Native Recorder for Node.js

High-performance, low-latency native audio recording SDK for Node.js. Supports both microphone input and system audio capture (loopback) on Windows and macOS.

Features

• Microphone Recording - Capture audio from any input device
• System Audio Capture - Record what's playing on your computer (loopback)
• Cross-Platform - Windows (WASAPI) and macOS (AVFoundation + ScreenCaptureKit)
• High Performance - Native C++ implementation with minimal latency
• Prebuilt Binaries - No compilation required for most platforms
• Type Safe - Full TypeScript support

Platform Support

Platform Differences

Installation

npm install native-recorder-nodejs

Prebuilt binaries are available for most platforms. If a prebuild is not available, the package will compile from source (requires CMake and a C++ compiler).

Electron Support

This package includes prebuilt binaries for Electron 29-40. In most cases, it will work out of the box. If you encounter issues:

# Option 1: Use electron-rebuild (recommended)
npx electron-rebuild

# Option 2: Rebuild using cmake-js with Electron runtime
npx cmake-js compile --runtime=electron --runtime-version=YOUR_ELECTRON_VERSION --arch=x64

Note: Make sure to rebuild native modules after updating Electron version.

Quick Start

import { AudioRecorder, SYSTEM_AUDIO_DEVICE_ID } from 'native-recorder-nodejs';
import * as fs from 'fs';

const recorder = new AudioRecorder();

// List available devices
const inputs = AudioRecorder.getDevices('input');
const outputs = AudioRecorder.getDevices('output');

console.log('Microphones:', inputs);
console.log('Output devices:', outputs);

// Record from default microphone
const mic = inputs.find(d => d.isDefault);
if (mic) {
  const output = fs.createWriteStream('recording.raw');
  
  recorder.on('data', (buffer) => {
    output.write(buffer);
  });
  
  recorder.on('error', (error) => {
    console.error('Recording error:', error);
  });
  
  await recorder.start({
    deviceType: 'input',
    deviceId: mic.id
  });
  
  // Record for 5 seconds
  setTimeout(async () => {
    await recorder.stop();
    output.end();
    console.log('Recording saved!');
  }, 5000);
}

Record System Audio

// Get system audio device
const outputs = AudioRecorder.getDevices('output');
const systemAudio = outputs.find(d => d.id === SYSTEM_AUDIO_DEVICE_ID) 
                 || outputs.find(d => d.isDefault);

if (systemAudio) {
  await recorder.start({
    deviceType: 'output',
    deviceId: systemAudio.id
  });
}

API Reference

AudioRecorder

The main class for controlling audio recording.

Constructor

const recorder = new AudioRecorder();

Instance Methods

start(config: RecordingConfig): Promise<void>

Starts recording from the specified device.

interface RecordingConfig {
  deviceType: 'input' | 'output';  // Required
  deviceId: string;                 // Required
}

await recorder.start({
  deviceType: 'input',
  deviceId: 'device-uuid'
});

stop(): Promise<void>

Stops the recording session.

await recorder.stop();

Static Methods

getDevices(type?: DeviceType): AudioDevice[]

Lists available audio devices.

interface AudioDevice {
  id: string;           // Unique identifier
  name: string;         // Human-readable name
  type: 'input' | 'output';
  isDefault: boolean;
}

const allDevices = AudioRecorder.getDevices();
const microphones = AudioRecorder.getDevices('input');
const outputs = AudioRecorder.getDevices('output');

getDeviceFormat(deviceId: string): AudioFormat

Gets the audio format of a device.

interface AudioFormat {
  sampleRate: number;   // e.g., 48000
  channels: number;     // 1 (mono) or 2 (stereo)
  bitDepth: number;     // Output bit depth (16)
  rawBitDepth: number;  // Native device bit depth
}

const format = AudioRecorder.getDeviceFormat(device.id);

checkPermission(): PermissionStatus

Checks current permission status.

interface PermissionStatus {
  mic: boolean;     // Microphone permission
  system: boolean;  // System audio permission
}

const status = AudioRecorder.checkPermission();

requestPermission(type: 'mic' | 'system'): boolean

Requests permission for recording.

const granted = AudioRecorder.requestPermission('mic');
const systemGranted = AudioRecorder.requestPermission('system');

Events

'data'

Emitted when audio data is available.

recorder.on('data', (buffer: Buffer) => {
  // Raw PCM 16-bit LE audio data
});

'error'

Emitted when an error occurs.

recorder.on('error', (error: Error) => {
  console.error(error.message);
});

Constants

// Special device ID for system-wide audio capture on macOS
export const SYSTEM_AUDIO_DEVICE_ID = 'system';

Audio Format

The output is always:

• Format: Raw PCM
• Bit Depth: 16-bit signed integer
• Endianness: Little Endian
• Sample Rate: 48kHz on macOS (fixed), native device rate on Windows (commonly 44.1kHz or 48kHz)
• Channels: Stereo on macOS (fixed), preserved from source on Windows

Playing Raw Audio

You can play the recorded raw audio using FFplay:

ffplay -f s16le -ar 48000 -ch_layout stereo recording.raw

Or convert to WAV using FFmpeg:

ffmpeg -f s16le -ar 48000 -ac 2 -i recording.raw output.wav

Permissions (macOS)

On macOS, you need to grant permissions:

1. Microphone - Required for input device recording
2. Screen Recording - Required for system audio capture

The SDK will automatically prompt for permissions when needed, or you can request them programmatically:

const permissions = AudioRecorder.checkPermission();

if (!permissions.mic) {
  AudioRecorder.requestPermission('mic');
}

if (!permissions.system) {
  AudioRecorder.requestPermission('system');
}

Error Handling

Building from Source

Prerequisites

• Node.js 16+
• CMake 3.15+
• C++17 compiler
  • Windows: Visual Studio 2019+ or MSVC Build Tools
  • macOS: Xcode Command Line Tools

Build Commands

# Install dependencies
npm install

# Build native module
npm run build:native

# Build TypeScript
npm run build:ts

# Build everything
npm run build

# Run tests
npm test

Publishing

# Build prebuilt binaries for current platform
npm run prebuild

# Upload prebuilts to GitHub Release (requires GITHUB_TOKEN)
npm run prebuild:upload

# Publish to npm (prebuilts should be uploaded first)
npm publish

CI/CD with GitHub Actions

The project uses GitHub Actions for automated builds and releases. The workflow is triggered by:

To release a new version:

# 1. Update version in package.json
npm version patch  # or minor, major

# 2. Push the tag to trigger the release workflow
git push origin main --tags

The CI will automatically:

1. Build native modules for all platforms (macOS arm64/x64, Windows x64/ia32)
2. Run tests on each platform
3. Publish the package to npm with prebuilt binaries
4. Create a GitHub Release with the native binaries attached

Note: Ensure NPM_TOKEN is configured in repository secrets for npm publishing.

Architecture

┌─────────────────────────────────────────────────┐
│              Node.js Application                 │
├─────────────────────────────────────────────────┤
│              TypeScript Wrapper                  │
├─────────────────────────────────────────────────┤
│              N-API Binding Layer                 │
├────────────────────┬────────────────────────────┤
│   Windows          │         macOS              │
│   (WASAPI)         │   (AVFoundation +          │
│                    │    ScreenCaptureKit)       │
└────────────────────┴────────────────────────────┘

For detailed architecture documentation, see docs/architecture.md.

For complete API documentation, see docs/api.md.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

MIT License - see LICENSE for details.

Acknowledgments

• node-addon-api - N-API C++ wrapper
• cmake-js - CMake build system for Node.js addons