Speech to Text for Previously Recorded Audio

Is something like the Whisper app suitable to make a transcription of previously recorded audio? It sounds like it would work with Talk, but I have meetings that have been recorded on Zoom that I would like to transcribe. Would that be accessed somehow via the smart picker? I’m not sure where to find out more about whether this use case is supported.


I figured out how to do this.

You install the Whisper Speech-To-Text app, in addition to the Speech to Text Helper App. Then you have to go into the Talk app, and use the picker ("/') to select speech to text, then instead of recording a file, select a file. Then you have to wait for the notification with the result. In the meantime, the admin status page for the Whisper app will show that there are no requests queued, although you will note that the CPU is grinding away actually doing the work.

Interestingly, I might have expected a file menu item to do initiate this function, rather than going to Talk. It won’t matter which conversation you choose, because the transcription doesn’t seem to affect the conversation at all, or actually have anything to do with Talk at all. It’s just the place where you have to make the request.

I might be wrong, so YMMV. I haven’t really found any decent documentation on this process yet.

1 Like

This topic was automatically closed 8 days after the last reply. New replies are no longer allowed.