APEX Cloth - few technical questions (rigging, modeling).
polycounter lvl 6
According to NVIDIA guidlines for aritsts:
Ie. I can get and long dress (robe), to do not have any intersections on default pose (or even to few frames in), but with more extreme, dynamic animations it seems to be impossible to prevent intersections simple due to volume of mesh.
How do you model clothes for simulation ? I'm not talking about simple clothes like shirts, trench coats, but about things that in real life have more volume. Like dresses, robes. I have tried several approaches so far, and in all cases cloth doesn't seem to behave that good in fast running animation. IDK if that because of my APEX Cloth setup or because of how I modelded meshes. Any tips on that ? Or general guide line.
Should I try to keep real life volume (and run non-realtime simulation to get base mesh), or just model it how it is supposed to look and plainly apply APEX to it.
I honestly tried both and it really didn't worked out that well.
What does it mean ? It is only about starting position (from default pose), or is it about entire animation ?Data in |Data out If the asset is in a bad state before cloth is added, it will still be in a bad state after the cloth is added. The skinned positions of the cloth should be outside any collision volume and should not be intersecting itself.
Ie. I can get and long dress (robe), to do not have any intersections on default pose (or even to few frames in), but with more extreme, dynamic animations it seems to be impossible to prevent intersections simple due to volume of mesh.
How do you model clothes for simulation ? I'm not talking about simple clothes like shirts, trench coats, but about things that in real life have more volume. Like dresses, robes. I have tried several approaches so far, and in all cases cloth doesn't seem to behave that good in fast running animation. IDK if that because of my APEX Cloth setup or because of how I modelded meshes. Any tips on that ? Or general guide line.
Should I try to keep real life volume (and run non-realtime simulation to get base mesh), or just model it how it is supposed to look and plainly apply APEX to it.
I honestly tried both and it really didn't worked out that well.