Has anyone experience with facial animations for dialogue? If so, what has been the easiest/quickest/cheapest method?
i am currently experimenting with mocap in Blender, with so far some diabolical results, but it is showing promise.
I have seen programs like iclone, but am trying to do it without spending too much money
Blender seems to be the easiest so far.
You can follow this guy’s tutorial.Blender 2.8 Facial motion capture tutorial - YouTube
The cheapest way to film is a phone camera attached to a bike helmet on a stick. Something like this.How To - DIY Facial Motion Capture Rig for $15 - YouTube
Don’t skimp out of the lighting. Good lighting will basically make the work much smoother.
The other alternative is using the Iphone’s exclusive thinghttps://youtube.com/watch?v=AIHoDo7Y4_gHowever for that one, you need morph targets
There are 51 to set up, making it a time consuming effort
https://developer.apple.com/documentation/arkit/arfaceanchor/blendshapelocation
@MostHostLA Thanks. The CGMatter tutorial was the one I followed, but I moved my head around too much and got some bad results. I’ll certainly give the helmet idea a try
one thing I’m wondering though, will I be able to export the head mesh with animations just as a mesh, so I can use it in game without having to make separate sequences/animations for dialogue? So the character is just in the level normally, but with the facial animations
If you take care to use the same facial rig across all your models, then yes. And you’ll also be able to retarget and re-use them on slightly different skeletal meshes.
Just make sure to have the bones facing the correct way to begin with, or animate the proper rig directly.
The helmet works great for avoiding twist/angle issues. The motion controller idea is great to preserve the head bone rotations, however getting that data to a bone isn’t easy.
as a quick fix for that, maybe a kinect would be more helpful.