ich habe ein Video (1,5std) worin 5 verschiedenen Menschen auf abwechselnd deutsch und englisch reden.
Ich kann aber immer nur entweder das deutsche oder das englische Transkript erstellen, das zuvor erstellte wird dann immer überschrieben, somit hab ich nie ein komplettes Transkript des ganzen Videos auf den richtigen Sprachen.
Perhaps it can be done by exporting the transcripts as .txt or .csv, but first I looked at working with the captions. If I were exploring those options, I think I would export csv, use the speaker designations to delete one language or the other, then use macros to merge csv's. But here's one method.
First, transcribe English with the selection "Recognize when different speakers are talking." Use the 3 dots "overflow menu" and export .txt and/or .csv so you have a reference later.
Delete all the transcriptions with speakers in German. (Are you kidding me? Well, I did say it was labor intensive.) The transcription is actually all English. But some of it will be odd because it is German speech transcribed as if it were English. The more a speaker talks at once, the fewer edits will be required. And because the transcription (unlike captions) are generally longer sections, it may be more easily recognizable as "German transcribed as English." And once you identify a speaker as English or German, you can just look for that speaker. You can "edit speakers" to rename. Or you can delete a speaker in the rename dialogue, but it only makes the speaker name "N/A." But nothing actually deletes that speaker's text.
Other than being labor intensive because of length, there is no way to directly delete a transcript timecode section, other than to delete all the text of that section. Double click in the text. You will be in edit mode, with the text in an edit box and one word selected. Triple click and all text will be highlighted. I also found that simple left-clicking 5 times selected all the text. After selecting all the text, hit the Delete key. The text is gone. Click outside that text, and you will see "<Type your caption here>" That is the same as an empty transcript section and will generate no captions.
Click CC and generate the captions for that language.
You can run through the timeline and identify captions misidentified as English that are actually German. Go back to the transcript and delete them. Delete the caption track and regenerate the captions.
In the timeline, turn off that caption track. I would also rename it for convenience (right click on the "Subtitle" name).
Go back to the transcript tab, 3 dots, "re-transcribe." Do German. Delete entries for English. Create captions. Rename the caption track for convenience.
You now have 2 caption tracks, one with one language and one with the other.
You could now export srt for each track and modify there. But I think the following method will be easier.
"Duplicate" the sequence in the Project Panel to create a backup copy.
In the main sequence, Drag to select all the captions in the lower (C1) track. Copy. Set the CTI to the beginning of that first caption. Delete that C1 caption track. C2, the other caption track now becomes C1, and when you paste, the captions merge in that one track.
Hey 🙂 , thanks for this detailed answer. Sorry for my late reply. Your description helped and I was able to do it that way :). Nevertheless, it's a pity that there is no easier way at Premiere. Thanks again and greetings 🙂