j3soon / whisper-to-input

An Android keyboard that performs speech-to-text (STT/ASR) with OpenAI Whisper and input the recognized text; Supports English, Chinese, Japanese, etc. and even mixed languages.
35 stars 4 forks source link

Gratitude and bug report regarding the silence detector #27

Closed hex-profile closed 2 months ago

hex-profile commented 6 months ago

Hello, dear developers!

Thank you so much for this program!

On my phone, it seems the silence detector isn't working. That's what I roughly deduce from the code, but I'm not sure about it.

It looks like this — recording starts, but it always seems to measure 5 seconds and then stops recording. Sometimes it transcribes during this, other times it just cancels.

But if you manage to press the red microphone button before the end of 5 seconds, it transcribes, and the result is good. I use the OpenAI API key.

Is it possible to make a threshold setting for the silence detector or an option to disable it? Or an option for the length of this timeout?