Facial rigging test
I’ve been seeing some really nice facial rigs lately and I was wondering how such rigs were made. This movie shows my latest attempt. I want as much control for this short as Ninoly possible. Blendshapes are still a possibility but I think that I’ll have more control for facial expressions if the main facial elements could be moved dynamically without affecting the current expression.
The test below is still being modified. But so far it has worked better than my old way of using blendshapes. I haven’t run into any major issues yet so thats always a good sign.

Click Image to view (13MB MOV 2 Mins)
Apologies for the terrible compression.
Some of the useful links that made me think about this:

Comments(2)
hey man,
i can see where you’re going with the facial bones under a master object, but.. there are significant disadvantages unless you further automate those controls.. like setdrivenkeyed poses that you can control on a slider, otherwise youll be sitting there posing one facial expression all day long.. and what about blinks and subtle eye stuff? (recommended reading: stop staring by jason osipa)
nice design esthetic, and solid weighting.. :)
looking forward to seeing some animated pieces.. keep up the solid work, good luck!
Hey Jordan, thanks for the advice.
Driving the controls by SDK will be the next step. I think mixing Osipas technique with this one will work very well. Using SDK with the controls I have now to automate most expressions then parenting each control to another controller located on the exact same place to further drive those expression if needed.