Facial mocap is not new. There are industry titans with products that serve the likes of Pixar and Marvel. So, would why would our studio choose the path of building our own tool? The short answer is related to our production process. Since we make VR movies, and VR is uniquely never truly passive, our pipeline is centered around constant performance of motion and voice, along with daily critiques. So we needed something custom made for that use; really frequent usage by non-technical performers and a super quick pipeline to get performances in the movie.
Why our VR film needed a different facial mocap solution.
Real-time rendered, not pre-rendered (for example Pixar movies are pre-rendered).
Fully computer generated (not live-action 360 video, though we do that for other movies).
Lightly interactive: We’re trying to walk the line between a passive movie, and what fees right in VR.
In this film, we have actors who perform both voice and motion capture performance. In pre-production, we searched for a good solution to capture the faces of actors who are 1) Not developers. So some might be very technically literate, and some might not even like to use computers. 2) Remote from us, and maybe remote from each other.