IMPORT CORRECTED TRANSCRIPT

[Exciting_magic5735]

I'm trying to upload a corrected transcript file to PR in order to create both a .scc and .srt file for captions. I have a .doc file, with timecodes matching the video. Adobe rep told me to convert to .txt file, but can't get the file to import. 

Right now, I'm having to correct the AI transcript line by line to match the corrected file, which is taking an absurdly long time. 

I know this question has come up a lot in the past, was hoping maybe Adobe has figured it out - any insights on this one? 

[derekwatson]

I would export theSRT from PP and open both in Word and do a file comparison, it should be much faster to check text then re-import the saved updated SRT file 

[Stan Jones]

@Exciting_magic5735,

Without knowing the format of each file, it is a bit of a guess.

Are you saying that you have an edited transcript in Word (.docx) form? Or is it more like captions in Word form?

I agree with @derekwatson that it may be simply to just work with fixing the captions. However, if you have long segments (as in the transcript), I'd opt for the following method.

I just tested this in the latest Beta (25.4.0.25), and it is still working. You create a version of your edited transcript with no timecodes, no speakers, no line breaks etc. I do this with Notepad++ macros, and it takes only a few seconds. We'd have to figure out what format your Word document is actually in.

When "imported as corrected," PR looks at its text timing, not the segment timecodes, and fits the words in. Almost by magic, the original segments are the same, but with the edits.

This will NOT work, if the edits are too extensive. It must be pretty much exactly what was said in the original audio.

I'll send you a direct message.

Stan

[Stan Jones · Accepted solution]

@Exciting_magic5735,

This will be challenging no matter how you proceed. There are 2 speakers and the dialogue overlaps a lot. And I suspect they are speaking rapidly enough that it is hard to meet the standards for accessible captions.

I would label each speaker's captions with the name. I might vary the position, particularly if they are on different sides of the screen. PR is getting there, but currently, a) you cannot carry the speaker names from the transcript to the captions, b) srt export from PR carries no positioning information, and c) even though multiple caption streams can now be enabled on one timeline, you cannot export more than one in an SRT sidecar. I would use Subtitle Edit or similar. You can import your text file to get the text in, and then time it in the app. If you have a transcript in PR, you can create captions and import that timing with your other text file.

Stan