Even though the technology has taken ginormous steps to transcend the limits of the uncanny valley, my own tech skills got me crawling on its floor.
For my Performative Avatars class, I had to take the 3D scan of myself, fix the mesh and polygons, and rig and animate it. So instead of using the recommended rigging shortcut (mixamo), I followed this route and did it in Maya… well… at least I tried and have experience now…
Hell is called Maya
The process in Wrap 3 was perfect and seamless thanks to Matt’s tutorial. But nothing can save you from Maya. It’s fair to say that with a generated model (instead of a 3D scan) in actual T-pose (instead of A-pose) the quick rig works like a charm. But if you don’t have that… let’s just say that I spent over 3 hours and the results are not very nice. If you have the patience, just look at the 30x sped up video of all my attempts at the end of the post.
Anyway, I had to do something so this last one was “good enough” (until I saw it in T-pose, ugh –see the post featured photo) (it really helped that I had to restart to change to Windows, -cause Motion Builder does not exist for Mac- and that my Maya for Windows is not working, cause otherwise I would’ve kept trying). That bit was easy enough, and the result in Unreal looks very fun. But I cannot get over how bad this model looks. I did not find a way of changing the rotation of each node, which I’m sure is what caused everything to look utterly horrible. But well, I’ll try to get a better one later (or have someone do it for me?).