Hey There!
I know that many people have suggested that the best ways to deal with how zbrush shrinks the lowest subdivision level after subdividing the model is to store a morph target before subdivision, or to import the base mesh back into zbrush before baking. My question is a little more nuanced and tricky.
Problem 1: Volume loss in animation pipeline.
If I want the sculpted nuanced changes to be propagated down to the lowest level and include those in Maya (or any animation package) after sculpting, the volume loss makes this problematic. Using a morph target simply reverts back to what I originally had in Maya, not letting me get those added details with the volume I also originally had. The cage feature sometimes warps the mesh far too much, making it difficult for rigging to work with.
Problem 2: Using cage, while moving back and forth between maya. -
The cage feature seems to be a pretty good solution, especially if your base is fairly dense, however it still will not be identical with the original base. If I do choose to output the cage result back into maya, as it has the sculpted details, and the volume I want, bringing this back into zbrush again, to propagate a new change from maya into zbrush, will change the volume again, this time the hirez will become warped and become too large. You can see that if you keep using cage going back and forth, eventually the mesh will have gotten far too large, and become distorted.
Problem 3:Accepting volume loss, with later changes to base in an external program. -
Let’s say I am ok with the volume loss as I’m getting some of the secondary forms created from sculpting in zbrush, and implement that into Maya. If I make a change in maya to this mesh, and want to transfer this edit back into zbrush, my volume will change again on the highest, and if I go back down to level 1 again in zbrush to export that back to maya, the volume on the lowest will change one more time in Maya. This is very problematic in a production setting, where rigging and animation need things to be relatively stable and locked down, but may want the sculpting detail to come through as a minor update.
I hate to bring up the M word…but I do know that mudbox is able to address these issues. It would appear that mudbox maintains an offset between each level, respecting the volume differences between each level. So if you make an edit to level 3, level 1 will pick up that edit, but also keep its volume.
I’ve attached 2 images to illustration the issues. I hope it’s not too confusing!
If anyone has any ideas,workarounds, or maybe even workflows that might not fit exactly with what I propose, but would be workable solutions to these issues I would be so grateful! This is probably the only thing about zbrush that has mystified me.
Thanks!