problem with sticky node

[jim c]

I've got a series of sticky nodes attached to a UV'd mesh. Each sticky node has a null attached to it. The mesh is deformed with bones, all of which are ultimately parented to a root null. When the root null is rotated on Y, I can see that the nulls attached to the sticky nodes start rotating around various axes. When I try to duplicate this with a simple model I can't. Any ideas on what might cause the child nulls to behave this way? The sticky nodes are using the default parameters with the exception of values for the UV coords. I've tried to parent the sticky node, and experimented with the various parameters, but no luck.

[edward]

Have you looked at the UVs at where those Sticky objects are positioned at? Perhaps they're near/on seam boundaries? Just grasping at straws here.

[jim c]

OK I've been able to reproduce this. I create a geo node, added mesh. Added some bones. Parented the bones to a master null. Added a sticky to the mesh. Rotate the master null along Y and sticky itself starts to spin around - look through Cam6 and you'll see this clearly. Is this normal behavior? Is there a way around it? I want the sticky to stay in place on the mesh.

biomech2-rig-err.hipnc

[jim c]

SOLVED!! At least I think so! In the Deform SOP, if you turn on "Deform Normals", then it works perfectly! Is this the correct solution? If so, why would this matter? Or are the normals remaining static and continually pointing the original direction prior to deformation? I suppose that makes sense, the Stick obj is trying to orient itself based on the normals of the mesh, correct?

[anim]

yes it does use N to orient Sticky object (as well as it uses uv attrib as an up vector)

but for your case it may be better to delete N attribute prior to capture completely because it seems to be wrong anyway

no N attribute means no problem (unless you need custom N)

and in that case no need for Deform Normals as they will be recomputed automatically and Sticky will work correctly