@synthlet/chorus

A multi-voice stereo chorus, as an audio worklet

Part of Synthlet.

Several delayed copies of the input, each read at a position its own LFO moves, summed to a stereo pair. Mono in, stereo out — the width comes from spreading the voices' LFOs apart rather than from panning.

All four controls are normalised 0…1 knob positions, not physical quantities: this is a pedal, not a modular utility.

Install

npm i @synthlet/chorus

Or npm i synthlet for every module at once.

Usage

import { registerChorusWorklet, Chorus } from "@synthlet/chorus";

const ac = new AudioContext();
await registerChorusWorklet(ac);

const chorus = Chorus(ac, {
  delay: 0.5,
  rate: 0.3,
  depth: 0.6,
  deviation: 0.5,
});

source.connect(chorus).connect(ac.destination);

Parameters

| Param | Default | Range | Rate | Meaning | | ----------- | ------- | ----- | ------ | ----------------------------------------------------- | | delay | 0.5 | 0 … 1 | k-rate | Where the moving read position is centred | | rate | 0.5 | 0 … 1 | k-rate | How fast the internal LFOs run | | depth | 0.5 | 0 … 1 | k-rate | How far they swing | | deviation | 0.5 | 0 … 1 | k-rate | How far the voices are spread apart in phase and rate |

All four are k-rate and handed to the engine's per-block update — the engine's own LFOs are what move per sample, which is what makes this a chorus rather than a delay you have to modulate yourself.

Only the first input channel is read, and the output is always stereo.

Credits

The DSP is compiled from Faust. The committed Faust source (dsp/chorus.dsp, not published to npm) contains a chorus_mono definition copied from the Faust distribution's examples/SAM/chorus/chorusForBrowser.dsp, Copyright (C) 2003-2024 GRAME, Centre National de Creation Musicale and (C) 2023-2024 INRIA, LGPL 2.1-or-later.

See the repository's THIRD-PARTY-LICENSES.md.

License

MIT © danigb