ByteBolt prototype dialogue audio

12 lines synthesized entirely on the local computer using three newly designed original voices. These are NOT the existing approved Robert, Iris or Evelyn voice identities. Character names refer only to game roles. No reference recordings, private voice models, speaker embeddings or training assets are included.

The input text is preserved exactly in manifest.json. Speech pronunciation may vary for technical identifiers. WAV files are mono PCM 16-bit with level normalization and peak limiting by gain only.

Backend/model: Qwen3-TTS-12Hz-1.7B-VoiceDesign, Apache 2.0 per installed third-party notices. Application: Kira Labs Voice Designer, MIT. No independent legal audit or guarantee of exclusive rights in generated speech is claimed. Retain the bundled notices when redistributing relevant software/model components; none are included here. Disclosure as synthetic prototype dialogue is recommended.

Best-effort speaker continuity; human listening review recommended before public release.
