Copy link to clipboard
Copied
Hey Community,
Here is a workflow quandry for you - I will do this many times, so this would be a huge timesaver...
I ingest and create a multicam interview clip (two cameras, each with audio source, synced by audio) and then need to export to a low res h264 file to go to trint.com for transcription. Trint is a pretty great service that does solid automated transcription, and has a function to highlight (with different colors) and then shift highlighted sections around (called "stories") to make a very basic rough of my interview. You can export to XML or IDL which import just fine, but use the low res h264 for the video in a new sequence, with "highlighted" clips put together just like I want them.
But now, I have no way of getting my master multicam timeline to be clipped out like this. I thought of Synchronizing each clip (cannot do more than one) back to the the multicam timeline sequence, but this is time consuming and defeats the purpose of moving the highlighted clips around.
Does any one have any ideas on how to achieve this a workflow where I can export a single cam track (mostly important for transcribing audio) then import the edits back via XML/IDL? It seems like this is kind of a corner case, but I am so ready to be amazed by a wizard who knows how to get those edits from the xml or idl file back into the sequence I am already working on....
Much appreciation for any help!
Copy link to clipboard
Copied
Very interesting workflow. I have one question and maybe a solution depending on the answer:
When you import the XML and it links back to your lowres H.264 files, is it linking to TWO of them? i.e. it's recreating that edited version by cutting back and forth to the two different clips from the multicam?
If the answer is YES then try this: offline the H.264 files. Then Reconnect Media and manually choose the original camera clip that matches that angle. You may need to adjust the size of the sequence to match (if your H.264s were smaller) but as long as the frame rate and timecode matches between the camera originals and the H.264 versions then this *should* work.
If the answer is NO then... back to the drawing board!
JVK
Copy link to clipboard
Copied
Unfortuneatly, no, I am not linking two files. When I export the h264 files, it is just a good audio track and the ACAM interview angle. Thats the file that the Trint service takes in and transcribes.
I like where you are thinking - that went through my head to just offline, then relink to another file - but multicams are the way they are, so it wouldn't work
Yes, back to the drawing board. It is a tricky one, but I feel this could be valueable to others, as this isn't a too uncommon situation (multicam interview, export to transcription service, use transcription to get to rough, import back in...).
Copy link to clipboard
Copied
Hi there - Im in the exact same boat right now. Have you figured out a solution to this?
Copy link to clipboard
Copied
I had the same issue when exporting multicam sequences to Descript and bringing them back into Premiere. Here's the workflow I use:
Hope this helps!
Copy link to clipboard
Copied
thanks for the insight! I just tried out a program called Reduct that does the same transcription/paper edit features as Trint or Descript. But it has a specific workflow for multicam. I just gave it a test run and it worked really well. Will be assembling an hour long TV edit this way later this week.
Copy link to clipboard
Copied
It would be interesting to hear back about that, @corbettj25089153. Thanks!
Kevin
Copy link to clipboard
Copied
I have different approach, and it takes always 1 minutes no matter how many cuts i have (2, or 20 000).
I use same steps from 1 to 4
1. Create sequence of multicam and export it.
2. Import into software or online site for transcription and editing.
3. Export as XML.
4. Import XML into Premiere, wich will create bin with reference file and sequence, referencing to exported file(represent Multicam sequence)
Copy link to clipboard
Copied
I don't know how a multicam could possibly be sent via XML or EDL. Those need to work with basically flattened tracks. They cannot handle any complexity like multiple video tracks with cuts from one track to another.
So ... yea, this is a hard thing to do.