npm i wavesurfer.js-async

Motivation

This piece of code is a plugin for wavesurfer.js which prevent blocking main browser thread during analyzing audio file. This problem occurs during analyzing long audio files or high resolution analyzing eg. minPxPerSec = 400 what's means that each second of file produce an array of 800 items for each channel. During analyzing 9 min audio this producing arrays of ~~ 432 000 elements. Computing so long arrays at once in browser environment blocking main tread for hundreds of second.

WARNING Implement only for WebAudio backend. (default)

Modifications

• modification - wavesurfer.drawBuffer- to using getPeeksAsync
• adding - wavesurfer.Backend.prototype.getPeeksAsync

How it works

During analysis audio data for each channel data of audio buffer is return promise. This promise run setTimeout with function iterating through whole buffer. If iterating take more that yieldIteration (16 ms by default) then stops loop and resolve promise with value of index to which peeks are computed. If is not whole buffer then function is run again but starting from point on which previous call was ended. In such way iteration of audio buffer is split to 16 ms chunks what's allow keeping 60 frames per second.

Usage

import WaveSurfer from "wavesurfer.js"
import WaveSurferAsync from "wavesurfer.js-async"


const config = {
    container: document.getElementyById("container"),
    minPxPerSec : 300, // the benefits are visible in big zooms 
    plugins: [WaveSurferAsync.create()]
}

const ws = WaveSurfer.create(config);

You can adjust time after witch loop will be break and computes will be moved to next call/iteration

plugins: [
    WaveSurferAsync.create({
        yieldIteration: 16 // optional break point time - default 16 ms
    })
]