audioJS

@audio/voice-glottis

Status calibrated

npm1.0.2licenseMIT

Glottal source models — LF model, Rosenberg pulse (excitation for tract/formant synthesis)

Overview

Install
Updated
Family
voice, 3 packages under it
Tags
make

LF model (Fant–Liljencrants–Lin 1985, Rd-controlled via Fant 1994's regression) and the Rosenberg (1971) pulse, cycle-repeated with jitter/shimmer/aspiration noise.

import glottis, { lfCycle, rosenbergCycle } from '@audio/voice-glottis'

let excitation = glottis({ f0: 120, duration: 0.5 })   // → Float32Array

glottis(opts: {f0=120, duration=1, fs=44100, model='lf', Rd=1.7, jitter=0.005, shimmer=0.03, aspiration=0.01, seed}) → Float32Array

lfCycle(n, {Rd=1.7}) / rosenbergCycle(n) — single-cycle waveforms, for custom sequencing.

Install