Merge branch 'main' into supradough

packages/superdough/superdough.mjs

@@ -7,7 +7,7 @@ This program is free software: you can redistribute it and/or modify it under th
 import './feedbackdelay.mjs';
 import './reverb.mjs';
 import './vowel.mjs';
-import { clamp, nanFallback, _mod, cycleToSeconds } from './util.mjs';
+import { clamp, nanFallback, _mod, cycleToSeconds, secondsToCycle } from './util.mjs';
 import workletsUrl from './worklets.mjs?audioworklet';
 import { createFilter, gainNode, getCompressor, getWorklet } from './helpers.mjs';
 import { map } from 'nanostores';
@@ -28,6 +28,13 @@ export function setMultiChannelOrbits(bool) {
   multiChannelOrbits = bool == true;
 }
 
+function getModulationShapeInput(val) {
+  if (typeof val === 'number') {
+    return val % 5;
+  }
+  return { tri: 0, triangle: 0, sine: 1, ramp: 2, saw: 3, square: 4 }[val] ?? 0;
+}
+
 export const soundMap = map();
 
 export function registerSound(key, onTrigger, data = {}) {
@@ -342,9 +349,33 @@ function getDelay(orbit, delaytime, delayfeedback, t, channels) {
   return delays[orbit];
 }
 
-export function getLfo(audioContext, time, end, properties = {}) {
-  return getWorklet(audioContext, 'lfo-processor', {
+export function getLfo(audioContext, begin, end, properties = {}) {
+  const { shape = 0, ...props } = properties;
+  const { dcoffset = -0.5, depth = 1 } = properties;
+  const lfoprops = {
     frequency: 1,
+    depth,
+    skew: 0.5,
+    phaseoffset: 0,
+    time: begin,
+    begin,
+    end,
+    shape: getModulationShapeInput(shape),
+    dcoffset,
+    min: dcoffset * depth,
+    max: dcoffset * depth + depth,
+    curve: 1,
+    ...props,
+  };
+
+  return getWorklet(audioContext, 'lfo-processor', lfoprops);
+}
+
+export function getSyncedLfo(audioContext, time, end, cps, cycle, properties = {}) {
+  const frequency = cycle / cps;
+
+  return getWorklet(audioContext, 'lfo-processor', {
+    frequency,
     depth: 1,
     skew: 0,
     phaseoffset: 0,
@@ -467,9 +498,10 @@ function mapChannelNumbers(channels) {
   return (Array.isArray(channels) ? channels : [channels]).map((ch) => ch - 1);
 }
 
-export const superdough = async (value, t, hapDuration, cps = 0.5) => {
+export const superdough = async (value, t, hapDuration, cps = 0.5, cycle = 0.5) => {
   // new: t is always expected to be the absolute target onset time
   const ac = getAudioContext();
+
   let { stretch } = value;
   if (stretch != null) {
     //account for phase vocoder latency
@@ -495,6 +527,12 @@ export const superdough = async (value, t, hapDuration, cps = 0.5) => {
   }
   // destructure
   let {
+    tremolo,
+    tremolosync,
+    tremolodepth = 1,
+    tremoloskew,
+    tremolophase = 0,
+    tremoloshape,
     s = getDefaultValue('s'),
     bank,
     source,
@@ -504,6 +542,7 @@ export const superdough = async (value, t, hapDuration, cps = 0.5) => {
     // filters
     fanchor = getDefaultValue('fanchor'),
     drive = 0.69,
+    release = 0,
     // low pass
     cutoff,
     lpenv,
@@ -536,6 +575,7 @@ export const superdough = async (value, t, hapDuration, cps = 0.5) => {
     phasercenter,
     //
     coarse,
+
     crush,
     shape,
     shapevol = getDefaultValue('shapevol'),
@@ -578,8 +618,11 @@ export const superdough = async (value, t, hapDuration, cps = 0.5) => {
   distortvol = applyGainCurve(distortvol);
   delay = applyGainCurve(delay);
   velocity = applyGainCurve(velocity);
+  tremolodepth = applyGainCurve(tremolodepth);
   gain *= velocity; // velocity currently only multiplies with gain. it might do other things in the future
 
+  const end = t + hapDuration;
+  const endWithRelease = end + release;
   const chainID = Math.round(Math.random() * 1000000);
 
   // oldest audio nodes will be destroyed if maximum polyphony is exceeded
@@ -654,7 +697,7 @@ export const superdough = async (value, t, hapDuration, cps = 0.5) => {
         lprelease,
         lpenv,
         t,
-        t + hapDuration,
+        end,
         fanchor,
         ftype,
         drive,
@@ -678,7 +721,7 @@ export const superdough = async (value, t, hapDuration, cps = 0.5) => {
         hprelease,
         hpenv,
         t,
-        t + hapDuration,
+        end,
         fanchor,
       );
     chain.push(hp());
@@ -689,20 +732,7 @@ export const superdough = async (value, t, hapDuration, cps = 0.5) => {
 
   if (bandf !== undefined) {
     let bp = () =>
-      createFilter(
-        ac,
-        'bandpass',
-        bandf,
-        bandq,
-        bpattack,
-        bpdecay,
-        bpsustain,
-        bprelease,
-        bpenv,
-        t,
-        t + hapDuration,
-        fanchor,
-      );
+      createFilter(ac, 'bandpass', bandf, bandq, bpattack, bpdecay, bpsustain, bprelease, bpenv, t, end, fanchor);
     chain.push(bp());
     if (ftype === '24db') {
       chain.push(bp());
@@ -720,6 +750,33 @@ export const superdough = async (value, t, hapDuration, cps = 0.5) => {
   shape !== undefined && chain.push(getWorklet(ac, 'shape-processor', { shape, postgain: shapevol }));
   distort !== undefined && chain.push(getWorklet(ac, 'distort-processor', { distort, postgain: distortvol }));
 
+  if (tremolosync != null) {
+    tremolo = cps * tremolosync;
+  }
+
+  if (tremolo !== undefined) {
+    // Allow clipping of modulator for more dynamic possiblities, and to prevent speaker overload
+    // EX:  a triangle waveform will clip like this /-\ when the depth is above 1
+    const gain = Math.max(1 - tremolodepth, 0);
+    const amGain = new GainNode(ac, { gain });
+
+    const time = cycle / cps;
+    const lfo = getLfo(ac, t, endWithRelease, {
+      skew: tremoloskew ?? (tremoloshape != null ? 0.5 : 1),
+      frequency: tremolo,
+      depth: tremolodepth,
+      time,
+      dcoffset: 0,
+      shape: tremoloshape,
+      phaseoffset: tremolophase,
+      min: 0,
+      max: 1,
+      curve: 1.5,
+    });
+    lfo.connect(amGain.gain);
+    chain.push(amGain);
+  }
+
   compressorThreshold !== undefined &&
     chain.push(
       getCompressor(ac, compressorThreshold, compressorRatio, compressorKnee, compressorAttack, compressorRelease),
@@ -733,7 +790,7 @@ export const superdough = async (value, t, hapDuration, cps = 0.5) => {
   }
   // phaser
   if (phaser !== undefined && phaserdepth > 0) {
-    const phaserFX = getPhaser(t, t + hapDuration, phaser, phaserdepth, phasercenter, phasersweep);
+    const phaserFX = getPhaser(t, endWithRelease, phaser, phaserdepth, phasercenter, phasersweep);
     chain.push(phaserFX);
   }
 

packages/superdough/util.mjs

@@ -72,3 +72,7 @@ export const getSoundIndex = (n, numSounds) => {
 export function cycleToSeconds(cycle, cps) {
   return cycle / cps;
 }
+
+export function secondsToCycle(t, cps) {
+  return t * cps;
+}

packages/superdough/worklets.mjs

@@ -31,7 +31,7 @@ function polyBlep(phase, dt) {
     return 0;
   }
 }
-
+// The order is important for dough integration
 const waveshapes = {
   tri(phase, skew = 0.5) {
     const x = 1 - skew;
@@ -81,10 +81,12 @@ function getParamValue(block, param) {
   }
   return param[0];
 }
+
 const waveShapeNames = Object.keys(waveshapes);
 class LFOProcessor extends AudioWorkletProcessor {
   static get parameterDescriptors() {
     return [
+      { name: 'begin', defaultValue: 0 },
       { name: 'time', defaultValue: 0 },
       { name: 'end', defaultValue: 0 },
       { name: 'frequency', defaultValue: 0.5 },
@@ -92,7 +94,10 @@ class LFOProcessor extends AudioWorkletProcessor {
       { name: 'depth', defaultValue: 1 },
       { name: 'phaseoffset', defaultValue: 0 },
       { name: 'shape', defaultValue: 0 },
+      { name: 'curve', defaultValue: 1 },
       { name: 'dcoffset', defaultValue: 0 },
+      { name: 'min', defaultValue: 0 },
+      { name: 'max', defaultValue: 1 },
     ];
   }
 
@@ -109,10 +114,14 @@ class LFOProcessor extends AudioWorkletProcessor {
   }
 
   process(inputs, outputs, parameters) {
+    const begin = parameters['begin'][0];
     // eslint-disable-next-line no-undef
     if (currentTime >= parameters.end[0]) {
       return false;
     }
+    if (currentTime <= begin) {
+      return true;
+    }
 
     const output = outputs[0];
     const frequency = parameters['frequency'][0];
@@ -122,7 +131,11 @@ class LFOProcessor extends AudioWorkletProcessor {
     const skew = parameters['skew'][0];
     const phaseoffset = parameters['phaseoffset'][0];
 
+    const curve = parameters['curve'][0];
+
     const dcoffset = parameters['dcoffset'][0];
+    const min = parameters['min'][0];
+    const max = parameters['max'][0];
     const shape = waveShapeNames[parameters['shape'][0]];
 
     const blockSize = output[0].length ?? 0;
@@ -134,8 +147,9 @@ class LFOProcessor extends AudioWorkletProcessor {
     const dt = frequency / sampleRate;
     for (let n = 0; n < blockSize; n++) {
       for (let i = 0; i < output.length; i++) {
-        const modval = (waveshapes[shape](this.phase, skew) + dcoffset) * depth;
-        output[i][n] = modval;
+        let modval = (waveshapes[shape](this.phase, skew) + dcoffset) * depth;
+        modval = Math.pow(modval, curve);
+        output[i][n] = clamp(modval, min, max);
       }
       this.incrementPhase(dt);
     }