Separating speakers when Premiere doesn't recognize them

[TiaM213]

I have a short clip with two speakers.  Premiere's transcription is doing fairly good understanding the words, but it doesn't recognize that there are two people speaking.  I have the feature turned on to where it is supposed to recognize them automatically, but the voices are pretty similar, so I suppose it isn't distinguishing a difference.  It puts all the lines on one speaker.  I just need to know how to tell it where the second person starts talking, how to separate what it sees as one line into two speakers.

[mattchristensen · Accepted solution]

@TiaM213 Make sure you're looking at the Source clip transcript, and then you can do the following:

1. Put your cursor somewhere at the point where the other person starts talking.
2. Click the "Split Segment" button found above the transcript
3. In the newly created segment click on the "..." next to the Speaker name and change the speaker there

[TiaM213]

Thank you!

[t_8601]

What if it is only transcribing in monitor? I dont realize why it wont show in source

[Eucatastrophe]

I gave this a try. While it allows you to alter the speaker by inserting a placeholder, updating the speaker in one section actually causes the system to overwrite the speaker tags I had already adjusted manually.

[Stan Jones]

@Eucatastrophe,

> While it allows you to alter the speaker by inserting a placeholder, updating the speaker in one section actually causes the system to overwrite the speaker tags I had already adjusted manually.

I'm not sure what you mean by this, and I suspect we all need to be clear how speaker naming/editing works. I have not paid enough attention to this previously. Creating a new speaker name does not appear to assign it to the transcript segment you selected it from. And changing a speaker name (i.e. changing Speaker 1 to something else) will change that speaker name for EVERY segment that has it.

Assume a transcript with only Speaker 1, and that it is a Source Media transcript, so I must be in Source Monitor view for that transcript to edit the speaker. (If you have a STATIC sequence transcript it works the same way, just in program monitor view.)

Click on the speaker name for any transcript segment, pick "Edit speaker name," and change that speaker name from Speaker 1 to Thomas. Because we have EDITED a current speaker name, it changes every instance of Speaker 1 to Thomas.

Click on the speaker name for any transcript segment (all now Thomas), pick "Edit speaker name," pick "Add speaker" - it shows Speaker 2 by default - and change that to "Joseph." There is no change to any segment speaker names. But now, when you click on any of the speaker names, you see a choice of Joseph, Thomas, and Edit speaker name.

If you click a segment and "split" at that point, it creates two segments, both with Thomas as the speaker. Click the "Thomas" for the second of those 2 segments and pick Joseph to make Joseph that speaker.

If you already made a bunch of changes, be sure you are picking from your list, not making new names when you already have a name for one of the speakers.

Stan