MIDI Control #1

A modular polyphonic MIDI synth voice plugin built with the JUCE C++ framework.
Designed for rapid experimentation with per-voice DSP modules, control mapping, and automation.
Part of an ongoing research and engineering project integrating hardware MIDI controllers with modern synthesis architectures.

---

Overview

MIDIControl001 implements a real‑time plugin environment in which each active MIDI note instantiates a self‑contained VoiceX — a modular voice that combines an OscillatorX and EnvelopeX (or their future variants). The system is designed so that adding new synthesis modes (VoiceB, VoiceC, etc.) requires no change to the plugin’s overall structure.

• Cross‑platform: Built with JUCE (VST3 / AU / Standalone).
• Thread‑safe: Uses AudioProcessorValueTreeState for GUI ↔ DSP synchronization.
• Extensible: Swap in new oscillator/envelope modules per voice.
• Realtime‑safe: Parameter snapshots and smoothed values prevent clicks or zipper noise.
• Human‑tuned: All CC controls normalized and perceptually mapped to musical ranges.

---

Key Features

Category	Description
Voice Model	Modular VoiceX structure with dedicated OscillatorX + EnvelopeX per note
Parameter Handling	AudioProcessorValueTreeState + per‑block ParameterSnapshot
Polyphony Management	VoiceManager handles allocation, stealing, and global gain smoothing
MIDI Control	CCs normalized (0–127 → 0–1); persistent state between notes
Extensibility	Add new voice types (VoiceB, VoiceC, …) without changing host logic
Visualization	Lock‑free GUI feedback loop for meters and waveform previews
DSP Safety	All parameter reads lock‑free; all updates real‑time safe

---

Architecture Summary

See the full engineering whitepaper in docs/architecture.md.
That document replaces older internal diagrams and fully details:

• The PluginProcessor ↔ VoiceManager ↔ VoiceX dataflow
• Real‑time parameter snapshotting and smoothing
• Modular DSP composition and controller mapping
• Design rules for new voice types and extensions

graph TD
    Host["Host (DAW / Standalone / CLI)"] --> Processor["PluginProcessor"]
    Processor --> VM["VoiceManager"]
    VM --> VX["VoiceX (per-note)"]
    VX --> OSC["OscillatorX"]
    VX --> ENV["EnvelopeX"]
    VM -->|mixdown| Audio["AudioBuffer<float>"]
    Audio --> Host

Loading

---

Getting Started

Build Requirements

• JUCE 7.x or newer
• CMake 3.25+
• C++17 or newer compiler (tested with AppleClang 13+)

Build Instructions

git clone https://github.com/TYLERSFOSTER/MIDIControl001.git
cd MIDIControl001
cmake -S . -B build -DCMAKE_BUILD_TYPE=Release
cmake --build build -j

Run Standalone

./build/MIDIControl001_artefacts/Release/Standalone/MIDIControl001.app/Contents/MacOS/MIDIControl001

Or load the VST3/AU binary in your DAW.

---

Contributing

1. Fork the repository.
2. Implement new voice logic under Source/dsp/voices/.
3. Add tests under Tests/dsp/voices/.
4. Document architectural changes in docs/architecture.md.
5. Submit a pull request to main with a short engineering summary.

---

Voice Types Roadmap

See the full Voice Type Roadmap in docs/VOICE_TYPES_ROADMAP.md.

That document details the planned evolution from the baseline VoiceA / OscillatorA / EnvelopeA to more advanced synthesis architectures such as:

• VoiceDopp — dual-modulator voice with cross-coupled envelopes
• VoiceLET — Lorentz-equivalent synthesis model
• VoiceEns — ensemble and frequency-band layering system

Each new type builds on the modular VoiceX framework described above, extending MIDIControl001 into a flexible, research-grade synthesis platform.

---

License

This project is distributed under the MIT License.
See LICENSE for details.

---

Created and maintained by Tyler Foster.