fix: support existing voicing dicts in new logic

+ rename voiceBelow to renderVoicing
This commit is contained in:
Felix Roos 2023-07-12 22:38:44 +02:00
parent 0b3a8a5f65
commit b0cbeda796
3 changed files with 26 additions and 14 deletions

View file

@ -1,14 +1,16 @@
import { isNote, isNoteWithOctave, _mod } from '@strudel.cycles/core';
import { Interval } from '@tonaljs/tonal';
// https://codesandbox.io/s/stateless-voicings-g2tmz0?file=/src/lib.js:0-2515
const flats = ['C', 'Db', 'D', 'Eb', 'E', 'F', 'Gb', 'G', 'Ab', 'A', 'Bb', 'B'];
const pcs = ['c', 'db', 'd', 'eb', 'e', 'f', 'gb', 'g', 'ab', 'a', 'bb', 'b'];
const sharps = ['C', 'C#', 'D', 'D#', 'E', 'F', 'F#', 'G', 'G#', 'A', 'A#', 'B'];
const accs = { b: -1, '#': 1 };
export const pc2chroma = (pc) => {
const [letter, ...rest] = pc.split('');
return flats.indexOf(letter) + rest.reduce((sum, sign) => sum + accs[sign], 0);
return pcs.indexOf(letter.toLowerCase()) + rest.reduce((sum, sign) => sum + accs[sign], 0);
};
export const rotateChroma = (chroma, steps) => (chroma + steps) % 12;
@ -54,6 +56,14 @@ export const x2chroma = (x) => {
}
};
export const step2semitones = (x) => {
let num = Number(x);
if (!isNaN(num)) {
return num;
}
return Interval.semitones(x);
};
export const x2midi = (x) => {
if (typeof x === 'number') {
return x;
@ -77,13 +87,13 @@ export function scaleStep(notes, offset) {
return notes[offset % notes.length] + octOffset;
}
export function voiceBelow(maxNote, chord, voicingDictionary, offset = 0, n) {
export function renderVoicing(chord, anchor, voicingDictionary, offset = 0, n) {
const [root, symbol] = tokenizeChord(chord);
const maxPc = note2pc(maxNote);
const maxPc = note2pc(anchor);
const maxChroma = pc2chroma(maxPc);
const rootChroma = pc2chroma(root);
const voicings = voicingDictionary[symbol].map((voicing) =>
typeof voicing === 'string' ? voicing.split(' ').map((n) => parseInt(n, 10)) : voicing,
(typeof voicing === 'string' ? voicing.split(' ') : voicing).map(step2semitones),
);
let minDistance, bestIndex;
@ -100,7 +110,7 @@ export function voiceBelow(maxNote, chord, voicingDictionary, offset = 0, n) {
const octDiff = Math.ceil(offset / voicings.length) * 12;
const indexWithOffset = _mod(bestIndex + offset, voicings.length);
const voicing = voicings[indexWithOffset];
const maxMidi = note2midi(maxNote);
const maxMidi = note2midi(anchor);
const topMidi = maxMidi - chromaDiffs[indexWithOffset] + octDiff;
const voicingMidi = voicing.map((v) => topMidi - voicing[voicing.length - 1] + v);