gopherCG

Hi,
I'm doing particle instance simple crowd. I have randomly rotated/positioned static particles. A box slides over and adds collided particles into a “hit_group”, while other particles are in a “nohit_group”.

I noticed in SOP, instancing uses Normal and upvector, while POP instance uses velocity and upvector, correct?
In SOP I defined N and up correctly, and the rotation works for SOP instancing.
But in POP, how do I give instancePOP the orientation without the particle flying away.
Or is it possible to convert N & up axes into “fixed axes” in instancePOP?

Things I tried:
source->animated grouping-> attribute v=$N -> instance -> attribute v=0
Instance rotates to correct angle, but with last attributePOP activated to reset v, it affects upstream instancePOP and rotation reverts to zero. Shouldn't nodes only affect downstream? Same if I try velocityPOP.

suppressPOP, suppress velocity=1, doesn't work. suppress position=1, keeps position while velocity gives correct rotation, but suppressPOP somehow deactivates animated groupPOPs way upstream.
Also tried turning suppress on and off along the POP connections, but order really doesn't matter it seams. Only the last suppress in the connections matter.

Same issue with statePOP, stopped=1.

I know suppressPOP affects pstate, but it doesn't make $DEAD=1, and even so group SOPs should still work.

Hey mate thx a ton! been away for awhile, but got a chance to revisit it. Another guy was asking this as well (Peter Claes' nice dvd was for 9 i think).

I didn't know I can open the cloth object.
The right click menu's got allow editing contents,match content def
So once I manually changed it like the file, it all works. I suppose the latter means reverting to previous type definition, so I should touch that after making changes to cloth obj? Or would it run slower if I keep the custom fixed node open?

Here's the H10 test btw, thanks again!
http://www.youtube.com/watch?v=UVCT45NmwGg [youtube.com]

Here's the scene, it's for a Houdini version that didn't have tear tools (turn off the cache ROP to simulate).
So when I use H10 to replicate the effect the cloth solver doesn't respect separate connectivity (or prim group?).

On the obj_import node I made sure I can select prim groups properly, so somewhere in the DOPnet the groups disappeared. Does dop not import groups, but only pure geometry?

tried most of the settings, crashed it several times (geo error?), I'm now guessing the DOPnet alters the geometry somewhere. I found tearing works fine, but doesn't follow any borders of my primitive groups. If I only obj_import a few groups physically separated (far enough i think), then the pieces fly off individually. This might be a threshold issue.

Stitch/seam tools are confusing to me but maybe I can convert group to seam? :shock: