Crate atomr_agents_tts_audio 

Output-side audio I/O for atomr-agents text-to-speech.

Mirrors atomr-agents-stt-audio for the inverse direction: PCM -> container encoders, and (under speaker) a cpal-based speaker-output stream that mirrors MicCaptureSession.

Two independent feature surfaces:

• encode-wav (default-on): hound-backed WAV writer.
• speaker (default-off, needs libasound2-dev on Linux): cpal output stream with a bounded mpsc producer queue.

Reuses AudioFormat, PcmBuffer, SampleType from stt-core so the two stacks share one vocabulary.

Modules

encode

PCM -> container encoders.