InfiniteDSP Core

A modular, high-performance audio DSP library for Rust, designed for real-time synthesis and effects processing. It is no_std compatible (requires alloc), making it suitable for embedded audio applications as well as desktop software.

Features

• no_std Compatible: Built for portability using libm and alloc.
• Type-Safe Channel System: Explicit Mono and Stereo types prevent routing errors.
• Modular Architecture: Build complex audio chains using DspChain and Mixer.
• Static Dispatch: Use StaticDspChain for zero-overhead composition of processors, allowing for aggressive compiler inlining.
• AudioParam System: All parameters can be static, linked to thread-safe controls (atomics), or modulated by other audio signals (e.g., LFOs, Envelopes) at sample rate.
• Signal Math: Combine signals easily with Add and Multiply processors.
• SIMD Optimization: Uses wide for SIMD-accelerated processing where applicable.
• Graph Visualization: Generate ASCII diagrams of your signal chain for easy debugging (chain.get_graph()).
• Spectral Processing: Includes a robust Overlap-Add (OLA) engine for FFT-based effects.
• Comprehensive Effect Suite:
  • Time: Delay (Standard & LowMem), Tape Delay (with saturation & flutter), PingPongDelay, Reverb (Schroeder, Standard & LowMem), Stutter.
  • Filter: Biquad (LowPass, HighPass, BandPass, Notch), Ladder Filter (Moog-style, both Iterative and Predictive ZDF), State Variable Filter (TPT/ZDF), Vowel Filter.
  • Dynamics: Compressor, Limiter, Distortion (Soft/Hard Clip, BitCrush, Foldback).
  • Modulation: Phaser, Tremolo, Ring Modulator, Chorus, Flanger.
  • Spectral: FFT Pitch Shift, Granular Pitch Shift, Spectral Filter.
  • Utility: Gain, Offset, Stereo Panner, Stereo Widener, MapRange, TimedGate.
• Synthesis:
  • Oscillators: Sine, Triangle, Saw, Square (PolyBLEP anti-aliased), Noise, Stack (Detuned Multi-Osc).
  • Vocal: Speech Synthesizer (Formant-based).
  • Physical Modeling: Karplus-Strong (String), Brass Model.
  • Control: LFO, ADSR Envelope (with retrigger support).

Benchmarks

Performance is tracked over time to ensure no regressions. View Benchmark Charts

Documentation

View Documentation

Demos

Listen to some of the examples generated with this library:

Showcase

Check out these projects built with infinitedsp-core:

• InfiniteTrak
• picoDSP
• picoDSP-Edit

Project Structure

• src/core: Core traits and infrastructure (FrameProcessor, AudioParam, DspChain, Ola, ParallelMixer, SummingMixer, Stereo).
• src/effects: Audio effects implementations.
• src/synthesis: Sound generators and control signals.
• examples_app: A separate workspace member containing runnable examples using cpal.

Usage

Add infinitedsp-core to your dependencies.

use infinitedsp_core::core::dsp_chain::DspChain;
use infinitedsp_core::core::audio_param::AudioParam;
use infinitedsp_core::core::channels::Mono;
use infinitedsp_core::synthesis::oscillator::{Oscillator, Waveform};
use infinitedsp_core::effects::time::delay::Delay;

// Create an oscillator (Mono source)
let osc = Oscillator::new(AudioParam::hz(440.0), Waveform::Saw);

// Create a delay effect (Mono effect)
let delay = Delay::new(
    1.0, // Max delay time in seconds
    AudioParam::ms(350.0),   // Delay time
    AudioParam::linear(0.5), // Feedback
    AudioParam::linear(0.3)  // Mix
);

// Chain them together. The chain is typed as DspChain<Mono>.
let mut chain = DspChain::new(osc, 44100.0).and(delay);

// Print the signal chain (requires 'debug_visualize' feature)
println!("{}", chain.get_graph());

// Process a buffer
let mut buffer = [0.0; 512];
chain.process(&mut buffer, 0);

Feature Flags

• debug_visualize: Enables get_graph() and visualize() methods for debugging signal chains. Disabled by default to minimize binary size for embedded targets.

Running Examples

The project includes several runnable examples in the examples_app folder that demonstrate different capabilities using cpal for real-time audio output.

Run an example using:

cargo run --release -p infinitedsp-examples --bin <example_name>

Available Examples:

• infinitedsp_demo: A complex polyphonic demo showcasing 30 voices, filters, envelopes, and effects (Stereo).
• speech_synth: Formant-based vocal synthesizer demo with rhythmic glitch effects.
• filter_sweep: Compares PredictiveLadderFilter vs LadderFilter with an LFO sweep (Mono).
• dual_mono_demo: Demonstrates independent processing of Left/Right channels (Ping-Pong Delay).
• ping_pong_demo: Demonstrates the stereo PingPongDelay effect.
• trance_synth: A massive stereo supersaw trance pluck with delay, reverb, and a sequencer.
• karplus_demo: Physical modeling of a guitar string (Karplus-Strong algorithm) (Mono).
• svf_demo: State Variable Filter demonstration (BandPass sweep).
• spectral_demo: FFT-based Pitch Shifting using the Overlap-Add (OLA) engine.
• granular_demo: Time-domain Granular Pitch Shifting.
• modulation_demo: Showcases Tremolo, Chorus, and Tape Delay.
• phaser_demo: 6-stage Phaser effect.
• effects_demo: Demonstrates signal math (Add/Multiply) and distortion.

Documentation

To generate and view the API documentation:

cargo doc --open

AI Contribution Policy

This project allows and encourages experimentation with AI agents for code generation and optimization. However, all AI-generated contributions must be strictly verified by a human maintainer. This verification includes:

1. Code Review: Ensuring the code is idiomatic, safe, and follows project standards.
2. Audio Verification: Listening to the output to ensure correctness and high audio quality (no artifacts, correct DSP behavior).

License

This project is licensed under the MIT License - see the LICENSE file for details.