facial: shape vs bones vs volume deform

[ankheilw]

I don't use XSI, but I think the discussion that you want to start is a great one.

After working with a bunch of different techniques, I'm not sure that anyone is the end all, be all of animation techniques. I think I personally feel most comfortable when merging 2 or 3 techniques on a character to get expressions and deformations that really work on every level.

I have also found that by starting with bones, you can rig a high-poly character in a relatively short amount of time. Once you bind the joints to the characters face, it is much easier to deform the characters face into an expression or phoneme than tweaking individual polys. When you ge the bound face looking nice, you duplicate that and then make tweaks on the duplicate face. Once that looks nice, you can make that a blend shape for an expression or mouth movement. When all is said and done and target shapes are done, you still have the bones to overkey or add to the expressions of the blend shapes.

Having a little extra to push the animation is definately handy.

Overall though, you can do great facial animation with just about any method, I think some might just be a little bit more efficent than others.

That's my 2 cents, great topic though.

[Wigaru Wiyamoto]

Here's an interesting setup using bones: http://www.cane-toad.com/tuteRig_Facial.htm

[pur9e]

Thanks for replying, I was thinking maybe no one would!

There was a decent response to more or less the same thread over on XSIbase:

here (http://www.xsibase.com/forum/index.php?board=11;action=display;threadid=5971)

I was looking through my characters and I realized, that of five of them, this is the only one with a real face!

So this would be the one I'm going to rig for lots of expressions. Maybe it's too low res right now, I am not sure. I Haven't done any facial yet.

Your technique sounds like one I can actually try without too much trouble! I was looking at the taron pron rig and others like it and feeling totally daunted. I am all about efficiency.

Those things look like they have 50 bones in them, to say nothing of the volume deforms. And the bones scale automatically! Gaaaah!

[blakboks]

hey pur9e,

I've only rigged up a face once...but i think i can offer a few suggestions. I tried the rig by adam sale (joncrow), and i wasn't really impressed with it very much. It seems kind of redundant...it seems like it'd be just as easy to just move the nulls around, rather than doing a shape animation on the splines. Also, I kind of ran into problems with the eyes...I had a spline for the ring around the lids, but when they closed, the lids passed through the eyeball a little bit. I could've just put a shape animation on the lid, and linked it to the shape of the eye-spline, but i wasn't really that big of a problem in my animation...but could be a real big problem in other animations.

For XSI, I would suggest just creating a few point clusters, and deforming them. i.e. a cluster for the eyelids, a cluster for the mouth, etc. And then just put some shape animations the clusters. Once you do that you can either just load up the animation mixer with a bunch of tracks, and animate the weight of the tracks, or you can create custom sliders for particular facial poses (like surprise, yawn, etc.).

Another thought that i've had recently...and it's just theory...i haven't actually put it into practice yet... But it even almost seems to make sense to just animate the points as you go, like a traditional animator... I'm actually curious to hear if anyone has tried that at all yet? If so, how'd it work out?

Later,
Chris

[CGTalk Moderation]

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.