README.md

Generate TTS Audio

Requirements

• A local piper binary available in PATH (or set PIPER_BIN).
• A TTS model path passed via --voice, or the en_GB key in voices.json.

Install Piper

• macOS (Homebrew):
  • brew install piper-tts
• Linux (prebuilt):
  • Download the latest release from https://github.com/rhasspy/piper/releases
  • Extract and add the piper binary to your PATH
• Optional:
  • Set PIPER_BIN=/absolute/path/to/piper if the binary is not in PATH

Download Voices

1. Pick a voice/model from the Piper releases page (look for .onnx model files).
2. Download the .onnx file (and the matching .onnx.json metadata if provided).
3. Place the model somewhere on disk (e.g. ~/models/piper/).
4. Pass the model path to --voice or map it in voices.json.

Usage

npm run generate -- --manifest ../../assets/audio-manifest.json --out ../../sd-card/audio --voice ~/models/en_GB-alba-medium.onnx --force

The script writes WAV files to the output directory for every entry in the manifest.

Regenerate Existing Files

Pass --force to overwrite existing WAVs:

npm run generate -- --manifest ../../assets/audio-manifest.json --out ../../sd-card/audio --voice en_US --force