@audio/tune-snap

Scale/key pitch correction — YIN track → note segments → scale snap → PSOLA retune (Auto-Tune class)

npm install @audio/tune-snap

import snap from '@audio/tune-snap'

YIN F0 tracking per frame, segmented into note runs (a run splits wherever pitch jumps more than 80¢ from its running median), each segment's median pitch snapped to the nearest note in scale/root, then shifted there with PSOLA and crossfaded back over the dry signal. v1 corrects per note segment — no intra-note glide/portamento; unvoiced audio passes through untouched.

snap(data, { scale: 'major', root: 0 })              // classic Auto-Tune-style scale correction
snap(data, { scale: 'chromatic', strength: 0.5 })    // half-strength chromatic snap ("natural" mode)

| Param | Default | | |---|---|---| | fs | 44100 | Hz | | scale | 'chromatic' | 'chromatic', 'major', 'minor', 'harmonic-minor', 'melodic-minor', 'pentatonic-major', 'pentatonic-minor', 'blues', 'dorian', 'mixolydian', 'whole' | | root | 0 | 0–11, root note (0 = C) | | a4 | 440 | Hz, tuning reference | | tolerance | 8 | cents; segments already this close to a scale note are left alone | | strength | 1 | 0–1, correction amount (0 = no shift, 1 = full snap) | | frameSize | 2048 | YIN analysis frame | | hop | 512 | frame hop | | fade | 0.01 | s, crossfade at each segment boundary |

Built on @audio/pitch-yin (F0 tracking) and @audio/shift-psola (pitch shifting). The melody-following complement is @audio/tune-midi, which retunes to a reference note guide instead of a scale.

Use when: vocal/instrument scale correction, Auto-Tune-style effects. Not for: retuning to a specific melody/guide track — use tune-midi.

Part of @audio/tune — the tune family umbrella.

MIT © audiojs