Nostr'a Katılın
2026-07-03 18:11:11 UTC

Rain on Nostr: Decided to do another full smoke test on all of [audio.cpp]() backends for ...

Decided to do another full smoke test on all of [audio.cpp](https://github.com/0xShug0/audio.cpp) backends for "voice-cloning" TTS. I do find longer gen times and quality crops with audio.cpp vs qwentts.cpp on 1.7B but maybe it is my settings. Either way pretty nifty and I think a suitable for my voice cloning/making front end I have been noodling with