RobynsVeil opened this issue on Dec 03, 2010 · 409 posts
johnpf posted Sat, 04 December 2010 at 1:30 PM
The hem of the skirt refuses to lie down, no matter what reasonable changes I make to the cloth parameters. Fold resistance reduced to 0 gives lots of crumples; fold resistance maxed out produces so many collisions, the calculation took almost for ever. This is 35.
Shear resistance at 0 gave me the opposite of 0 fold resistance; that is, lots of collisions that made me think I'd crashed the simulation. It was taking so long I aborted that one. Shear resistance at 500 (half the maximum?) gave me very unrealistic folding. This is shear resistance at 200.
Cloth density at 0.2 gave some pointing-directly-at-the-ground effects on the bits of the skirt overhanding the chair, and c.d. at 0.01 made the hem float around as if it didn't want anything to do with gravity. Here, I've got cloth density at 0.1.
For the bits that worked, it looks okay. I like the folding at the hip-thigh bend and the way the skirt has moved out in response to the underlying hip being widened (I used a "sit down" 3rd-party morph as base V4 bends poorly without help).
However! The hem refuses to fall onto her lap, and the back has gone straight through the chair. In all the simulations I've done with this today, I noticed that the chair-skirtback collision was okay up to a certain point (around about frame 24 of 30) and then it seems to reach a certain point and think to itself "What the hell, let's just ignore the chair and forget that I have to collide with it." And the result is as the picture shows. I tried those cryptic "polygon against vertex" options in all four combinations and either produced no change or, when the second box was ticked, a crash occurred at frame 5 of the simulation.
I was hoping to use dynamics for some scenes in a story I'm illustrating where conformers just won't work, and the woman-sitting-with-crossed-legs pose at this point is looking even further away than I had feared.