Thanks for the links @clesage! I came across these links too before posting. They were sort of helpful but I think I have come up with an alternative solution that doesn’t depend on chaining deformations. I know this sounds cryptic so let me explain what I’m working on.
I’m working on a joint eyelid setup and started using some examples out there that involve some sort of blendshape approach that drives a motion system. I was trying to add in another layer deformation (skinning) to offset the blendshape while trying to have a motion system maintain the same space but ended up with some double transformations. This setup gets pretty ugly trying to offset this double transformation issue so I’m thinking of taking a different approach all together. I’ll kind of dig in a bit of my train of thought here.
I’ve been approaching this setup with keeping motion systems inline with deformations which can get a bit hairy and convoluted. Instead of relying on ‘deformations’ to drive motion, sticking with a pure motion system setup will provide much more flexibility and easier to debug and graph. I’m using Maya 2020 and really need to tap in more to a more ‘matrix’ approach to the setup. So instead of relying on deformations to get from AtoB (uplid <-> lolid), blend the matrices to these locations. Then I’d have much more control over the results of these matrices at any given point and layer on some more matrix operations.
Once I’ve wrapped my head around the solution, I’ll work on a video/write up or example file to share with the community. Maybe it can inspire others to take it and improve upon so I can make this setup even better !