Are we finally reaching the end of the uncanny valley?
No. While the faces look great, the rest of the movements the characters make in their cut scenes aren't on par. As much as I want to play this game I've very concerned that the difference in facial animation fidelity to the body animation/modeling/texturing fidelity is going to keep pulling me out of the experience. They've obviously sacrificed the quality of the rest of the body to put all the processing time into the facial stuff.
No. While the faces look great, the rest of the movements the characters make in their cut scenes aren't on par. As much as I want to play this game I've very concerned that the difference in facial animation fidelity to the body animation/modeling/texturing fidelity is going to keep pulling me out of the experience. They've obviously sacrificed the quality of the rest of the body to put all the processing time into the facial stuff.
I'm not saying this game specifically will surpass the valley, but will we see similar technology used to that effect in the near future? Is it even possible to reach that point?
From what I can tell, they are essentially projecting the recorded actor's face onto their character models through a mix of mesh morphing and texturing. So this doesn't even attempt to solve the uncanny valley problem, any more than recording a real actor on a 2D plane* does; it just sidesteps the whole issue.
*What some of you older types might call a "movie" or a "talkie"
Comments
*What some of you older types might call a "movie" or a "talkie"