Years ago I rigged some braids to deform by low res poly strips. There is a simpler way. Cage is overkill.
You could constrain a null per polygon (cluster) and turn on all the options (in the Constrain->Object to Cluster constraint) so it locks to the poly orientation well. You can then use those nulls as an envelope for your proper mesh. The default weighting with Envelope->Smooth Envelope Weights and some minor slider fiddling will work pretty much perfect. You can also build the poly constraint in ICE kinematics too, if you like, though you won't see any improvements as ICE only excels when iterating/multithreading over a big dataset Of course you can build your own cage deform — the factory "Hull Deformer" isn't very good — but you will find out soon enough that it doesn't interpolate and although your geo will follow, it will only bend as nice as the number of divisions in your low res stripes, forcing you to increase them, but then what's the point, right? Hope that helps. Cheers, -- Alan On Thu, Jan 10, 2013 at 11:32 AM, Nuno Conceicao < nunoalexconcei...@gmail.com> wrote: > Hi guys, have a problem here that havent found a solution yet. > > Problem: > Basically I want to do a special cage deform where I have several braids > of hair being deformed by a simple simulated mesh of low res stripes. > My problem is when i want to transfer the animation from the low res mesh > to the high res, where the high res braids get badly deformed by the low > res cage because basically they dont know how to split properly since they > are very close together and where a single hi res braid can be intersecting > one or two different lowres. > > Possible solution:? > So I thought the solution could be in separating each braid by a cluster > in both the high res and low res cage, then somehow change the hull cage > deformer to filter by clusters. > The issue here is that i can filter the location by cluster on the high > res mesh, but i dont know how to just pick the equivalent locations per > cluster on my lores cage deformer. > > Does this make sense? Or maybe am i being silly and there is a simpler > solution? > > >
