Well, for the moment I've stalled against the performance issues of auto-sort. It seems it can only reasonably handle so many shapes. Is there any way to bind layers to the 3Drig that only utilize the z-depth of the binding pt bone?
I'm thinking I could use fa_meshinstance for the x/y plane motion and the 3Drig for the layer z-depth. If I can get it to work, it'll reduce the workload of the auto-sort script. I'll only need to use it for fine detail like facial features.
I tried to parent and extra bone between the pt bone and the layer, but that doesn't seem to work.
I'm not sure the mesh instance script would work. I think it is based on actual keys of the points. Since there are no keys using Genetes script... nothing to "instance".
synthsin75, as far as I know meshinstance duplicates the point motion channel. There is no point motion with the 3D rig! each point is driven by a bone so there is not point motion at all!!
I think meshinstance just works based on the global point coordinates. Anyway, here's a quick example. It probably doesn't instance z-depth motion, but I'm not looking to utilize that.
It's coming back now. I remember that there is more than one way to move a point. There are different ways to find the "real" motion of a point in the scripting interface. I have encountered this "problem" before where the points move when I don't want them to or they don't move when I think they should. (my favorite problem was getting caught in a loop and the point keeps going until it disappears off the screen ).
I think it has to do somehow with animated position versus position or something like that.
Glad this works. It could be very helpful with this shading thing.
Glad this works. It could be very helpful with this shading thing.
Well, I still haven't figured out how to utilize the 3Drig z-depth without having the bound layer's instanced motion being added to the layer motion. As it stands, these bits would just move right off the face.