Transcript-Based Lip Sync; Error Message

[woodya9882332]

I've been using the transcript-based lip sync (following the process from the Okay Samuria YouTube video). Overall, it works great but I keep running into the following error message:

"compute lip sync failed: check that the transcript matches the audio"

The transcript does tend to correctly display the words from teh audio but I get the error message regardless. Does anyone know how to fix this? Or know why I'm getting the error message since the words and audio appear to correctly match?

Thanks!

[cwest222111]

Going through the same thing, does anyone know anything?

[Mittensbrains]

I was having the same issue.  My first audio with .SRT file worked perfectly.  But the next two were getting this error: "compute lip sync failed: check that the transcript matches the audio"

I looked at the imported SRT file and noticed that the times were not starting from zero.  So I went back to the Premiere Pro file, placed all the audio and transcript at the beginning of the timeline THEN exported the SRT file ONLY again.  When I imported this new SRT file into Character Animator and hit Timeline > Compute Lip Sync Take from Audio and Transcript it finally worked.

I guess you have to have the SRT file starting from 0:00:00.  Otherwise Character Animator doesn't know what to do with it.

[David25727707wpd6]

Yeah, it's maddening. I run into all of the time. Like right now, for instance. Makes no sense. Works great for a few sentences. Forgets everything it knows about English in the next sentence. Comes back to its senses and works fine. Loses its mind again. I waste sooo much time having to manually fix it. I really wish Adobe would actually speak to these issues. It's great to have a forum for checking with each other. But I am customer paying for a product that does not always work and I would like to hear from the vendor about how they are addressing these deficiencies. It frustrates me to no end when companies refuse to acknowledge their customers legitimate problems. "We apologize for the inconvenience" only goes so far in helping, folks.

[imaginizer]

This feature is not ready for prime-time. I've found it best to ignore using the transcript, becasue when the software gets confused, it decides not to try the lip sync at all. Instead of falling back to the audio for lip sync, it just chooses to do NO lip sync. The feauture creates more work, because unlike with audio lip sync, using the transcript creates numerous errors and sections where no lip synch is attempted at all, requiring time-consuming maunual fixes. Lip sync with both audio and SRT transcript is a great idea in theory, but in my expeience thus far, it fails in actual practice.

[David25727707wpd6]

100%!

• Record in Adobe Audition
• Create captions and SRT in Adobe Premiere Pro
• Export SRT verbatim as Adobe rendered
• Import into Adobe Character Animator
• Create lip sync from transcript
  -aaaaaaaand-
• Adobe claims that its own SRT doesn't match 😠

[Michael30303535yegb]

As the OP and several commenters have said this is a problem that neeeds to be fixed. As a workaround, I will run sync with transcript, and then if/when there are gaps, I will run the sync again with just scene audio, and then trim it down to fill in the gaps. As a tip, I've noticed that numbers don't play well with the transcript function, so you may see if that is an issue. 

[Inspired Kindergartens]

A work around for me is to transcribe without transcript the failed portions:

1. Duplicate the audio layer - use the original audio layer as a guide in case you move the duplicate audio layer
2. Cut up the audio layer for the failed portions
3. Compute Lip Sync from Audio (only)
4. adjust in and out for next failed portion of audio in the duplicate audio layer.

I do prefer the transcript result and the extra time spent fixing it with this method isn't too painful.

Protip: For characters with an edited voice I use the original voice recording for clarity.