Justifier

A desktop application for exploring just intonation through direct manipulation of sound.

Build soundscapes by layering voices tuned to JI ratios from a reference frequency, tweaking parameters, and hearing the result in real time. Every parameter is visible and tweakable — you learn what's possible by experimenting.

Features

• 12 waveform types: sine, triangle, saw, square, pulse, white/pink/brown noise, lfnoise0/1/2, FM synthesis
• Just intonation ratios: set numerator/denominator per voice, octave selector, detune in cents
• Per-voice controls: amplitude, pan, waveform crossfade (20ms), enable/mute toggle
• Per-voice multi-mode filter: lowpass, highpass, bandpass, notch with cutoff and resonance
• Per-voice reverb send: stereo send/return bus with Faust Zita Rev1 reverb, global return level
• ADSR envelope: per-voice attack, decay, sustain, release with proper voice release lifecycle
• FM synthesis controls: mod ratio and mod index (FM waveform only)
• Sleep-before-destroy: voices fade out over their release time with undo option
• Wave grouping: organize voices into named, colored wave columns
• 32-voice polyphony: pre-allocated pool, zero allocation on the audio thread
• Lock-free audio: SPSC command queue, atomic panic flag, stack-allocated buffers
• Cross-platform: Linux, macOS, Windows (miniaudio backend)

Architecture

• Audio engine (native/): C/C++ library using Faust DSP and miniaudio
  • 12 voice DSPs compiled from Faust at build time via CMake
  • Reverb effect as singleton Faust DSP (Zita Rev1 stereo), separate from voice pool
  • Pre-allocated 32-voice pool, lock-free SPSC command queue (moodycamel::ReaderWriterQueue)
  • 20ms waveform crossfade, stereo send bus for reverb, atomic panic flag
  • Zero allocation on the audio thread
• FFI bridge (lib/audio/): Dart FFI bindings to the native audio engine, platform-specific library loading
• UI (lib/): Flutter with Riverpod state management, Material 3 dark theme
  • Voice cards with inline sliders for all parameters
  • Waveform selector, ratio input, octave controls
  • Filter type buttons (LP/HP/BP/NT), cutoff/resonance sliders
  • Reverb send slider, ADSR envelope sliders

Building

Prerequisites

• Faust compiler (pacman -S faust / brew install faust)
• CMake 3.14+
• ALSA dev headers (Linux only)
• Flutter SDK

Native audio engine

cd native
mkdir -p build && cd build
cmake ..
make

Flutter app

flutter run -d linux   # or -d macos, -d windows

Tests

# C audio engine tests
cd native/build
./test_audio           # full mode (~30s, audible verification)
./test_audio --quick   # quick mode (~10s)

# Flutter widget tests
flutter test

Status

• Phase 1 (C audio engine): complete
• Phase 2 (Dart FFI bridge): complete
• Phase 3 (Flutter UI): complete — voice cards, wave columns, all controls
• Phase 4 (Audio features):
  • 4a Per-voice filters: complete
  • 4b1 ADSR envelope: complete
  • 4c Effects bus (reverb): complete
  • 4b2 Filter envelope: planned
  • 4d JI lattice explorer: planned