Refine input audio signals #55

New issue

Open

opened 2026-02-26 11:24:17 -08:00 by puregarlic · 0 comments

puregarlic commented

2026-02-26 11:24:17 -08:00

Owner

However, it's also worth investigating applying these optimizations locally, to see if we can get better or at least more customizable defaults. Possible options include sonora, a Rust port of the WebRTC audio improvement features, or perhaps seeing what features WASAPI provides directly.

Our audio streams are currently raw, and could be improved with echo cancellation, background suppression, and possible automatic gain control. These improvements are available out-of-the-box in WebRTC, and can be specified when creating the local audio track. However, it's also worth investigating applying these optimizations locally, to see if we can get better or at least more customizable defaults. Possible options include [sonora](https://github.com/dignifiedquire/sonora), a Rust port of the WebRTC audio improvement features, or perhaps seeing [what features WASAPI provides directly](https://docs.rs/wasapi/latest/wasapi/struct.AcousticEchoCancellationControl.html).