Transcription in "Titles to Keywords Toolbox"

[samplue]

The titles to keywords toolbox could benefit from transcription.

Chris Hocking: Adding transcription is too much work. Whisper is amazing, but requires far too many post-fixes. Better to decide on a tool like Transcriber and build upon that. Maybe you export your Transcriber titles into FCPX on your Titles to Keywords timeline and we do something with it?

[latenitefilms]

@samplue - The question is... what do you want it to do, and how would you want it to work? You could quite easily add transcribed titles to your Titles to Keywords Timeline today using Transcriber (or any other 3rd party tool) - but what would we actually do with that data?

If you can give specific examples - ideally manually created before and afters (including screenshots and FCPXMLs) - that would be very helpful.

For example, do you want to convert those transcriptions into markers/notes like Lumberjack Builder and Simons Says does, so you can search for text in the timeline? Or do you just want to create Keywords for each line of dialogue, etc?

[samplue]

I'd want to speed up our "Titles to Keywords" workflow.

One way is to get the transcription back into the timeline with one title per line per speaker. The titles would have different colors based on the speaker. This would already speed up adjusting the "Titles to Keywords" titles as it's a more visual approach. I need to check if Transcriber has speaker detection, this would bring us a lot closer.

The main idea is getting a kind of Script Sync workflow, where you feed in a script and the transcription is checked against that script. The outcome would still be line-by-line keywords as we currently use them with "Titles to Keywords".