Import auto captions into markers

[Fly Line Prods]

I'm loving Premiere's new auto caption feature, and I've found it to be surprisingly accurate.

In addition to creating auto captions for my final edited product, another use of this technology that would be super helpful to my workflow would be to auto transcribe my raw interview footage and have them placed into markers.  Other than manually cutting and pasting from the Captions section, is there a way to easily do this?

I'm currently running Premiere 22.0.0

Thanks!

[Mike Dziennik]

Premiere currently doesn't have the ability to IMPORT markers, only export them.

Once your captions are generated, you can export an SRT. It should be fairly trivial to reformat it into a marker or CSV document. You will need to use a third-party extension/plugin to be able to import markers back into premiere. There is a free one called Markerbox but I haven't used it so can't comment on its efficacy.

[Fly Line Prods]

Great suggestion, however... I downloaded Markerbox, installed it as an extension, but then when I tried to run it, nothing happened.  I checked the support boards and several people were reporting this issue.  I emailed the app developer and he said that several people have reported that problem on Premiere 2022, and his suggestion was to use Premiere 2020 -- which is not an option for me.

[Mike Dziennik]

That's a shame.
You can download and install previous versions of Premiere from the
Creative Cloud App (and run them alongside your current version). Can you
do that, simply to get the markers imported...then open that file in your
newer version?

[Fly Line Prods]

Just to follow up on this post from last week, I ended up addressing the problem by slightly changing my workflow. 

What I did was string out the entire interview on a timeline, auto caption the entire thing, and then I had a timecoded transcript.  The only problem was that when I extracted chunks of the interview I won't use, the new timeline timecode in the transcript did not update.  So I simply started lifting the segments instead of extracting them, so the timecode wouldn't change.  Then once I'd eliminated a significant portion of the interview footage (usually more than half on my first pass), I extracted the gaps from the timeline and then re-auto captioned it.  Feels like a bit of a low tech solution, but it has ended up saving me time so far.