Timestamps of each chunk

[dgoryeo]

Hi @flutydeer ,

Thank you for this package. Would it be possible to store / retrieve the timestamps (start - end in miliseconds) of each sliced chunk?

Best regards

[flutydeer]

Yes, but why do you need this feature? If you want to edit markers in other applications, you can try OpenVPI's dataset-tools. It can save slice markers into a csv file, which can be loaded by Audition.

[dgoryeo]

Thanks @flutydeer . My use case is to: slice audio to chapters, transcribe each chapter as subtitles, join/stitch the subtitiles of each chapter to have complete and correct timetsamps for entire audio. I think I should be able to do that by using the .csv markers file.

[dgoryeo]

Hi @flutydeer , can the python code be used to same the chunks and the markers --i.e non-gui code.

[flutydeer]

Python code in this repo does not support markers.

[dgoryeo]

@flutydeer , I just tried the OpenVPI's dataset-tools to save audio chunks and markers. Couple of quuck questions:

• are the markers in number of frames, or miliseconds?
• what does minimum interval mean?

Thanks

[flutydeer]

1. frames.
2. Please check out README in this repo.

[dgoryeo]

Thank you @flutydeer !