Web Audio Analyzer

A lightweight, dependency-free library for analyzing audio buffers in the browser. It extracts musical features (RMS energy, spectral brightness, texture) and robustly classifies audio as Reverb or Delay based on peak detection and density.

Installation

This package is intended to be used as a library.

# Install via npm/yarn (once published)
yarn add web-audio-analyzer

Usage

import { analyzeAudioBuffer } from 'web-audio-analyzer';

// Assuming you have an AudioBuffer (e.g., from AudioContext.decodeAudioData)
const audioBuffer = ...;

const result = analyzeAudioBuffer(audioBuffer);

console.log(result.classification);
// Output:
// {
//   type: 'Delay',
//   duration_category: 'Short',
//   brightness_category: 'Low',
//   texture_category: 'Energetic'
// }

console.log(result.analysis_data.musical_features);
// Output:
// {
//   rms_energy: 0.002,
//   peak_count: 4,
//   estimated_decay_time: '1.5s',
//   ...
// }

Features

• Robust Peak Detection: Uses prominence checks and adaptive thresholds to distinguish distinct delay taps from reverb reflections, even in sparse signals.
• Classification: Automatically categorizes audio into types (Reverb/Delay) and descriptors (Brightness, Texture, Duration).
• Spectral Analysis: Provides frequency band energy distribution and spectral centroid data.
• Zero Dependencies: Pure TypeScript implementation using standard Web Audio API structures.

Development

Setup

yarn install

Commands

• yarn dev: Start the development server (interactive demo).
• yarn build: Build the library for production (outputs to dist/).
• yarn test run: Run unit tests once.
• yarn lint: Run ESLint.
• yarn format: Format code with Prettier.
• yarn typecheck: Run TypeScript type checking.

CI/CD

This project uses GitHub Actions to ensure code quality. On every push and pull request to main, it runs:

• Linting & Formatting checks
• Type checking
• Unit tests (Vitest)

License

MIT