideal number. Then, in each SoundStreamBlock, use the
difference between the ideal and actual numbers of samples as of the end of the prior SWF frame as the value
of SeekSamples. This will enable Flash Player to begin sound playback at exactly the right point after a seek
occurs. Here is an illustration of this example:

The SoundStreamBlock in SWF Frame 1 contains one MP3 frame and has SeekSamples set to zero. Frame 2
contains two MP3 frames and has SeekSamples set to 919 - 576 = 343. Frame 3 contains one MP3 frame and
has SeekSamples set to 1838 - 1728 = 110.
In continuous playback, Flash Player will string all of the MP3 frames together and play them at their natural
sample rate, reading ahead in the SWF bitstream to build up a buffer of sound data (this is why it is acceptable
to include less than the ideal number of samples in a SWF frame). After a seek to a particular frame, such as is
prompted by an ActionGotoFrame, Flash Player will skip the number of samples indicated by SeekSamples. For
185

example, after a seek to Frame 2, it will skip 343 samples of the SoundStreamBlock data from Frame 2, which
will cause sound playback to begin at sample 919, the ideal value.
If the ideal number of MP3 frames per SWF frame is less than one, there will be SWF frames whose
SoundStreamBlocks cannot accommodate any MP3 frames without exceeding the ideal number of samples. In
this case, write a SoundStreamBlock with SampleCount = 0, SeekSamples = 0, and no MP3 data.
Some MP3 encoders have an initial latency, generating a number of silent or meaningless samples before the
desired sound data begins. This can help the Flash Player MP3 decoder as well, providing some ramp-up data
before the samples that are needed. In this situation, determine how many samples the initial latency occupies,
and supply that number for SeekSamples in the first SoundStreamBlock. Flash Player will add this number to the
SeekSamples for any other frame when performing a seek. Latency also affects the decision as to how many
MP3 frames to put into a SoundStreamBlock. Here is a modification of the above example with a latency of 940
samples:

The SoundStreamBlock in SWF Frame 1 contains three MP3 frames, the maximum that can be accommodated
without exceeding the ideal number of samples after adjusting for latency (represented by the leftward shift of
the MP3 timeline above). The value of SeekSamples in Frame 1 is special; it represents the latency. Frame 2
contains one MP3 frame and has SeekSamples set to 919 - (1728 - 940) = 131.

ADPCM compression
ADPCM (Adaptive Differential Pulse Code Modulation) is a family of audio compression and decompression
algorithms. It is a simple but efficient compression scheme that avoids any licensing or patent issues that arise
with more sophisticated sound compression schemes, and helps to keep player implementations small.
For SWF 4 or later files, MP3 compression is a preferable format (see “MP3 compression” ). MP3 produces
substantially better sound for a given bitrate.
ADPCM uses a modified Differential Pulse Code Modulation (DPCM) sampling technique where the encoding of
a each sample is derived by calculating a “difference” (DPCM) value, and applying to this a complex formula
which includes the previous quantization value. The result is a compressed code, which can recreate almost the
same subjective audio quality.
A common implementation takes 16-bit linear PCM samples and converts them to 4-bit codes, yielding a
compression rate of 4:1. Public domain C code written by Jack Jansen is available at
www.cwi.nl/ftp/audio/adpcm.zip.
186

The SWF file format extends Jansen’s implementation to support 2-, 3-, 4- and 5-bit ADPCM codes. When
choosing a code size, there is the usual trade-off between file size and audio quality. The code tables used in
SWF file format are as follows (note that each structure here provides only the unique lower half of the range,
the upper half being an exact duplicate):
int indexTable2[2] = {-1, 2};
int indexTable3[4] = {-1, -1, 2, 4};
int indexTable4[8] = {-1, -1, -1, -1, 2, 4, 6, 8};
int indexTable5[16] = {-1, -1, -1, -1, -1, -1, -1, -1, 1, 2,
4, 6, 8, 10, 13, 16};

ADPCM sound data
The ADPCMSOUNDATA record defines the size of the ADPCM codes used, and an array of ADPCMPACKETs which
contain the ADPCM data.
Field

Type

Comment
Bits per ADPCM code less 2.
The actual size of each code
is AdpcmCodeSize + 2.

AdpcmCodeSize UB[2]
0 = 2 bits/sample
1 = 3 bits/sample
2 = 4 bits/sample
3 = 5 bits/sample
AdpcmPackets

If SoundType = mono, ADPCMMONOPACKET [one or more]

Array of ADPCMPACKETs.

If SoundType = stereo, ADPCMSTEREOPACKET [one or more]

ADPCMPACKETs vary in structure depending on whether the sound is mono or stereo.
ADPCMMONOPACKET
Field

Type

Comment

InitialSample

SI16

First sample. Identical to first sample in
uncompressed sound.

InitialIndex

UB[6]

Initial index into the ADPCM StepSizeTable.*

AdpcmCodeData

UB[4095 * (AdpcmCodeSize+2)]

4095 ADPCM codes. Each sample is
(AdpcmCodeSize + 2) bits.

187

ADPCMSTEREOPACKET
Field

Type

Comment

InitialSampleLeft

SI16

First sample for left channel. Identical to first
sample in uncompressed sound.

InitialIndexLeft

UB[6]

Initial index into the ADPCM StepSizeTable* for
left channel.

InitialSampleRight

SI16

First sample for right channel. Identical to first
sample in uncompressed sound.

InitialIndexRight

UB[6]

Initial index into the ADPCM StepSizeTable* for
right channel

AdpcmCodeData

UB[8190 * (AdpcmCodeSize+2)]

4095 ADPCM codes per channel, total 8190. Each
sample is (AdpcmCodeSize + 2) bits. Channel data
is interleaved left, then right.

* For an explanation of StepSizeTable, see the Jansen source code.

MP3 compression
MP3 is a sophisticated and complex audio compression algorithm supported in SWF 4 and later. It produces
superior audio quality at better compression ratios than ADPCM. Generally speaking, MP3 refers to MPEG1
Layer 3; however, the SWF file format supports later versions of MPEG (V2 and 2.5) that were designed to
support lower bitrates. Flash Player supports both CBR (constant bitrate) and VBR (variable bitrate) MP3
encoding.
For more information on MP3, see www.mp3-tech.org/ and www.iis.fhg.de/amm/techinf/layer3/index.html.
Writing an MP3 encoder is quite difficult, but public-domain MP3 encoding libraries may be available.
Note: Be aware that software and hardware MP3 encoders and decoders might have their own licensing
requirements.

MP3 sound data
MP3 sound data is described in the following table:
Field

Type

Comment

SeekSamples

SI16

Number of samples to skip.

Mp3Frames

MP3FRAME[zero or more]

Array of MP3 frames.

188

For an explanation of the the SeekSamples field, see “Frame subdivision for streaming sound”.
Note: For event sounds, SeekSamples is limited to specifying initial latency.

189

MP3 frame
The MP3FRAME record corresponds exactly to an MPEG audio frame that you would find in an MP3 music file.
The first 32 bits of the frame contain header information, followed by an array of bytes, which are the encoded
audio samples.
Field

Type

Comment

Syncword

UB[11]

Frame sync. All bits must be set.

MpegVersion

UB[2]

MPEG2.5 is an extension to MPEG2
that handles very low bitrates,
allowing the use of lower sampling
frequencies. 0 = MPEG Version 2.5;
1 = reserved; 2 = MPEG Version 2; 3
= MPEG Version 1

Layer

UB[2]

Layer is always equal to 1 for MP3
headers in SWF files. The “3” in
MP3 refers to the Layer, not the
MpegVersion.: 0 = reserved; 1 =
Layer III; 2 = Layer II; 3 = Layer I

ProtectionBit

UB[1]

If ProtectionBit == 0, a 16-bit CRC
follows the header: 0 = Protected
by CRC; 1 = Not protected

Bitrate

UB[4]

Bitrates are in thousands of bits per
second. For example, 128 means
128000 bps.
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15

SamplingRate

190

UB[2]

free
32
40
48
56
64
80
96
112
128
160
192
224
256
320
bad

free
8
16
24
32
40
48
56
64
80
96
112
128
144
160
bad

Sampling rate in Hz.
Value MPEG1 MPEG2.x
MPEG2.5

--------------------0 44100 22050 11025
1 48000 24000 12000
2 32000 16000 8000
-- -- -Padding is used to fit the bitrate
exactly. 0 = frame is not padded; 1
= frame is padded with one extra
slot

PaddingBit

UB[1]

Reserved

UB[1]

ChannelMode

UB[2]

ModeExtension

UB[2]

Copyright

UB[1]

0 = Audio is not copyrighted; 1 =
Audio is copyrighted

Original

UB[1]

0 = Copy of original media; 1 =
Original media

Emphasis

UB[2]

0 = none; 1 = 50/15 ms; 2 =
reserved; 3 = CCIT J.17

SampleData

UB[size of sample data*]

The encoded audio samples.

Dual-channel files are made of two
independent mono channels. Each
one uses exactly half the bitrate of
the file. 0 = Stereo; 1 = Joint stereo
(Stereo); 2 = Dual channel; 2 =
Single channel (Mono)

* The size of the sample data is calculated like this (using integer arithmetic):
Size = (((MpegVersion == MPEG1 ? 144
- 4

: 72) * Bitrate) / SamplingRate) + PaddingBit

For example: The size of the sample data for an MPEG1 frame with a Bitrate of 128000, a SamplingRate of
44100, and PaddingBit of 1 is:
Size = (144 * 128000) / 44100 + 1 – 4 = 414 bytes

191

Nellymoser compression
Starting with SWF 6, a compressed sound format called Nellymoser Asao is available. This is a single-channel
(mono) format optimized for low-bitrate transmission of speech. The format was developed by Nellymoser Inc.
at www.nellymoser.com.
A summary of the Nellymoser Asao encoding process is provided here. For full details of the Asao format,
contact Nellymoser.
Asao uses frequency-domain characteristics of sound for compression. Sound data is grouped into frames of 256
samples. Each frame is converted into the frequency domain and the most significant (highest-amplitude)
frequencies are identified. A number of frequency bands are selected for encoding; the rest are discarded. The
bitstream for each frame then encodes which frequency bands are in use and what their amplitudes are.

Speex compression
Starting with SWF 10, a SWF file can store audio samples that have been compressed using the free, open source
Speex voice codec (see speex.org). Speex audio is stored as format 11 in a DefineSound tag. While Speex
supports a range of sample rates, Speex audio encoded in SWF is always encoded at 16 kHz; the SoundRate field
of the DefineSound tag is disregarded. The SoundType and SoundSize fields are also ingored in the case of
Speex. Speex in SWF is always mono and always decodes to 16-bit audio samples internally.
Speex 1.2 beta 3 is compiled into the Flash Player as of version 10 (10.0.12).

192

Chapter 12: Buttons
Button characters in the SWF file format serve as interactive elements. They can react programmatically to
events that occur. The most common event to handle is a simple click from the mouse, but more complex
events can be trapped as well.

Button states
A button object can be displayed in one of three states: up, over, or down.
The up state is the default appearance of the button. The up state is displayed when the SWF file starts playing,
and whenever the mouse is outside the button. The over state is displayed when the mouse is moved inside the
button. This allows rollover or hover buttons in a SWF file. The down state is the clicked appearance of the
button. It is displayed when the mouse is clicked inside the button.
A fourth state—the hit state—defines the active area of the button. This is an invisible state and is never
displayed. It defines the area of the button that reacts to mouse clicks. This hit area is not necessarily
rectangular and need not reflect the visible shape of the button.
Each state is made up of a collection of instances of characters from the dictionary. Each such instance is defined
using a Button record, which, within a button definition, acts like a PlaceObject tag. For the up, over, and down
states, these characters are placed on the display list when the button enters that state. For the hit-area state,
these characters define the active area of the button.
The following is an example of a typical button and its four states. The button is initially blue. When the mouse is
moved over the button, it changes to a purple color. When the mouse is pressed inside the button, the shading
changes to simulate a depressed button. The fourth state—the hit area—is a simple rectangle. Anything outside
this shape is outside the button, and anything inside this shape is inside the button.

The SWF file format has no native support for radio buttons or check boxes. There is no “checked” (selected)
state, and buttons cannot “stick” down after the mouse is released. In addition, there is no way to group buttons
into mutually exclusive choices. However, both of these behaviors can be simulated by using button actions.

Button tracking
Button tracking refers to how a button behaves as it tracks the movement of the mouse. A button object can
track the mouse in one of two modes, as a push button or as a menu button.
If a push button is clicked, all mouse movement events are directed to the push button until the mouse button is
released. This is called capturing the mouse. For example, if you click a push button and drag outside the button
193

(without releasing), the button changes to the over state, and the pointer remains a pointing hand.
Menu buttons do not capture the mouse. If you click a menu button and drag outside, the button changes to the
up state, and the pointer reverts to an arrow.

Events, state transitions, and actions
A button object can perform an action whenever a state transition occurs (that is, when the button changes
from one state to another). A state transition occurs in response to some event, such as a mouse click, or mouse
entering the button. In the SWF file format, events are described as state transitions. The following table shows
possible state transitions and corresponding Flash Player events:
State Transition

Event

Description

Visual Effect

IdleToOverUp

Roll Over

Mouse enters the hit area while the
mouse button is up.

Button changes from up
to over state.

OverUpToIdle

Roll Out

Mouse leaves the hit area while the
mouse button is up.

Button changes from over
to up state.

OverUpToOverDown

Press

Mouse button is pressed while the
mouse is inside the hit area.

Button changes from over
to down state.

OverDownToOverUp

Release

Mouse button is released while the
mouse is inside the hit area.

Button changes from
down to over state.

The following transitions only apply when tracking Push buttons:
State Transition

Event

Description

Visual Effect

OutDownToOverDown

Drag Over

Mouse is dragged inside the hit area
while the mouse button is down.

Button changes from over
to down state.

OverDownToOutDown

Drag Out

Mouse is dragged outside the hit
area while the mouse button is
down.

Button changes from
down to over state.

OutDownToIdle

ReleaseOutside

Mouse button is released outside
the hit area while the mouse is
captured.

Button changes from over
to up state.

The following transitions apply only when tracking Menu buttons:
194

State Transition

Event

Description

Visual Effect

IdleToOverDown

Drag Over

Mouse is dragged inside the hit area
while the mouse button is down.

Button changes from up to
down state.

OverDownToIdle

Drag Out

Mouse is dragged outside the hit area
while the mouse button is down.

Button changes from down to
up state.

Often button actions are performed only on OverDownToOverUp (when the mouse button is released), but
DefineButton2 allows actions to be triggered by any state transition. A button object can perform any action
supported by the SWF 3 actions (see “SWF 3 actions” on page 64).

Button tags
Buttonrecord
A button record defines a character to be displayed in one or more button states. The ButtonState flags indicate
which state (or states) the character belongs to.
A one-to-one relationship does not exist between button records and button states. A single button record can
apply to more than one button state (by setting multiple ButtonState flags), and multiple button records can be
present for any button state.
Each button record also includes a transformation matrix and depth (stacking-order) information. These apply
just as in a PlaceObject tag, except that both pieces of information are relative to the button character itself.
SWF 8 and later supports the new ButtonHasBlendMode and ButtonHasFilterList fields to support blend modes
and bitmap filters on buttons. Flash Player 7 and earlier ignores these two fields.
Field

Type

Comment

ButtonReserved

UB[2]

Reserved bits; always 0

ButtonHasBlendMode

UB[1]

0 = No blend mode; 1 = Has blend mode (SWF
8 and later only)

ButtonHasFilterList

UB[1]

0 = No filter list; 1 = Has filter list (SWF 8 and
later only)

ButtonStateHitTest

UB[1]

Present in hit test state

ButtonStateDown

UB[1]

Present in down state

195

ButtonStateOver

UB[1]

Present in over state

ButtonStateUp

UB[1]

Present in up state

CharacterID

UI16

ID of character to place

PlaceDepth

UI16

Depth at which to place character

PlaceMatrix

MATRIX

Transformation matrix for character
placement

ColorTransform

If within DefineButton2,
CXFORMWITHALPHA

Character color transform

FilterList

If within DefineButton2 and
ButtonHasFilterList = 1, FILTERLIST

List of filters on this button

BlendMode

If within DefineButton2 and
ButtonHasBlendMode = 1, UI8

0 or 1 = normal; 2 = layer; 3 = multiply; 4 =
screen; 5 = lighten; 6 = darken; 7 = difference;
8 = add; 9 = subtract; 10 = invert; 11 = alpha;
12 = erase; 13 = overlay; 14 = hardlight;
Values 15 to 255 are reserved.

DefineButton
The DefineButton tag defines a button character for later use by control tags such as PlaceObject.
DefineButton includes an array of Button records that represent the four button shapes: an up character, a
mouse-over character, a down character, and a hit-area character. It is not necessary to define all four states,
but at least one button record must be present. For example, if the same button record defines both the up and
over states, only three button records are required to describe the button.
More than one button record per state is allowed. If two button records refer to the same state, both are
displayed for that state.
DefineButton also includes an array of ACTIONRECORDs, which are performed when the button is clicked and
released (see “SWF 3 actions” on page 64).
The minimum file format version is SWF 1.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 7

ButtonId

UI16

ID for this character

196

Characters

BUTTONRECORD[one or more]

Characters that make up the button

CharacterEndFlag

UI8

Must be 0

Actions

ACTIONRECORD[zero or more]

Actions to perform

ActionEndFlag

UI8

Must be 0

DefineButton2
DefineButton2 extends the capabilities of DefineButton by allowing any state transition to trigger actions.
The minimum file format version is SWF 3:
Starting with SWF 9, if the ActionScript3 field of the FileAttributes tag is 1, there must be no
BUTTONCONDACTION fields in the DefineButton2 tag. ActionOffset must be 0. This structure is not supported
because it is not permitted to mix ActionScript 1/2 and ActionScript 3.0 code within the same SWF file.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 34

ButtonId

UI16

ID for this character

ReservedFlags

UB[7]

Always 0

TrackAsMenu

UB[1]

0 = track as normal button; 1 = track as menu
button

ActionOffset

UI16

Offset in bytes from start of this field to the first
BUTTONCONDACTION, or 0 if no actions occur

Characters

BUTTONRECORD [one or more]

Characters that make up the button

CharacterEndFlag

UI8

Must be 0

Actions

BUTTONCONDACTION [zero or more]

Actions to execute at particular button events

The actions associated with DefineButton2 are specified as follows:
BUTTONCONDACTION
Field

Type

Comment

CondActionSize

UI16

Offset in bytes from start of this field to next

197

BUTTONCONDACTION, or 0 if last action
CondIdleToOverDown

UB[1]

Idle to OverDown

CondOutDownToIdle

UB[1]

OutDown to Idle

CondOutDownToOverDown

UB[1]

OutDown to OverDown

CondOverDownToOutDown

UB[1]

OverDown to OutDown

CondOverDownToOverUp

UB[1]

OverDown to OverUp

CondOverUpToOverDown

UB[1]

OverUp to OverDown

CondOverUpToIdle

UB[1]

OverUp to Idle

CondIdleToOverUp

UB[1]

Idle to OverUp

CondKeyPress

UB[7]

SWF 4 or later: key codeOtherwise: always 0
Valid key codes:

198

•

1 = left arrow

•

2 = right arrow

•

3 = home

•

4 = end

•

5 = insert

•

6 = delete

•

8 = backspace

•

13 = enter

•

14 = up arrow

•

15 = down arrow

•

16 = page up

•

17 = page down

•

18 = tab

•

19 = escape

•

32 to 126: follows ASCII

CondOverDownToIdle

UB[1]

OverDown to Idle

Actions

ACTIONRECORD [zero or more]

Actions to perform. See DoAction.

ActionEndFlag

UI8

Must be 0

For each event handler (each BUTTONCONDACTION), one or more of the Cond bit fields should be filled in. This
specifies when the event handler should be executed.
CondKeyPress specifies a particular key to trap. A CondKeyPress event handler is executed even if the button
that it applies to does not have input focus. For the 32 to 126 ASCII key codes, the key event that is trapped is
composite—it takes into account the effect of the Shift key. To trap raw key events, corresponding directly to
keys on the keyboard (including the modifier keys themselves), use clip event handlers instead.

DefineButtonCxform
DefineButtonCxform defines the color transform for each shape and text character in a button. This is not used
for DefineButton2, which includes its own CXFORM.
The minimum file format version is SWF 2.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 23

ButtonId

UI16

Button ID for this information

ButtonColorTransform

CXFORM

Character color transform

DefineButtonSound
The DefineButtonSound tag defines which sounds (if any) are played on state transitions. The minimum file
format version is SWF 2.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 17

ButtonId

UI16

The ID of the button these sounds apply to.

199

ButtonSoundChar0

UI16

Sound ID for OverUpToIdle

ButtonSoundInfo0

SOUNDINFO (if ButtonSoundChar0 is
nonzero)

Sound style for OverUpToIdle

ButtonSoundChar1

UI16

Sound ID for IdleToOverUp

ButtonSoundInfo1

SOUNDINFO (if ButtonSoundChar1 is
nonzero)

Sound style for IdleToOverUp

ButtonSoundChar2

UI16

Sound ID for OverUpToOverDown

ButtonSoundInfo2

SOUNDINFO (if ButtonSoundChar2 is
nonzero)

Sound style for OverUpToOverDown

ButtonSoundChar3

UI16

Sound ID for OverDownToOverUp

ButtonSoundInfo3

SOUNDINFO (if ButtonSoundChar3 is
nonzero)

Sound style for OverDownToOverUp

200

Chapter 13: Sprites and Movie Clips
A sprite corresponds to a movie clip in the Adobe Flash authoring application. It is a SWF file contained within
another SWF file, and supports many of the features of a regular SWF file, such as the following:
•

Most of the control tags that can be used in the main file.

•

A timeline that can stop, start, and play independently of the main file.

•

A streaming sound track that is automatically mixed with the main sound track.

A sprite object is defined with a DefineSprite tag. It consists of a character ID, a frame count, and a series of
control tags. Definition tags (such as DefineShape) are not allowed in the DefineSprite tag. All of the characters
that control tags refer to in the sprite must be defined outside the sprite, and before the DefineSprite tag.
Once defined, a sprite is displayed with a PlaceObject2 tag in the main file. The transform (specified in
PlaceObject) is concatenated with the transforms of objects placed inside the sprite. These objects behave like
children of the sprite, so when the sprite is moved, the objects inside the sprite move too. Similarly, when the
sprite is scaled or rotated, the child objects are also scaled or rotated. A sprite object stops playing automatically
when it is removed from the display list.

Sprite names
When a sprite is placed on the display list, it can be given a name with the PlaceObject2 tag. This name is used to
identify the sprite so that the main file (or other sprites) can perform actions inside the sprite. This is achieved
with the SetTarget action (see ActionSetTarget).
For example, say a sprite object is placed in the main file with the name "spinner". The main file can send this
sprite to the first frame in its timeline with the following action sequence:
1. SetTarget "spinner"
2. GotoFrame zero
3. SetTarget "" (empty string)
4. End of actions. (Action code = 0)
Note: All actions following SetTarget “spinner” apply to the spinner object until SetTarget “”, which sets the
action context back to the main file.
The SWF file format supports placing sprites within sprites, which can lead to complex hierarchies of objects. To
handle this complexity, the SWF file format uses a naming convention similar to that used by file systems to
identify sprites.
201

For example, the following outline shows four sprites defined within the main file:
MainMovie.swf
SpriteA (name: Jack)
SpriteA1 (name: Bert)
SpriteA2 (name: Ernie)
SpriteB (name: Jill)

The following SetTarget paths identify the preceding sprites:
•

/Jack targets SpriteA from the main file.

•

../ targets the main file from SpriteA.

•

/Jack/Bert targets SpriteA1 from any other sprite or the main file.

•

Bert targets SpriteA1 from SpriteA.

•

../Ernie targets SpriteA2 from SpriteA1.

•

../../Jill targets SpriteB from SpriteA1.

DefineSprite
The DefineSprite tag defines a sprite character. It consists of a character ID and a frame count, followed by a
series of control tags. The sprite is terminated with an End tag.
The length specified in the Header reflects the length of the entire DefineSprite tag, including the ControlTags
field.
Definition tags (such as DefineShape) are not allowed in the DefineSprite tag. All of the characters that control
tags refer to in the sprite must be defined in the main body of the file before the sprite is defined.
The minimum file format version is SWF 3.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 39

Sprite ID

UI16

Character ID of sprite

FrameCount

UI16

Number of frames in sprite

ControlTags

TAG[one or more]

A series of tags

202

The following tags are valid within a DefineSprite tag:
• ShowFrame

• StartSound

• PlaceObject

• FrameLabel

• PlaceObject2

• SoundStreamHead

• RemoveObject

• SoundStreamHead2

• RemoveObject2

• SoundStreamBlock

• All Actions (see Actions)

• End

203

Chapter 14: Video
Adobe Flash Player 6 and later supports video playback. Video can be provided to Flash Player in the following
ways:
•

Embed video within a SWF file by using the SWF video tags.

•

Deliver a video stream over RTMP through the Adobe Flash Media Server, which, as one option, can
obtain the video data from an FLV file format file.

•

Load an FLV file directly into Flash Player by using the NetStream.play ActionScript method. This method
is only available in Flash Player 7 and later. The SWF and FLV file formats share a common video
encoding format.

For complete information about the FLV file format, refer to the FLV File Format Specification at
www.adobe.com/go/video_file_format.

Sorenson H.263 Bitstream Format
As of SWF 6, a single video format, called Sorenson H.263, is available. This format is based on H.263, an open
video encoding standard that is maintained by the ITU. Copies of the H.263 standard can be obtained at
www.itu.int/.
All references to the H.263 standard in this document refer to the draft version of H.263, dated May 1996,
sometimes referred to as H.263v1. This is distinct from the revised version of H.263, dated February 1998,
sometimes referred to as H.263v2 or H.263+, and currently the in-force version of H.263 according to the ITU.
The Sorenson H.263 video format differs slightly from H.263. For the most part, it is a subset of H.263, with
some advanced features removed and a few additions. These changes are described in this section.
The Sorenson H.263 video format was developed by Sorenson Media (www.sorenson.com). Existing products
that can produce video for playback in Flash Player are the Adobe Flash authoring application, and Sorenson
Squeeze for Adobe Flash 8, a professional video compression application. You can license the Sorenson Spark
codec to perform video encoding for Flash Player; contact Sorenson Media for details.

Summary of differences from H.263
The following H.263 features are removed from the Sorenson H.263 video format:

204

•

GOB (group of blocks) layer

•

Split-screen indicator

•

Document camera indicator

•

Picture freeze release

•

Syntax-based arithmetic coding

•

PB frames

•

Continuous-presence multipoint

•

Overlapped block-motion compensation

The following non-H.263 features are added to the Sorenson H.263 video format:
•

Disposable frames (difference frames with no future dependencies)

•

Arbitrary picture width and height up to 65535 pixels

•

Unrestricted motion vector support is always on

•

A deblocking flag is available to suggest the use of a deblocking filter

To support these differences, the Sorenson H.263 video format uses different headers than H.263 at both the
picture layer and the Macroblock layer. The GOB layer is absent.
Two versions of the Sorenson H.263 video format are defined. In version 0, the block layer is identical to H.263.
In version 1, escape codes in transform coefficients are encoded differently than in H.263. Version 0 and version
1 have no other differences

Video packet
The video packet is the top-level structural element in a Sorenson H.263 video packet. It corresponds to the
picture layer in H.263 section 5.1. This structure is included within the VideoFrame tag in the SWF file format,
and also within the VIDEODATA structure in the FLV file format.
H263VIDEOPACKET
Field

Type

Comment

PictureStartCode

UB[17]

Similar to H.263 5.1.1; 0000 0000 0000
0000 1

Version

UB[5]

Video format version; Flash Player 6
supports 0 and 1

TemporalReference

UB[8]

See H.263 5.1.2

PictureSize

UB[3]

000: custom, 1 byte; 001: custom, 2
bytes; 010: CIF (352x288); 011: QCIF

205

(176x144); 100: SQCIF (128x96); 101:
320x240; 110: 160x120; 111: reserved
CustomWidth

If PictureSize = 000, UB[8] If PictureSize =
001, UB[16] Otherwise absent. Note:
UB[16] is not the same as UI16; there is no
byte swapping.

Width in pixels

CustomHeight

If PictureSize = 000, UB[8] If PictureSize =
001, UB[16] Otherwise absent. Note:
UB[16] is not the same as UI16; there is no
byte swapping.

Height in pixels

PictureType

UB[2]

00: intra frame; 01: inter frame; 10:
disposable inter frame; 11: reserved

DeblockingFlag

UB[1]

Requests use of deblocking filter
(advisory only, Flash Player may ignore)

Quantizer

UB[5]

See H.263 5.1.4

ExtraInformationFlag

UB[1]

See H.263 5.1.9

ExtraInformation

If ExtraInformationFlag = 1, UB[8],
Otherwise absent

See H.263 5.1.10
The ExtraInformationFlagExtraInformation sequence repeats until
an ExtraInformationFlag of 0 is
encountered

...

Macroblock

MACROBLOCK

See following

PictureStuffing

varies

See H.263 5.1.13

Macro block
The macro block is the next layer down in the video structure. It corresponds to the macro block layer in H.263
section 5.3.
MACROBLOCK
Field

Type

Comment

CodedMacroblockFlag

UB[1]

See H.263 5.3.1. If 1, macro block ends here

206

MacroblockType

varies

See H.263 5.3.2. Can cause various fields (see following)
to be absent

BlockPattern

varies

See H.263 5.3.5

QuantizerInformation

UB[2]

See H.263 5.3.6; 00: -1; 01: -2; 10: +1; 11: +2

MotionVectorData

varies[2]

See H.263 5.3.7. A horizontal code followed by a vertical
code

ExtraMotionVectorData

varies[6]

See H.263 5.3.8. Three more MotionVectorData code
pairs are included when MacroblockType is INTER4V

BlockData

BLOCKDATA[6]

See H.263 5.4. Four luminance blocks followed by two
chrominance blocks

Block data
Block data is the lowest layer in the video structure. In version 0 of the Sorenson H.263 video format, this layer
follows H.263 section 5.4 exactly.
In version 1 of the Sorenson H.263 video format, escape codes in transform coefficients (see H.263 section 5.4.2)
are encoded differently. When the ESCAPE code 0000 011 appears, the next bit is a format bit that indicates the
subsequent bit layout for LAST, RUN, and LEVEL. In both cases, one bit is used for LAST and six bits are used for
RUN. If the format bit is 0, seven bits are used for LEVEL; if the format bit is 1, eleven bits are used for LEVEL. The
7-bit and 11-bit LEVEL tables, which replace table 14 in H.263, as the following table shows:
7-bit LEVELs

11-bit LEVELs

Index

Level

Code

Index

Level

Code

-

-64

FORBIDDEN

-

-1024

FORBIDDEN

0

-63

1000 001

0

-1023

1000 0000
001

.

.

.

.

.

.

61

-2

1111 110

1021

-2

1111 1111 110

62

-1

1111 111

1022

-1

1111 1111 111

-

0

FORBIDDEN

-

0

FORBIDDEN

63

1

0000 001

1023

1

0000 0000

207

001
64

2

0000 010

1024

2

0000 0000
010

.

.

.

.

.

.

125

63

0111 111

2045

1023

0111 1111 111

Screen Video bitstream format
As of SWF 7, an additional video format, called screen video, is available. Screen video is a simple lossless
sequential-bitmap format with blocked interframing. It is designed for sending captures of computer screens in
action.
Pixel data in the screen video format is compressed by using the ZLIB open standard.

Block format
Each frame in a screen video sequence is formatted as a series of blocks. These blocks form a grid over the
image. In a keyframe, information for every block is sent. In an interframe, there might be blocks that are
unchanged from the previous frame and special information can be sent to indicate this.
Blocks have width and height that range from 16 to 256 in multiples of 16. Block height is not required to match
block width. The block size must not change except at a keyframe.
Blocks are ordered from the bottom left of the image to the top right, in rows. A fixed layout of blocks exists for
any given combination of block size and image size. To determine the number of blocks in a row of the grid,
divide the image width by the block width. If the result is not an integer, the end of each row has one partial
block, which contains only the number of pixels necessary to fill the remaining width of the image. The same
logic applies to the image height, block height, number of rows in the grid, and partial blocks in the final row. It
is important to understand the partial-block algorithm to create correct blocks, since the pixels within a partial
block are extracted with implicit knowledge of the width and height of the block.

208

The following is an example of blocking. The image in this example is 120 x 80 pixels, and the block size is 32 x
32.

Video packet
The video packet is the top-level structural element in a screen video packet. This structure is included within
the VideoFrame tag in the SWF file format, and also within the VIDEODATA structure in the FLV file format.
The data consists of information about the image sub-block dimensions and grid size, followed by the data for
each block.
SCREENVIDEOPACKET
Field

Type

Comment

BlockWidth

UB[4]

Pixel width of each block in the grid. This value is stored
as (actualWidth / 16) - 1, so possible block sizes are a
multiple of 16 and not more than 256.

ImageWidth

UB[12]

Pixel width of the full image.

BlockHeight

UB[4]

Pixel height of each block in the grid. This value is stored
as (actualHeight / 16) - 1, so possible block sizes are a
multiple of 16 and not more than 256.

ImageHeight

UB[12]

Pixel height of the full image.

ImageBlocks

IMAGEBLOCK[n]

Blocks of image data. See preceding for details of how to
calculate n. Blocks are ordered from bottom left to top
right, in rows.

209

Image block
The image block represents one block in a frame.
IMAGEBLOCK
Field

Type

Comment

DataSize

UB[16]
Note: UB[16] is not the same as
UI16; no byte swapping occurs

Size of the compressed block data
that follows. If this is an interframe,
and this block is not changed since
the last keyframe, DataSize is 0 and
the Data field is absent.

Data

If DataSize > 0, UI8[DataSize]

Pixel data compressed using ZLIB.
Pixels are ordered from bottom left
to top right in rows. Each pixel is
three bytes: B, G, R.

Screen Video V2 bitstream format
SWF also supports a new screen video format, Screen Video Version 2, which is an extension of the Screen Video
bitstream format and is supported in Flash Player 8 and later. Screen Video v2 uses several techniques to reduce
the amount of data for each screen block.
In the initial Screen Video version, each block of screen data is a complete buffer of compressed data that can be
decompressed to a full 24-bit color image for that block. In the Screen Video v2 format, the screen data blocks
can be incomplete updates of the image area, similar to the concept of keyframes and interframes. Further, the
v2 format introduces a hybrid 15/7-bit hybrid colorspace in addition to the usual 24-bit RGB colorspace. The
15/7- bit hybrid colorspace is useful for encoding images with a small number of unique colors (less than 256).In
the Screen Video v2 format, block data comes in two types:
Keyblock contains complete information for the block. The contents can be decompressed to obtain the
complete block image.
Interblock requires additional data, either from a previous image or the current image, to construct the full block
image.

V2 Colorspace
The Screen Video v2 can encode video data using either a 24-bit RGB colorspace, as in v1, or using a 15/7-bit
hybrid colorspace. Using the latter colorspace, an image has a corresponding 128-entry palette. Each pixel in a
decoded image is represented by either 1 or 2 bytes. Generally, a decoder will want to convert a 15/7-bit hybrid
colorspace to a 24-bit RGB colorspace. The process for doing so is:
210

•

Fetch next byte from decoded image

•

If next byte has its high bit set, clear high bit and fetch next byte from decoded image; form a 15-bit
color by placing the low 7 bits of the current byte in bits 14-8 of the color, and place the 8 bits from the
next byte in bits 7-0 of the color; convert the 15-bit RGB color to 24-bit RGB

•

If next byte has its high bit clear, use the lower 7 bits as an index into the 128-entry palette and retrieve
the corresponding 24-bit RGB color

A v2 video packet is free to define a new palette at any time, which is transmitted as a v1 IMAGEBLOCK. In the
absence of a stream-defined palette, the v2 decoder will fall back to a default palette. For the default palette,
see Appendix C, “Screen Video v2 Palette.”

V2 Video Packet
Video Packet v2 is the top-level structural element in a screen video packet for Screen Video Version 2. This
structure is included within the VideoFrame tag in SWF file format, and also within the VIDEODATA structure in
FLV file format.
The data consists of information about the image sub-block dimensions and grid size, followed by the data for
each block.
SCREENV2VIDEOPACKET
Field

Type

Comment

BlockWidth

UB[4]

Pixel width of each block in the grid. This value is stored
as (actualWidth / 16) - 1, so possible block sizes are a
multiple of 16 and not more than 256.

ImageWidth

UB[12]

Pixel width of the full image.

BlockHeight

UB[4]

Pixel height of each block in the grid. This value is stored
as (actualHeight / 16) - 1, so possible block sizes are a
multiple of 16 and not more than 256.

ImageHeight

UB[12]

Pixel height of the full image.

Reserved

UB[6]

Must be 0

HasIFrameImage

UB[1]

If 1, has IFrameImage

HasPaletteInfo

UB[1]

If 1, has PaletteInfo

PaletteInfo

If HasPaletteInfo,
IMAGEBLOCK

One block of data to describe the palette.

211

ImageBlocks

IMAGEBLOCKV2[n]

Blocks of image data. See Block format for details of how
to calculate n. Blocks are ordered from bottom left to top
right in rows and can be a combination of keyblocks and
interblocks.

IFrameImage

If HasIFrameImage,
IMAGEBLOCKV2[n]

Blocks of image data representing interblocks that must
be combined with the previous keyblocks to produce the
image. See Block format for details of how to calculate n.
Blocks are ordered from bottom left to top right in rows.

Image Block V2
The Image Block v2 structure represents one block in a frame.
IMAGEBLOCKV2
Field

Type

Comment

DataSize

UB[16] Note: UB[16] is not
the same as UI16; there is
no byte swapping.

Size of the compressed block data that follows,
including the ImageFormat, ImageBlockHeader and
Data fields. If this is an interframe, and this block has
not changed since the last keyframe, DataSize is 0 and
the Data field is absent.

Format

IMAGEFORMAT

Compression format of block data.

ImageBlockHeader

If Format’s HasDiffBlock = Describes the format and compression of Data
1, IMAGEDIFFPOSITION. If
Format’s
ZlibPrimeCompressCurrent
= 1,
IMAGEPRIMEPOSITION

Data

If DataSize > 0,
UI8[DataSize]

Pixel data compressed using ZLIB. Pixels are ordered
from bottom left to top right in rows. Each pixel is
three bytes: B, G, R.

Image format
The IMAGEFORMAT byte describes the color depth and compression of the IMAGEBLOCKV2 structure.
Field

212

Type

Comment

Reserved

UB[3]

Must be 0

ColorDepth

UB[2]

00: 24-bit RGB image. 10: 15/7-bit hybrid color image

HasDiffBlocks

UB[1]

If 1, the data starts and stops on specific rows within the
block and does not represent the entire block.

ZlibPrimeCompressCurrent

UB[1]

If 1, the current data block was produced with the ZLIB
priming technique of compression.

ZlibPrimeCompressPrevious

UB[1]

If 1, the previous data block was produced with the ZLIB
priming technique of compression.

Image block diff position
The image block diff position can be included in the IMAGEBLOCKV2 ImageBlockHeader field. This structure
describes the location and size of the diff block image data.
IMAGEDIFFPOSITION
Field

Type

Comment

RowStart

UI8

Indicates the first scan line of the block that contains the image data.

Height

UI8

Indicates the height, in contiguous scan lines, of the image data.

Image block prime position
The image block prime position is included in the IMAGEBLOCKV2 ImageBlockHeader field if the IMAGEFORMAT
structure indicates ZLIB priming is used. This structure specifies which image is used as the priming source.
IMAGEPRIMEPOSITION
Field

Type

Comment

Block column

Ui8

Indicates the position of the source block.

Block row

Ui8

Indicates the position of the source block.

On2 Truemotion VP6 bitstream format
SWF 6 or later supports the On2 Truemotion VP6 video format, which can be played in Flash Player 8 and later.
VP6 is a leading-edge video compression algorithm that combines traditional motion compensated prediction
and pseudo discrete cosine transform (DCT) coding and context-dependent entropy-coding techniques (based
213

on Huffman and arithmetic principles) along with novel approaches to surpass the quality of other codecs. VP6
applies extensive context modeling, in and out of loop filtering, and novel quantization methods to achieve a
high level of quality. Further information on this video format can be obtained from On2: www.on2.com.
Like the Sorenson H.263 video format, the On2 Truemotion VP6 video format uses color information encoded in
the YCbCr color space described in the ITU-R BT.601 standard. This color information is stored as YUV 4:2:2 using
unsigned 8-bit values for each color component. The following algorithm can be used to convert forward from
RGB color space pixel data to YUV color space:
FUNCTION SATURATE(A) = MIN(255,MAX(0,A))
Y = SATURATE(+ (0.257 * R) + (0.504 * G) + (0.098 * B)
U = SATURATE(- (0.148 * R) - (0.291 * G) + (0.439 * B)
V = SATURATE(+ (0.439 * R) - (0.368 * G) - (0.071 * B)

+ 16)
+ 128)
+ 128)

This algorithm can be used to convert YUV color space pixel data back to RGB color space:
B = SATURATE(1.164(Y - 16) + 2.018(U - 128))
G = SATURATE(1.164(Y - 16) - 0.813(V - 128) - 0.391(U - 128))
R = SATURATE(1.164(Y - 16) + 1.596(V - 128))

In addition to supporting standard On2 Truemotion VP6 video streams, Flash Player 8 adds support for an extra
alpha channel that is used to simulate transparency. The alpha channel is encoded by using a second On2
Truemotion VP6 stream that contains the alpha channel information. To encode this type of video stream, the
preceding RGB-to-YUV algorithm should be used on the RGB color channels of the premultiplied ARGB color
space pixel data. With the resulting YUV color space data, the video data can be encoded as the first video
stream.
For the second video stream, the following algorithm can be used to obtain the YUV video data from the alpha
channel of the premultiplied ARGB color space pixel data:
Y = A U = 0
V = 0

Note: At encode time, the second video stream must contain at least as many key frames as the first video
stream. Each key frame occurring in the first video stream must force a key frame in the second video stream at
encode time so that the combined video stream stays seekable.
To decode alpha channel video streams, assume that the first video stream returns YUV- encoded color channels
for a screen pixel in the form of three channels named Y1, U1, and V1. The second video stream returns this data
as Y2, U2, and V2. The resulting alpha premultiplied ARGB pixel values are obtained by using the following
algorithm:
B = MIN(Y2,SATURATE(1.164(Y1 - 16) + 2.018(U1 - 128)))
G = MIN(Y2,SATURATE(1.164(Y1 - 16) - 0.813(V1 - 128) - 0.391(U1 - 128)))
R = MIN(Y2,SATURATE(1.164(Y1 - 16) + 1.596(V1 - 128)))
A = Y2

The U2 and V2 channels are not currently used at decode time.

214

VP6 FLV video packet
The VP6 FLV video packet represents a VP6 video frame within an FLV file.
VP6FLVVIDEOPACKET
Field

Type

Comment

HorizontalAdjustment

UB[4]

Number of pixels to subtract from the total width. The resulting width is
used on the stage, and the rightmost pixels of the video is cropped.

VerticalAdjustment

UB[4]

Number of pixels to subtract from the total height. The resulting height
is used on the stage, and the rightmost pixels of the video is cropped.

Data

UI8[n]

Raw VP6 video stream data

VP6 FLV Alpha video packet
The VP6 FLV Alpha video packet represents a VP6 video frame with alpha channel information within FLV files.
VP6FLVALPHAVIDEOPACKET
Field

Type

Comment

HorizontalAdjustment

UB[4]

Number of pixels to subtract from the total width.
The resulting width is used on the stage, and the
rightmost pixels of the video is cropped.

VerticalAdjustment

UB[4]

Number of pixels to subtract from the from the total
height. The resulting height is used on the stage, and
the rightmost pixels of the video is cropped.

OffsetToAlpha

UI24

Offset in bytes to the alpha channel video data

Data

UI8[OffsetToAlpha]

Raw VP6 video stream data representing the color
channels

AlphaData

UI8[n]

Raw VP6 video stream data representing the alpha
channel

215

VP6 SWF video packet
The VP6 SWF video packet represents a VP6 video frame within SWF files.
VP6SWFVIDEOPACKET
Field

Type

Comment

Data

UI8[n]

Raw VP6 video stream data

VP6 SWF Alpha video packet
The VP6 SWF Alpha video packet represents a VP6 video frame with alpha channel information within SWF files.
VP6SWFALPHAVIDEOPACKET
Field

Type

Comment

OffsetToAlpha

UI24

Offset in bytes to the alpha
channel video data

Data

UI8[OffsetToAlpha]

Raw VP6 video stream data
representing the color channels

AlphaData

UI8[n]

Raw VP6 video stream data
representing the alpha channel

SWF video tags
The following tags define embedded video data within a SWF file. These tags are permissible only in SWF 6 or
later.
Video embedded in a SWF file is always streamed: video frames are located in the SWF frames with which they
are temporally associated, and video playback can begin before an entire video stream is downloaded. This
process is comparable to the way that streaming sounds are defined (see Streaming sound).

216

DefineVideoStream
DefineVideoStream defines a video character that can later be placed on the display list (see The Display List).
DefineVideoStream
Field

Type

Comment

Header

RECORDHEADER

Tag type = 60

CharacterID

UI16

ID for this video character

NumFrames

UI16

Number of VideoFrame tags that makes up this stream

Width

UI16

Width in pixels

Height

UI16

Height in pixels

VideoFlagsReserved

UB[4]

Must be 0

VideoFlagsDeblocking

UB[3]

000 = use VIDEOPACKET value: ; 001 = off; 010 = Level 1 (Fast
deblocking filter); 011 = Level 2 (VP6 only, better deblocking
filter); 100 = Level 3 (VP6 only, better deblocking plus fast
deringing filter); 101 = Level 4 (VP6 only, better deblocking
plus better deringing filter); 110 = Reserved; 111 = Reserved

VideoFlagsSmoothing

UB[1]

0 = smoothing off (faster); 1 = smoothing on (higher quality)

CodecID

UI8

2 = Sorenson H.263; 3 = Screen video (SWF 7 and later only);
4 = VP6 (SWF 8 and later only); 5 = VP6 video with alpha
channel (SWF 8 and later only)

VideoFrame
VideoFrame provides a single frame of video data for a video character that is already defined with
DefineVideoStream.
In playback, the time sequencing of video frames depends on the SWF frame rate only. When SWF playback
reaches a particular SWF frame, the video images from any VideoFrame tags in that SWF frame are rendered.
Any timing mechanisms built into the video payload are ignored.
A VideoFrame tag is not needed for every video character in every frame number specified. A VideoFrame tag
merely sets video data associated with a particular frame number; it does not automatically display a video
frame. To display a video frame, specify the frame number as the Ratio field in PlaceObject2 or PlaceObject3.

217

VideoFrame
Field

Type

Comment

Header

RECORDHEADER

Tag type = 61

StreamID

UI16

ID of video stream character of which this
frame is a part

FrameNum

UI16

Sequence number of this frame within its
video stream

VideoData

if CodecID = 2 H263VIDEOPACKET
if CodecID = 3 SCREENVIDEOPACKET
if CodecID = 4 VP6SWFVIDEOPACKET
if CodecID = 5 VP6SWFALPHAVIDEOPACKET
if CodecID = 6 SCREENV2VIDEOPACKET

Video frame payload

218

Chapter 15: Metadata
Starting with version 9, the SWF file format specification supports the inclusion of arbitrary blobs of binary data.

FileAttributes
If the FileAttributes tag is used, it must immediately follow the SWF header. It defines a set of bit fields that
configure the Flash Player. As such, it is only interpreted on the root SWF.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 69

Reserved

UB[3]

Must be 0

HasMetaData

UB[1]

If set, the SWF contains metadata somewhere. This
is ignored by the player, but used by search
engines.

SWFFlagsAS3

UB[1]

Set if the SWF contains AVM-2 (Tamarin) byte code.

SWFFlagsNoCrossDomainCache

UB[1]

If set, this SWF will never be placed in a cross
domain cache.

Reserved

UB[1]

Must be 0

SWFFlagsUseNetwork

UB[1]

If this is a SWF running from the local drive, give it
network access instead of of local access.

219

EnableTelemetry
Telemetry is a a Flash player feature that sends profiling information about the runtime and the currently
running content. The EnableTelemetry tag controls whether the advanced features of telemetry are included in
the profile data. If the tag isn’t present, only basic information is available.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 93

Reserved

UB[16]

Must be 0

PasswordHash

UI8[32]

Optional: SHA-256 hash of the UTF-8
representation of the password. If not present, SWF
opts in to advanced telemetry.

DefineBinaryData
The DefineBinaryData tag permits arbitrary binary data to be embedded in a SWF file. DefineBinaryData is a
definition tag, like DefineShape and DefineSprite. It associates a blob of binary data with a standard SWF 16-bit
character ID. The character ID is entered into the SWF file's character dictionary.
DefineBinaryData is intended to be used in conjunction with the SymbolClass tag. The SymbolClass tag can be
used to associate a DefineBinaryData tag with an AS3 class definition. The AS3 class must be a subclass of
ByteArray. When the class is instantiated, it will be populated automatically with the contents of the binary data
resource.
Field

Type

Comment

Header

RECORDHEADER

Tag type = 87

Tag

UI16

16-bit character ID

Reserved

U32

Reserved space; must be 0

Data

BINARY

A blob of binary data, up to the end of the tag

220

Appendix A: SWF Uncovered: A Simple SWF
File Dissected
To write SWF files, you must be able to read and understand the raw bits and bytes. This appendix examines a
simple, one-frame SWF file that contains only a rectangle.
Here is a hex dump of the SWF file:
000000
000010
000020
000030
000040

46
00
01
00
B4

57
00
00
01
00

53
0C
70
25
00

03
01
FB
C9
86

4F
00
49
92
06

00
43
97
0D
06

00
02
0D
21
01

00
FF
0C
ED
00

78
FF
7D
48
01

00
FF
50
87
00

05
BF
00
65
00

5F
00
01
30
40

00
23
14
3B
00

00
00
00
6D
00

0F
00
00
E1
00

A0
00
00
D8

A SWF file always begins with a header. It describes the file version, the length of the file in bytes, the frame size
in twips (twentieths of a pixel), frame rate in frames per second, and the frame count.
The types are defined in Chapter 1: Basic Data Types, on page 14.
SWF File Header
Field

Type

Comment

Signature

UI8

Signature byte: “F” indicated uncompressed; “C” indicates
compressed (SWF 6 or later only) using ZLib compression. “Z”
indicates compressed using LZMA compression.

Signature

UI8

Signature byte always “W”

Signature

UI8

Signature byte always “S”

Version

UI8

Single byte file version (for example, 0x06 for SWF 6)

FileLength

UI32

Length of entire file in bytes

FrameSize

RECT

Frame size in twips

FrameRate

UI16

Frame delay in 8.8 fixed number of frames per second

FrameCount

UI16

Total number of frames in file

The first three bytes are the standard signature for all SWF files. They are the ASCII values of the characters ‘F’
(or ‘C’), ‘W’, and ‘S’ in that order. The fourth byte indicates the version of the file.

221

0x46 ➜ ‘F’ 0x57 ➜ ‘W’ 0x53 ➜ ‘S’ 0x03 ➜ 3

The next four bytes represent an unsigned 32-bit integer indicating the file size. Here’s where it starts getting
tricky and machine architecture gets involved. The next four bytes are 0x4F000000 so that would imply that the
file length is 1325400064 bytes, a very large number which doesn’t make sense. What we failed to do is swap all
the bytes.
In SWF files, bytes are swapped whenever reading words and dwords such that a 32-bit value B1B2B3B4 is
written as B4B3B2B1, and a 16-bit value B1B2 is written as B2B1. Single bytes are written unchanged since there
is no bit-swapping. The reason for this is the differences in storage and retrieval between the Mac and PC
processors.
Reversing the bytes we can read the four bytes correctly and see that file is 79 bytes long.
0x4F000000 ➜

0x0000004F ➜

79

The next nine bytes represent a data structure used in the SWF format called a Rectangle. Here is the file
description of a rectangle:
RECT
Field

Type

Comment

Nbits

UB[5]

Bits in each rect value field

Xmin

SB[Nbits]

x minimum position for rect

Xmax

SB[Nbits]

x maximum position for rect

Ymin

SB[Nbits]

y minimum position for rect

Ymax

SB[Nbits]

y maximum position for rect

To understand these bytes, we need to look at the individual bits.
78

00

05

0111 1000
0000 0000

5F 00

00

0F A0 00

0000 0000
0000 1111

0000 0101
1010 0000

0101 1111
0000 0000

0000 0000

There are five fields in a rectangle structure: Nbits, Xmin, Xmax, Ymin, Ymax. The unsigned Nbits field occupies
the first five bits of the rectangle and indicates how long the next four signed fields are.
Another subtle point about the SWF file representation is that reading and writing bits is different from reading
and writing words and dwords. When reading and writing bits, no byte-swapping occurs. This is because when
Flash Player is reading an n-bit field, it reads a byte at a time until it has read all n bits. Therefore, the next five
222

bits are read in order and evaluate to 15.
01111 ➜ 15

What if Nbit has a value of sixteen? This is exactly the size of a word so do we read the following fields as words
and swap bytes? No. Fields described by bit size are always read a byte at a time. No swapping, just read the
next n bits in that order.
000000000000000

‹

0 = Xmin

010101011111000

‹

11000 = Xmax

000000000000000

‹

0 = Ymin

001111101000000

‹

8000 = Ymax

For the header, the rectangle is used to store the file dimensions with Xmax corresponding to the file width and
Ymax corresponding to the file height, both in twips. In SWF format, a twip is a twentieth of a pixel, so if we
convert to pixels, we see that our file is 550 x 400.
Now we have looked at all of the fields of the rectangle and evaluated them, but what about those last seven
bits which are all 0s? Those bits are filled with 0s so that the structure aligns to a byte boundary.
0000000 = padding bits
After the end of any structure, if the structure does not completely fill up its last byte, then that last byte is filled
with 0s to keep the next item byte aligned. So if the next item is a word or dword, you can read it as such and
not worry about being in the middle of a byte. In this case, only one bit in the last byte is used so the last seven
bits are filled with 0s.
Next in the header is the frame rate. It is supposed to be stored as a 16-bit integer, but the first byte (or last
depending on how you look at it) is completely ignored. So the frame rate is
12 fps.
0x000C ➜

0x0C00 ➜

0x0C ➜ 12 = frame rate

Next is the frame count, which is also a 16-bit integer. So the frame count is 1.
0x0100 ➜

223

0x0001(byte swapping) ➜

1 = frame count

Now we are finished with the header. After the header is a series of tagged data blocks. Here is a description of
a tag (this is simplifying somewhat; byte swapping is necessary):
RECORDHEADER (short)
Field

Type

Comment

TagCodeAndLength

UI16

Upper 10 bits: tag type; Lower 6 bits: tag length

RECORDHEADER (long)
Field

Type

Comment

TagCodeAndLength

UI16

Tag type and length of 0x3F Packed together as in short header

Length

UI32

Length of tag

There are two types of tags: the short and long tag header. Regardless of which case you have, you begin by
looking at the first word.
0x4302 ➜

0x0243 ➜

0000 0010 0100 0011

The first 10 bits of the tag are the unsigned tag type. The tag type indicates what type of data is to follow in the
body of the data block to follow. In this case, the value of the tag type is 9, which corresponds to a
SetBackgroundColor block. The last six unsigned bits of the tag header indicate the length of the data block to
follow if it is 62 bytes or less. If the length of the data block is more than 62 bytes, then this field has all 1s and
the length is indicated in the following dword. For the tag we are looking at, the field does not have all 1s, so it
does indicate the actual length which is 3 bytes.
0000001001 = 9 = SetBackgroundColor000011 = 3 = body length
Since we know that the length of the body is 3 bytes, let’s take a look at it. A SetBackgroundColor tag only
contains the 3-byte RGB color description so we evaluate it as such. A color is its own 3-byte data type so there
is no byte swapping.
0XFFFFFF = white
The next tag is a long tag and is a DefineShape tag.
0xBF00

0x00BF

0000000010 = 2 = DefineShape
224

0000 0000 1011 1111

111111 = 63 = body length (indicates that the body length value is in the next dword)
0x23000000

0x00000023

35 = body length

Here is the file description of DefineShape:
DefineShape
Field

Type

Comment

Header

RECORDHEADER

Tag type = 2

ShapeId

UI16

ID for this character

ShapeBounds

RECT

Bounds of the shape

Shapes

SHAPEWITHSTYLE

Shape information

The body of a DefineShape is composed of an unsigned 16-bit character ID, a rectangle defining the bounds for
the shape, and a SHAPEWITHSTYLE structure which contains shape information.
0x0100 ➜

0x0001 ➜

1 = shape ID

Now the Rect, which defines the boundaries:
70 FB 49 97 0D 0C 7D 50
0111 0000 1111 1011 0100 1001 1001 0111 0000 1101

0000

1100 0111 1101 0101 0000
01110 = 14 = Nbits
00011111011010 = 2010 = Xmin

/20 to covert to pixels from twips

100.5

01001100101110 = 4910 = Xmax

245.5

00011010000110 = 1670 = Ymin

83.5

00111110101010 = 4010 = Ymax

200.5

000 = fill bits

The following table describes the SHAPEWITHSTYLE structure:

225

SHAPEWITHSTYLE
Field

Type

Comment

FillStyles

FILLSTYLEARRAY

Array of fill styles

LineStyles

LINESTYLEARRAY

Array of line styles

NumFillBits

UB[4]

Number of fill index bits

NumLineBits

UB[4]

Number of line index bits

ShapeRecords

SHAPERECORD[one or more]

Shape records (see following)

A fill style array itself has three fields. The first field is an 8-bit integer count which indicates how many fill styles
are in the array. This count works similar to the tag’s length field in that if it is all 1s, you have to look at the next
16 bits to get the actual length. Here is the file description:
FILLSTYLEARRAY
Field

Type

Comment

FillStyleCount

UI8

Count of fill styles

FillStyleCountExtended

If FillStyleCount = 0xFF UI16 Extended count of fill
styles.
Supported only for Shape2 and Shape3.

FillStyles

FILLSTYLE[FillStyleCount]

Array of fill styles

In this case, the 8-bit count is equal to 0 so there is nothing to follow it.
0x00 = 0 = FillStyleCount ➜ This is the end of the fill style array because it has no elements
A line style array is exactly the same as a fill style array except it stores line styles. Here is the file description:
LINESTYLEARRAY
Field

Type

Comment

LineStyleCount

UI8

Count of line styles

LineStyleCountExtended

If LineStyleCount = 0xFF UI16

Extended count of line styles

226

LineStyles

LINESTYLE[count]

Array of line styles

0x01 = 1 = LineStyleCount ➜ So there is one line style in the array.
A line style has two parts, an unsigned 16-bit integer indicating the width of a line in twips, and a color. Here is
the file description:
LINESTYLE
Field

Type

Comment

Width

UI16

Width of line in twips

Color

RGB (Shape1 or Shape2) RGBA (Shape3)

Color value including alpha channel information
for Shape3

The color in this case is a 24-bit RGB, but if we were doing a DefineShape3, it would be a 32- bit RGBA where
alpha is the opacity of the color.
0x1400 ➜ 0x0014 = 20 = width = 1 pixel
0x000000 = RGB = black
Back to the ShapeWithStyle, the NumFillBits field is 4 bits, as is the NumLineBits.
0x0 = 0 = NumFillBits

0x1 = 1 = NumLineBits

Now for the array of shape records. The following four tables describe the four types of shape records. Here are
the file descriptions:
ENDSHAPERECORD
Field

Type

Comment

TypeFlag

UB[1]

Non-edge record flag. Always 0

EndOfShape

UB[5]

End of shape flag. Always 0

STYLECHANGERECORD
Field

227

Type

Comment

TypeFlag

UB[1]

Non-edge record flag, Always 0

StateNewStyles

UB[1]

New styles flag. Used by DefineShape2 and
DefineShape3 only.

StateLineStyle

UB[1]

Line style change flag

StateFillStyle1

UB[1]

Fill style 1 change flag

StateFillStyle0

UB[1]

Fill style 0 change flag

StateMoveTo

UB[1]

Move to flag

MoveBits

If StateMoveTo, UB[5]

Move bit count

MoveDeltaX

If StateMoveTo,
SB[MoveBits]

Delta X value

MoveDeltaY

If StateMoveTo,
SB[MoveBits]

Delta Y value

FillStyle0

If StateFillStyle0,
UB[FillBits]

Fill 0 Style

FillStyle1

If StateFillStyle1,
UB[FillBits]

Fill 1 Style

LineStyle

If StateLineStyle,
UB[LineBits]

Line Style

FillStyles

If StateNewStyles,
FILLSTYLEARRAY

Array of new fill styles

LineStyles

If StateNewStyles,
LINESTYLEARRAY

Array of new line styles

NumFillBits

If StateNewStyles, UB[4]

Number of fill index bits for new styles

NumLineBits

If StateNewStyles, UB[4]

Number of line index bits for new styles

STRAIGHTEDGERECORD
Field

Type

Comment

TypeFlag

UB[1]

This is an edge record. Always 1.

228

StraightFlag

UB[1]

Straight edge. Always 1.

NumBits

UB[4]

Number of bits per value (two less than the
actual number).

GeneralLineFlag

UB[1]

General Line equals 1. Vert/Horz Line equals
0.

DeltaX

If GeneralLineFlag, SB[NumBits+2]

X delta

DeltaY

If GeneralLineFlag, SB[NumBits+2]

Y delta

VertLineFlag

If GeneralLineFlag, SB[1]

Vertical Line equals 1. Horizontal Line equals
0.

DeltaX

If VertLineFlag, SB[NumBits+2]

X delta

DeltaY

If VertLineFlag, SB[NumBits+2]

Y delta

CURVEDEDGERECORD
Field

Type

Comment

TypeFlag

UB[1]

This is an edge record. Always 1.

StraightFlag

UB[1]

Curved edge. Always 0.

NumBits

UB[4]

Number of bits per value. (two less than the actual number)

ControlDeltaX

SB[NumBits+2]

X control point change

ControlDeltaY

SB[NumBits+2]

Y control point change

AnchorDeltaX

SB[NumBits+2]

X anchor point change

AnchorDeltaY

SB[NumBits+2]

Y anchor point change

ENDSHAPERECORD defines the end of the shape record array. STYLECHANGERECORD defines changes in line
style, fill style, position, or a new set of styles. STRAIGHTEDGERECORD and CURVEDEDGERECORD define a
straight or curved edge, respectively. The first bit in a shape record is a type flag. A 0 corresponds to a non-edge
record, and a 1 corresponds to an edge record. Looking at the first bit of our first shape record, we see that it is
not an edge record. Now we must look at the next five bits which are all flags that tell us what is to follow. If all
of the five bits are 0, then that is a type0 shape record and defines the end of the array of shape records.

229

25

C9 92

0010 0101

0D 21

1100 1001

1001 0010

0000 1101

0010 0001

0 = 0 = non edge record
01001 = 5 flags line style flag is true, and move to flag is true
Since the move to flag is true, the next five bits are the MoveBits field. This value is 14 so the next two fields
which are the MoveDeltaX, and the MoveDeltaY are of size 14. These are unsigned numbers.
01110 = MoveBits
01001100100100 = 4900 (twips) = 245 pixels = MoveDeltaX
00011010010000 = 1680 = 84 pixels = MoveDeltaY
Since the line style flag is true, the next field is a NumLineBits = 1 bit field representing the line style. This field is
equal to 1. This means that the line style for the line to follow is the first one in the line style array.
1 = 1 = line style
Now for the rest of the shape records:

ED 48 87 65 30 3B 6D E1 D8 B4 00 00

1110 1101 0100 1000 1000 0111 0110 0101 0011 0000 0011 1011 0110
1101 1110 0001 1101 1000 1011 0100 0000 0000 0000 0000

The next shape record begins with a 1, so it is an edge record.
The next bit indicates if it is a straight or curved edge. It is a 1, which stands for a straight edge. The next four
bits indicate the size of any delta fields which follow. The formula for the NumBits value is 2 + whatever the
value of that 4-bit field. In this case, the value of NumBits is 13. Following the NumBits field is a 1-bit line flag.
This indicates whether the line being described is a general line or horizontal/vertical line. The value of 0
corresponds to a hor/vert line, so the next bit is a VertLineFlag field and indicates whether the line is horizontal
or vertical. The value of the bit is 1 which corresponds to a vertical line. The next field for a vertical line is the
signed DeltaY field which is nbits = 13 bits long. The value corresponds to
116 pixels. That is the end of the shape record.
1 = 1 = edge record
1 = 1 = straight edge
230

1011 = 11 + 2 = 13 = NumBits
0 = 0 = hor/vert line
1 = 1 = vertical line
0100100010000 = 2320 twips = 116 pixels = DeltaY The next three records are very similar to the last one:
1 = 1 = edge record
1 = 1 = straight edge
1011 = 11 + 2 = 13 = NumBits
0 = 0 = hor/vert line
0 = 0 = horizontal line
1010011000000 = -2880 twips (2’s complement number) = -144 pixels = DeltaX
1 = 1 = edge record
1 = 1 = straight edge
1011 = 11 + 2 = 13 = NumBits
0 = 0 = hor/vert line
1 = 1 = vertical line
1011011110000 = -2320 twips = -116 pixels = DeltaY
1 = 1 = edge record
1 = 1 = straight edge
1011 = 11 + 2 = 13 = NumBits
0 = 0 = hor/vert line
0 = 0 = horizontal line
0101101000000 = 2880 twips = 144 pixels = DeltaX
Finally, the last shape record begins with a 0 which means it is not an edge record. Furthermore, all of its flag
bits are equal to 0, which means that it is the last shape record and we are through with our shape record array.
0 = 0 = non-edge record
000000 = flags (since they are all 0, this is the end of the shape record array
231

Since we are done with our structure, we must now fill our last byte with 0s to keep byte alignment.
000000 = padding bits so that the record aligns on a byte boundary
We are also done with our shape with style since the shape record array is the last element of the shape with
style. Since we are already byte aligned, we can go on to our next tagged data block.
The Tag type of the block is equal to 26 which corresponds to a PlaceObject2. The length field has a value of 6 so
the length of the data block is 6 bytes.
0x8606 ➜

0x0686 ➜

0000 0110 1000 0110

0000011010 = 26 = tag type = PlaceObject2
000110 = 6 = length

06 01 00 01 00 00
0000 0110 0000 0001 0000 0000 0000 0001 0000 0000 0000 0000
Here is the file description of the PlaceObject2 tag:
PlaceObject2
Field

Type

Comment

Header

RECORDHEADER

Tag type = 26.

PlaceFlagHasClipActions

UB[1]

SWF 5 or later: has clip actions (sprite characters
only). Otherwise: always 0.

PlaceFlagHasClipDepth

UB[1]

Has clip depth.

PlaceFlagHasName

UB[1]

Has name.

PlaceFlagHasRatio

UB[1]

Has ratio.

PlaceFlagHasColorTransform

UB[1]

Has color transform.

PlaceFlagHasMatrix

UB[1]

Has matrix.

PlaceFlagHasCharacter

UB[1]

Places a character.

PlaceFlagMove

UB[1]

Defines a character to be moved.

232

Depth

UI16

Depth of character.

CharacterId

If PlaceFlagHasCharacter
UI16

ID of character to place.

Matrix

If PlaceFlagHasMatrix
MATRIX

Transform matrix data.

ColorTransform

If
PlaceFlagHasColorTransform
CXFORMWITHALPHA

Color transform data.

Ratio

If PlaceFlagHasRatio UI16

Name

If PlaceFlagHasName
STRING Name of character.

ClipDepth

If PlaceFlagHasClipDepth
UI16

Clip depth (see Clipping layers).

ClipActions

If PlaceFlagHasClipActions
CLIPACTIONS

SWF 5 or later: Clip Actions Data.

The first eight bits of the body are all flags indicating what is to follow. A 1 in the sixth bit indicates that the body
has a transform matrix, and the 1 in the seventh bit indicates that the object to be placed has a character ID.
00000110 ➜ body has a transform matrix and object has a character ID
Following the flags is a 16-bit unsigned integer which indicates the depth of the character. In this case, the depth
is 1, which makes sense since the rectangle is the only object in the file.
0x0100 ➜

0x0001 ➜

depth = 1

Since the object has a character ID, the next field in the body is the unsigned 16-bit ID. Since the rectangle is the
only object in the file, the ID of the rectangle is 1.
0x0100 ➜

0x0001 ➜

character ID = 1

The final field for this PlaceObject2 is the transform matrix. Here is the file description:
MATRIX
Field

Type

Comment

HasScale

UB[1]

Has scale values if equal to 1.

233

NScaleBits

If HasScale = 1, UB[5]

Bits in each scale value field.

ScaleX

If HasScale = 1, FB[NScaleBits]

x scale value.

ScaleY

If HasScale = 1, FB[NScaleBits]

y scale value.

HasRotate

UB[1]

Has rotate and skew values if equal to 1.

NRotateBits

If HasRotate = 1, UB[5]

Bits in each rotate value field.

RotateSkew0

If HasRotate = 1,
FB[NRotateBits]

First rotate and skew value.

RotateSkew1

If HasRotate = 1,
FB[NRotateBits]

Second rotate and skew value.

NTranslateBits

UB[5]

Bits in each translate value field.

TranslateX

SB[NTranslateBits]

x translate value in twips.

TranslateY

SB[NTranslateBits]

y translate value in twips.

Since this shape has no transform information, the matrix is empty. All of its flag bits have values of zero. This is
not super efficient but it is valid.
0x00 ➜ completely empty matrix with leftover bits filled
Since we are done with our PlaceObject2, let’s take a look at our next tag.
0x4000 ➜

0x0040 ➜

0000 0000 0100 0000

Tag type = 1 = ShowFrame length = 0
We see that the tag is an instruction to show the frame. A ShowFrame has no body. Its length is 0, so we move
on to the next tag.
0x0000 0x0000 0000 0000 0000 0000
Tag type = 0 = end length = 0
We have reached the end tag which signals the end of our SWF file.

234

Appendix B: Reverse index of tag values
This table provides a quick lookup, allowing any tag in the SWF specification to be found by its tag value.
Tag
value

Tag name

0

End

1

ShowFrame

2

DefineShape

4

PlaceObject

5

RemoveObject

6

DefineBits

7

DefineButton

8

JPEGTables

9

SetBackgroundColor

10

DefineFont

11

DefineText

12

DoAction

13

DefineFontInfo

14

DefineSound

15

StartSound

17

DefineButtonSound

18

SoundStreamHead

19

SoundStreamBlock

20

DefineBitsLossless

21

DefineBitsJPEG2

22

DefineShape2

235

23

DefineButtonCxform

24

Protect

26

PlaceObject2

28

RemoveObject2

32

DefineShape3

33

DefineText2

34

DefineButton2

35

DefineBitsJPEG3

36

DefineBitsLossless2

37

DefineEditText

39

DefineSprite

43

FrameLabel

45

SoundStreamHead2

46

DefineMorphShape

48

DefineFont2

56

ExportAssets

57

ImportAssets

58

EnableDebugger

59

DoInitAction

60

DefineVideoStream

61

VideoFrame

62

DefineFontInfo2

64

EnableDebugger2

65

ScriptLimits

66

SetTabIndex

236

69

FileAttributes

70

PlaceObject3

71

ImportAssets2

73

DefineFontAlignZones

74

CSMTextSettings

75

DefineFont3

76

SymbolClass

77

Metadata

78

DefineScalingGrid

82

DoABC

83

DefineShape4

84

DefineMorphShape2

86

DefineSceneAndFrameLabelData

87

DefineBinaryData

88

DefineFontName

89

StartSound2

90

DefineBitsJPEG4

91

DefineFont4

93.

EnableTelemetry

237

Appendix C: Screen Video v2 Palette
The Screen Video v2 codec (see Chapter 14, “Video,”) can use a 15/7-bit hybrid colorspace in addition to a 24-bit
colorspace. The codec allows the bitstream to encode a 128-entry color palette. In the absence of a valid
palette, the code will fall back on the following 128-entry RGB palette.
Each of these 32-bit numbers is formatted as 0x00rrggbb.
unsigned int default_screen_video_v2_palette[128] = {
0x00000000,
0x00333333,
0x00666666,
0x00999999,
0x00CCCCCC,
0x00FFFFFF,
0x00330000,
0x00660000,
0x00990000,
0x00CC0000,
0x00FF0000,
0x00003300,
0x00006600,
0x00009900,
0x0000CC00,
0x0000FF00,
0x00000033,
0x00000066,
0x00000099,
0x000000CC,
0x000000FF,
0x00333300,
0x00666600,
0x00999900,
0x00CCCC00,
0x00FFFF00,
0x00003333,
0x00006666,
0x00009999,
0x0000CCCC,
0x0000FFFF,
0x00330033,
0x00660066,
0x00990099,
0x00CC00CC,
238

0x00FF00FF,
0x00FFFF33,
0x00FFFF66,
0x00FFFF99,
0x00FFFFCC,
0x00FF33FF,
0x00FF66FF,
0x00FF99FF,
0x00FFCCFF,
0x0033FFFF,
0x0066FFFF,
0x0099FFFF,
0x00CCFFFF,
0x00CCCC33,
0x00CCCC66,
0x00CCCC99,
0x00CCCCFF,
0x00CC33CC,
0x00CC66CC,
0x00CC99CC,
0x00CCFFCC,
0x0033CCCC,
0x0066CCCC,
0x0099CCCC,
0x00FFCCCC,
0x00999933,
0x00999966,
0x009999CC,
0x009999FF,
0x00993399,
0x00996699,
0x0099CC99,
0x0099FF99,
0x00339999,
0x00669999,
0x00CC9999,
0x00FF9999,
0x00666633,
0x00666699,
0x006666CC,
0x006666FF,
0x00663366,
0x00669966,
0x0066CC66,
0x0066FF66,
239

0x00336666,
0x00996666,
0x00CC6666,
0x00FF6666,
0x00333366,
0x00333399,
0x003333CC,
0x003333FF,
0x00336633,
0x00339933,
0x0033CC33,
0x0033FF33,
0x00663333,
0x00993333,
0x00CC3333,
0x00FF3333,
0x00003366,
0x00336600,
0x00660033,
0x00006633,
0x00330066,
0x00663300,
0x00336699,
0x00669933,
0x00993366,
0x00339966,
0x00663399,
0x00996633,
0x006699CC,
0x0099CC66,
0x00CC6699,
0x0066CC99,
0x009966CC,
0x00CC9966,
0x0099CCFF,
0x00CCFF99,
0x00FF99CC,
0x0099FFCC,
0x00CC99FF,
0x00FFCC99,
0x00111111,
0x00222222,
0x00444444,
0x00555555,
240

0x00AAAAAA,
0x00BBBBBB,
0x00DDDDDD,
0x00EEEEEE
};

241