Using FACS for facial setup

#1

Hi,

We are at work researching the use of FACS (Facial Action Coding System) for setting up our facial rigs. But it seems to be very little info around on that subject… We know that “big guys” use FACS in their facial rigs, but how?

Any idea where to start? (We are looking at getting that huge FACS manual)
We are thinking of joint-based setups…
Anyone have experience with that?

Any input would be very useful!
Thanks

#2

Hi Alexandra,
Im looking into similar stuff (joint based facial setup) and have also looked into FACS in the past, which is helpful, but shouldnt be followed dogmatically, since its science of the 70s and no blueprint for high quality expressions.
I found the approach of Jason Osipas “Stop Staring” book more practical, since the facial shapes are more modular, and there are less shapes required than with FACS. There are similarities.
However, FACS is interesting. You should check the Facial Animation Toolset, which comes from a research project from the Filmakademie Baden Wüttemberg, which is based on FACS and can be downloaded and looked at for free.
http://research.animationsinstitut.de/
On the site are also high quality reference videos from FACS expressions, and some interesting stuff, that might help you with your project.

#3

Hi Robert,

Yes, at present we use setup based on Osipas “Stop Staring” book, which works quite well for most our characters at the time. But our goal is to go into high quality realistic facial animation since projects we get lean in that direction more and more. We want our facial rigs to be driven by motion capture as well as keyframe animation.

And the reason we started looking into FACS is actually Filmakademie’s work :slight_smile: We did test their tools, but they dont’t give the result we are looking for, and quite hard to extend for production. Their reference library is very impressive though, and their methods are very interesting.

So, we where wandering what other do.
Your answer confirms our own findings, thank you.

#4

This was posted last week in the News section about the production of Benjamin Button. DD used FACS on this. No deep production insight, but interesting anyway…
http://www.cgworld.com/showthread.php?746

I also read user Laa-Yosh mentioning FACS in the topology research topic, maybe he can help you much better than I can.
http://forums.cgsociety.org/showpost.php?p=6861393&postcount=955

#5

Thank you for the info.
That was very interesting, at least one can see steps of the process.

#6

I don’t have a lot of experience, but from what I have read I understand that you model AUs (Action units). Then combining these action units will produce different facial expressions. However, you also need a good corrective system to correct small defects when combining expressions or some AUs.

#7

Hi Alex,

I myself use a method similar to to curve based facial rigging on digital tutors and methodologies from “The art of rigging” and joint based solutions I’ve made up. Curve based is really solid. I’ve had to use some modifications for things that twist or turn in such as the lips but that was quite simple. Of course blendhapes can be stacked on top of the curve based solutions. Also Eat 3d has a relatively recent bone based video but that’s for games and the articulation isn’t quite as nice (my opinion) since it is for lower poly work.

http://www.digitaltutors.com/11/training.php?pid=27&autoplay=1

#8

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.