@synthlet/clock

A tempo clock with two outputs: a phase ramp and a gate

Part of Synthlet.

The thing a sequenced patch is built on. It emits two different signals, because the two things that consume a clock want different shapes:

• the phase ramp, on the node's own output — a 0…1 sawtooth restarting each beat. Subdividing a clock means multiplying its phase, so this is what Euclid reads.
• the gate, on .gate — high for pulseWidth of each beat. This is what an envelope or a drum voice reads.

They are not interchangeable. Feeding the ramp to an envelope's trigger used to work by accident and does not any more.

Install

npm i @synthlet/clock

Or npm i synthlet for every module at once.

Usage

import { registerClockWorklet, Clock } from "@synthlet/clock";

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

const clock = Clock(ac, { bpm: 120, pulseWidth: 0.25 });

Euclid(ac, { clock }); // the ramp
KickDrum(ac, { trigger: clock.gate }); // the gate

clock.dispose(); // disposes the gate node with it

Parameters

| Param | Default | Range | Rate | Meaning | | ------------ | ------- | -------- | ------ | ------------------------------------------ | | bpm | 120 | 0 … 1000 | k-rate | Tempo in beats per minute | | pulseWidth | 0.5 | 0 … 1 | k-rate | Fraction of each beat the gate is high for |

Both are k-rate: the phase increment is derived from bpm and cached, and a tempo that changed every sample is frequency modulation of the clock rather than a tempo — Lfo and Param are the modules for that. What the clock emits is written per sample either way.

pulseWidth is a fraction of the beat rather than a fixed duration, so at 120 BPM the default is 250 ms — about 86 render quanta, wide enough that no consumer can miss it.

One phase accumulator per node: two Clock nodes drift unless they were built in the same render quantum, which is why the gate is a second output rather than a separate ClockGate module.

License

MIT © danigb