@zoran
Hello.
Unless I have overseen something it currently doesn’t seem to be possible to generate timestamps from a YouTube video if the text has not been generated via Whisper transcript. This is the case if there is already a transcript available which I prefer to use over the ai generated one, as the latter sometimes contains errors or doesn’t reflect certain formatting aspects, as they cannot be heard.
My current workflow to circumvent this is the following.
- Add the transcript as base of the new lesson
- Add the YouTube video link
- Add the audio version of the Video, which I have to create beforehand
- create timestamps based on the audio
- delete the audio from the lesson afterwards
The last step is necessary because otherwise the underlining of the currently spoken text will only work with audio, but not with video playback. I think it is needless to say that this is a cumbersome approach.
I would therefore like to ask whether it could be made possible to create the timestamps from the video (without transcribing the video!).