Kind of a weird question, but here goes…
My question is about morph targets when used for facial animation. I’ve been trying to mimic the Source engine approach to facial animation to some extent; I’m using their meshes to experiment because it’s better than making my own blend shapes.
Each HL2/L4D/whatever character has 36 blend shapes for different muscle groups in the face: lid_raiser, jaw_drop, whatever else. (image:
)
When imported into UE4, all those morph targets work just fine, but then you realise every morph target controls both sides of the face: there’s an “eyebrow raise” morph target, but it raises both eyebrows. In Source’s tools, you can raise both eyebrows as one, or you can “split” the animation track and have a different slider for each eyebrow. But at the Maya end, it’s the same morph target. Just one big ol’ eyebrow (or smile, or whatever) morph target.
So Source’s facial animation uses only 36 blend shapes but effectively has double that amount of flexibility because they’re (I guess) splitting down the middle of the face somehow. I have no idea how they’re doing it and no way to find out, but it’s not like I’d have to do it how they’re doing it, so I thought I’d post here to see if anyone has any ideas. How would you do it?
Cheers