Regarding the Ghost shadow / Onion skin function there are some addons that offer this feature:
Mh … thinking about a combination of few key Ghosts and Motion Paths (how poses propagate between Ghosts).
How does computer know which frames to chose? We put labels on keys. Keyframe Types.
Keyframe ||| Breakdown ||| Moving Hold ||| Extreme ||| Jitter
Maybe, these could be used to tell computer which one could be used for Ghosting?
Ghosts would be invisible, unless time approaches one of the ‘Ghosted keyframe type’ poses. Motion paths show the flow between the poses.
Just an idea.
I dont have experience with cascadeur. I find single figures doing awesome stuff. Cant find information about how to use cascadeur with many chars in full pbr, with environment and assets.
Its visually appealing. But I am not sure this need be an advantage. The problem with ghosts are the same. They overlap. First ones overlap and merge to something with 10 legs, I wont be able to see whats going on there. Except, 4 poses (mid area, with little overlaps).
12 rendered ghosts means: model has to render 12 times. Now, add many chars with hi-res data. A complete environment.
What do you think? Would there be an impact on playback speed?
The most exciting thing in that video to me is the dynamic tessellation that happens on the fly in the viewport. This is something Blender should prioritize implementing for Animation 2020. Gives a huge boost to performance while not sacrificing visual fidelity, which is important for animators.
I wonder if that kind of tech is something that also can be applied to live modeling/sculpting as well.
Does anyone know if developers have planned to implement it in blender?
The technology in the video is impressive!