Treat continuous multi-speaker audio as a VAD safety-boundary case

A hard maximum utterance duration can falsely terminate a healthy streaming STT session when bridged participants overlap or background speech prevents the VAD silence boundary. Retain the cap because it protects memory and latency, but do not let it become a fatal call-level event: isolate or deliberately mix participant tracks, log active-turn age and audio rate, and close, segment, or reset an overlong turn while keeping the transport alive. Validate with simultaneous speakers and sustained noise, not only single-speaker clips.

Applies when: one STT stream receives audio from a transferred or multi-participant voice room. Keywords: STT VAD utterance timeout overlapping speakers LiveKit transfer audio