Any idea what software does chrome use with SpeechSynthesis js API? I mean the built-in one powering this demo. I would like to capture synthesized voice into .wav (or other format), in my web app, but seems very inconvenient via available javascript apis. I was thinking if chromium uses open source software here, I could compile it into wasm...
I found some references to espeak-ng, however espeak-ng output quality is much worse compared to SpeechSynthesis js. There is also PATTS mentioned or chrome-tts-blaze but seems outdated and links broken.
I also compared quality with flite using available .flitevox voices, but quality not good either.
