Skip to main content

Audio Plugins - Voice Interaction Support

· 2 min read
Mark Dierolf
Mark Dierolf
Creator of TokenRing AI

TokenRing Coder now supports voice interaction with audio recording, playback, transcription, and text-to-speech.

Audio Plugin​

Abstract audio framework defining core capabilities:

Features​

  • Recording: Capture audio from microphone
  • Playback: Play audio files through system audio
  • Transcription: Convert audio to text using AI services
  • Text-to-Speech: Generate speech from text
  • Provider Architecture: Abstract interface for platform-specific implementations

Chat Commands​

/voice record    # Record audio
/voice transcribe audio.wav  # Convert to text
/voice speak "Hello world"   # Generate speech
/voice playback audio.mp3    # Play audio

Linux Audio Plugin​

Linux-specific implementation using naudiodon2:

Features​

  • Native Performance: Uses naudiodon2 for efficient audio I/O
  • OpenAI Integration: Whisper for transcription, TTS for speech
  • Format Support: WAV for recording/playback, MP3 for TTS
  • ALSA Support: Works with Linux audio system

Usage​

const provider = new LinuxAudioProvider({
  sampleRate: 48000,
  channels: 1,
  format: 'wav'
});

const audioService = new AudioService();
audioService.registerProvider('linux', provider);

// Record and transcribe
const recording = await audioService.record(abortSignal);
const transcription = await audioService.transcribe(recording);

// Text-to-speech and playback
const audioFile = await audioService.speak('Hello, world!');
await audioService.playback(audioFile);

Use Cases​

  • Voice commands for hands-free coding
  • Transcribe meeting notes and discussions
  • Generate audio documentation
  • Accessibility features for visually impaired developers

Voice interaction makes AI assistance more natural and accessible.

Mark Dierolf
Creator of TokenRing AI