BeatTracker

A lightweight, pure JavaScript beat tracking library for the browser. It uses Web Workers for off-main-thread processing and FFT for spectral analysis to detect beats in audio files.

Note: The current version performs best with piano music.

Features

• Web Worker Support: Performs heavy computations (FFT, spectral flux) in a background thread to keep the UI responsive.
• Spectral Flux Analysis: Uses frequency domain analysis to detect onset strength.
• Intelligent Beat Tracking: Implements an agent-based approach with tempo induction and phase alignment.
• Zero External Dependencies: Self-contained logic using only native Web APIs.

Installation

This is a standalone module. Copy the src folder to your project.

Usage

Basic Usage

import { beatsEval } from './path/to/BeatTracker.js';

// Get an AudioBuffer (e.g., from AudioContext.decodeAudioData)
const audioContext = new AudioContext();
const response = await fetch('music.mp3');
const arrayBuffer = await response.arrayBuffer();
const audioBuffer = await audioContext.decodeAudioData(arrayBuffer);

// Analyze beats
try {
    const beatSamples = await beatsEval(audioBuffer, (progress) => {
        console.log(`Progress: ${progress.progress * 100}%`);
    });
    
    console.log('Detected beat sample indices:', beatSamples);
    
    // Convert samples to seconds
    const beatTimes = beatSamples.map(sample => sample / audioBuffer.sampleRate);
    console.log('Beat times (seconds):', beatTimes);
    
} catch (err) {
    console.error('Beat tracking failed:', err);
}

Finding Nearest Beat in Window

If you need to find a beat near a specific timestamp (e.g., for quantization or syncing):

import { findNearestBeatInWindow } from './path/to/BeatTracker.js';

const centerSample = 44100 * 10; // 10 seconds in
const result = await findNearestBeatInWindow(audioBuffer, centerSample);

console.log('Nearest beat sample:', result.nearestBeatSample);

Running the Example

Due to browser security restrictions on Web Workers and ES Modules (CORS), you cannot run the example directly from the file system (file://). You must serve it via a local HTTP server.

1. Navigate to the project root:

   cd /path/to/BeatTracker

2. Start a local server. For example, using Python:

   # Python 3
   python3 -m http.server 8000

   Or using Node.js http-server:

   npx http-server

3. Open your browser and visit: http://localhost:8000/example/

Project Structure

BeatTracker/
├── src/
│   ├── BeatsTracker.js       # Main entry point and logic
│   ├── SpectralFluxWorker.js # Web Worker for processing
│   └── fftv5.js             # FFT implementation
├── example/
│   ├── index.html           # Demo interface
│   └── app.js               # Demo logic
└── README.md

License

MIT