4 ms·
> Even if this tool could abstract away some of the complexity, it is still very difficult to model textiles or other wrapped components on a solid. It's also
by bsder 9y ago
> Even if this tool could abstract away some of the complexity, it is still very difficult to model textiles or other wrapped components on a solid.
It's also remarkably difficult to simulate fabric draping. Look at how much trouble Pixar had modeling hair in "The Incredibles"--now make all of those hairs have cross linking hairs and simulate it--yikes!--that's fabric.
I'm surprised nobody has taken one of these big GPU clusters and tried something new.
- theoh 9y agoHmmm: actually cloth simulation is constrained relative to hair simulation. (It's an orderly grid -- a Chebyshev net? -- rather than spaghetti) With a simple Verlet integrator you can animate a "fishing net" of points with no difficulty -- because of the direct links between them. Someone else will have to fill in the details of recent research. I remember a paper called "Large Steps in Cloth Simulation" which aimed to remove the need for fine detail in simulations.
- abakker 9y agoFishing nets don't stretch, so I assume the fact that clothing does adds a lot of complexity, especially since seams won't stretch the same way? Also, different sizes have the same number of seams, but won't have the same overall drape/stretch.