Voice Conversion Microphone Processor

A powerful Electron app that creates a virtual microphone with real-time audio processing capabilities.

Features

• Microphone Detection: Automatically detects all available audio input devices
• Visual Selection: Click any microphone to select it as the active device
• Virtual Microphone: Creates a virtual audio device that processes the selected microphone's audio stream
• Real-time Processing: Audio processing pipeline for voice conversion and effects
• Default Selection: Automatically selects the system default microphone on first run
• Persistent Settings: Remembers your microphone selection across app restarts
• System Tray Integration: Lives in your system tray for quick access
• Modern UI: Clean, native-looking interface with smooth animations and virtual mic controls

Installation

1. Clone the repository:

   git clone https://github.com/sundai-club/voice-conversion.git
   cd voice-conversion

2. Set up environment variables:

   cp .env.example .env
   # Edit .env and add your API keys:
   # - SONIOX_API_KEY: Your Soniox real-time transcription API key
   # - ELEVENLABS_API_KEY: Your ElevenLabs text-to-speech API key

3. Install dependencies:

   brew install blackhole-2ch # reboot after this one!
   npm install

4. Run the app:

   npm start

Usage

• Click the microphone icon in your system tray to open the microphone list
• Click any microphone in the list to select it as your active device
• Click "Start Virtual Microphone" to begin processing audio from the selected device
• The virtual microphone will appear as a new audio input device in your system
• Your selection is automatically saved and restored when you restart the app
• Use the refresh button (↻) to reload the microphone list if devices change

Virtual Microphone

The app creates a virtual audio pipeline that:

• Captures audio from your selected physical microphone
• Processes the audio stream in real-time (voice conversion, effects, etc.)
• Routes the processed audio to a virtual audio device
• Makes the processed audio available as a system-wide microphone input

Virtual Audio Device Setup

To use this as a system-wide virtual microphone, you need to install a virtual audio driver:

macOS:

• BlackHole (Free): Download from GitHub
• Loopback (Paid): Professional audio routing tool

Windows:

• VB-Audio VoiceMeeter (Free): Virtual audio mixer
• Virtual Audio Cable (Free): Simple virtual audio routing

Linux:

• PulseAudio: Built-in virtual sink support
• JACK: Professional audio routing system

See VIRTUAL_AUDIO_SETUP.md for detailed installation instructions.

Development

• npm start - Start the Electron app
• npm run dev - Start in development mode (if available)

Requirements

• Node.js 14+
• Electron-compatible operating system (macOS, Windows, Linux)
• Microphone permissions granted to the application

License

MIT License