View Full Version : Lip sync and Frame rate.

06 June 2007, 09:52 AM
I would like to know more about Lip sync.

1. Any learning resources about setting up standard mouth shapes of vowels and constants?

2. I would like to know the relationships of lip syncs and frame rate. If I received a soundtrack or script that speak very fast while about the animation, I work with a moderately low frame rate (like 25 fps), Do professional will do fractional keyframe or fractional keyframing is necessary? Or just do the lipsync approximately?

06 June 2007, 03:07 AM

Lip Syncing is one of the more difficult aspects of animation, as not only does the motion of the mouth need to look good, but it als must sync up to the sound and dialogue you are using in a believable manner. I'll give you some of the things I go through when I do some lip sync animation and some links to some tutorials and articles you may want to use in general.

In general, when I set up the mouth shapes I use about six or seven basic visemes (Basic visual elements for sounds) for each Phoneme (Basic sound element). To get them I use a mirror and shape my blendshapes as such (if you wanted to go the blendshape route for your model that is) If you wanted to get every sound, you could go up to about nine shapes for each viseme to get the fll spectrum of sounds for your dialogue. However, you can also mix and match shapes to get even more sounds as well. This is suggested as this will give your syncs a more organic feel.

Now, you asked about Lip sync and frame rate. This is tricky a tricky part as some 3D programs will compress the voice track to fit into the frame rate you select as your default withint the program. So for example if you set your fram rate to 24 fps, the voice track will fit and look good at 24 fps. However, if you export it to another compositing program and the audio file itself wasn't made for 24fps, then your lip sync will be a lot faster than your audio and thus it will get all messed up. So it's important that you match your frame rate to the audio file, just to avoid this. I'm not familiar with fractional keyframing, and someone else might want to fill you in on that, but as for my experiences I just avoid it and work at the frame rate that the audio file is in. Usually is about 30 fps.

Now for some resources to help you out.

This first one was written by Michael Comet ( In this article, he talks about how to construct Visemes and some basics in lipsync and some things to look out for when doing facial animation in general.

This second article is written by Keith Lango and he talks about more about the process of creating a lip sync animation rather than the nuts and bolts of how to construct visemes etc.

I hope this provides some aid in your quest to get your characters talking. If you have any other questions, please don't hesitate to ask. Good luck!


CGTalk Moderation
06 June 2007, 03:07 AM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.