https://github.com/newfolder6/vctranslator
a demo for using vosk to provide live subtitles for my friends when I am speaking Chinese with another guy in the same VC
https://github.com/newfolder6/vctranslator
fastify nodejs vosk
Last synced: about 1 month ago
JSON representation
a demo for using vosk to provide live subtitles for my friends when I am speaking Chinese with another guy in the same VC
- Host: GitHub
- URL: https://github.com/newfolder6/vctranslator
- Owner: NewFolder6
- Created: 2025-02-17T08:35:01.000Z (8 months ago)
- Default Branch: master
- Last Pushed: 2025-03-06T07:27:59.000Z (7 months ago)
- Last Synced: 2025-03-15T13:16:48.668Z (7 months ago)
- Topics: fastify, nodejs, vosk
- Language: JavaScript
- Homepage:
- Size: 27.2 MB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# VCTranslator
Real-time audio translation to subtitles using Vosk speech recognition.
## Setup Requirements
### Prerequisites:
1. **Node.js**: Version 16.x or 18.x recommended
- **IMPORTANT**: Node.js 22+ has known compatibility issues with native modules
- If using Node.js 22+, see the "Compatibility Mode" section below
2. **Windows Users**:
- Visual C++ Build Tools (install "Desktop development with C++" workload from Visual Studio Installer)
- Python 3.x (added to PATH)
Alternatively, you can install the build tools via npm:
```
npm install --global --production windows-build-tools
```
3. **Audio Device**: A working microphone or audio input device
4. **Speech Recognition Model**:
- After installing dependencies, run: `npm run download-model`
- This will automatically download and extract the Vosk speech recognition model
### Installation:
```bash
# Clone the repository
git clone https://github.com/yourusername/VCTranslator.git
# Navigate to the project directory
cd VCTranslator
# Install dependencies
npm install
# Download speech recognition model
npm run download-model
```
## Running the Application
### Standard Mode (Node.js 16-18 recommended):
```bash
npm start
```
### Compatibility Mode (for Node.js 20+):
If you're using Node.js 22 or newer and encounter native module errors:
```bash
# Run with compatibility flags
npm run start:compat
# Or run with native modules disabled (limited functionality)
npm run start:safe
```
## Audio Loopback Setup
The project uses naudiodon for audio input/output functionality on Windows and other platforms.
### Requirements:
- For audio recording/loopback functionality, you need to have SoX installed:
- On Windows: Download from [SoX website](https://sourceforge.net/projects/sox/) and add to PATH
- On macOS: `brew install sox`
- On Linux: `apt-get install sox` or equivalent
### Usage:
```javascript
const { AudioLoopback } = require('./src/audio-wrapper');
// Create a loopback instance
const loopback = new AudioLoopback({
// Customize options if needed
sampleRate: 44100,
channels: 2,
deviceId: 0 // Use 0 for default device or get available devices with loopback.getDevices()
});
// Start audio loopback
const stream = loopback.start();
// When finished
loopback.stop();
// To get available audio devices
const devices = loopback.getDevices();
console.log('Input devices:', devices.input);
console.log('Output devices:', devices.output);
```
## Troubleshooting
If you encounter issues during installation or runtime:
1. **Node.js Version**:
- The recommended solution is to downgrade to Node.js 16.x or 18.x
- Run `nvm install 18` if you use nvm
- Or download from [nodejs.org](https://nodejs.org/)
2. **Native Module Errors**:
- Try running with compatibility mode: `npm run start:compat`
- For testing without native modules: `npm run start:safe`
- Make sure Visual C++ Build Tools and Python are installed
3. **ffi-napi Errors**:
- Run `npm run postinstall` to attempt an automatic patch
- If that fails, try a clean install: `rm -rf node_modules && npm install`
4. **Audio Device Access**:
- Ensure your application has permission to access audio devices