The "segment" time is wrong from the real time when the silence is removed

[NguyenDucLamK63]

I run the same as you, the results are good, but the "segment" time is wrong compared to the real time of the audio. So is there a way when I run the code, the silence is still removed, but the text segment time is still the same as the audio time? Hope you can help. Thank you very much

[EtienneAb3d]

@NguyenDucLamK63

1. use WhisperHallu options to produce both transcriptions, with and without silence/noise removal.
2. use WhisperTimeSync to send the good timestamps over the good text.
   https://github.com/EtienneAb3d/WhisperTimeSync