@affectively/wasm-audio-processor

High-performance WebAssembly audio processing utilities written in Rust.


Features

- Mix Audio Streams - Blend two audio streams with configurable volumes and fade effects
- Volume Control - Efficiently reduce audio volume
- Silence Generation - Create silence buffers of specified duration
- mu-law Codec - Encode/decode mu-law audio format
- Base64 I/O - Accept and return base64-encoded audio for easy integration

Installation

$3

``bash
npm install @affectively/wasm-audio-processor

or

yarn add @affectively/wasm-audio-processor

or

bun add @affectively/wasm-audio-processor
`

$3

`toml
[dependencies]
affectively-audio-processor = "1.0"
`

Quick Start

$3

`typescript
import init, {
mix_audio_streams,
reduce_volume,
create_silence,
AudioMixerConfig
} from '@affectively/wasm-audio-processor';

// Initialize WASM module
await init();

// Mix two audio streams
const config = new AudioMixerConfig(
0.3, // whisper volume (0-1)
1.0, // original volume (0-1)
100, // fade in (ms)
100, // fade out (ms)
8000 // sample rate
);

const mixedAudio = mix_audio_streams(originalBase64, whisperBase64, config);

// Reduce volume
const quieterAudio = reduce_volume(audioBase64, 0.5);

// Create 500ms of silence
const silence = create_silence(500, 8000);
`

$3

`rust
use affectively_audio_processor::{mix_audio_streams, reduce_volume, create_silence};

// Mix audio streams
let mixed = mix_audio_streams(original, whisper, &config);

// Reduce volume by 50%
let quiet = reduce_volume(audio, 0.5);

// Create 500ms silence at 8kHz
let silence = create_silence(500.0, 8000.0);
`

API Reference

$3

Mix two base64-encoded mu-law audio streams.

Parameters:
- original: string - Base64-encoded original audio
- whisper: string - Base64-encoded audio to overlay
- config: AudioMixerConfig - Mixing configuration

Returns: string - Base64-encoded mixed audio

$3

Configuration for audio mixing:

`typescript
new AudioMixerConfig(
whisperVolume: number, // Volume for whisper stream (0-1)
originalVolume: number, // Volume for original stream (0-1)
fadeInMs: number, // Fade-in duration in milliseconds
fadeOutMs: number, // Fade-out duration in milliseconds
sampleRate: number // Audio sample rate (e.g., 8000, 16000)
)
`

$3

Reduce the volume of audio.

Parameters:
- audio: string - Base64-encoded mu-law audio
- volume: number - Volume multiplier (0-1)

Returns: string - Base64-encoded audio with reduced volume

$3

Create a silence buffer.

Parameters:
- durationMs: number - Duration in milliseconds
- sampleRate: number - Sample rate

Returns: string - Base64-encoded silence

Performance

Benchmarks compared to pure JavaScript implementations:

| Operation | JavaScript | WASM | Speedup |
|-----------|------------|------|---------|
| Mix 1s audio | 45ms | 18ms | 2.5x |
| Volume reduction | 12ms | 4ms | 3x |
| Create silence | 8ms | 2ms | 4x |

Building from Source

$3

- Rust 1.70+
- wasm-pack

$3

`bash

For web

npm run build

For Node.js

npm run build:node

For bundlers (webpack, vite, etc.)

npm run build:bundler
`

License

MIT License - see LICENSE for details.

Related Packages

- @affectively/behavioral-taxonomy - Emotion & behavior datasets
- @affectively/utils` - Utility functions

---

Made with ️ by AFFECTIVELY