Starstruk Logo MarkStarstruk

v0.3.103

KoljaB/RealtimeSTTv0.3.103Apr 19, 2025by KoljaB

AI Summary

This release brings major infrastructure improvements including thread-safe IPC with SafePipe, audio normalization to -0.95 dBFS, and a complete callback overhaul to run asynchronously via helper threads, along with wake word and VAD enhancements.

Key Highlights

  • Thread-safe IPC with SafePipe replacing mp.Pipe
  • Audio normalization to -0.95 dBFS for consistent transcription quality
  • All callbacks now run asynchronously via helper threads
  • Wake word & VAD improvements with faster_whisper_vad_filter flag
  • Rich metadata with nanosecond-precision timestamps

New Features

  • SafePipe for thread-safe inter-process communication
  • normalize_audio option
  • Async callback execution via helper threads
  • wakeword_backend configuration
  • faster_whisper_vad_filter flag
  • Nanosecond-precision timestamps
  • --faster_whisper_vad_filter and --debug_websockets CLI flags

Full Release Notes

# RealtimeSTT 0.3.103

### Features & Improvements
- **Thread‑safe IPC**: Introduce `SafePipe` to replace `mp.Pipe`, hopefully ensuring robust inter-process communication (needs more tests).
- **Audio normalization**: New `normalize_audio` option scales input to –0.95 dBFS for consistent transcription quality.
- **Callback overhaul**: All event callbacks (VAD, wake‑word, turn detection, recording, realtime updates) now run asynchronously via helper threads.
- **Wake word & VAD**: Add `wakeword_backend` config and `faster_whisper_vad_filter` flag; improved error messages when misconfigured.
- **Rich metadata**: Embed nanosecond‑precision timestamps in both client and server, serialized as formatted strings.
- **CLI enhancements**: `--faster_whisper_vad_filter` and `--debug_websockets` flags give finer control over server behavior.
- **Testing updates**: Adjusted parameters in `realtimestt_test` and added a new `type_into_textbox.py` example.