Stick particles to emit location?

[Werner]

Hi,

I must be losing my mind....In the past I used to get the particle emit location and set the pointposition, so that particles don't swim over a deformed (bones) mesh. This does not seam to work anymore. I feel like and idiot because the only way I can get this to work is to do it in post sim tree, which creates other problems I don't need.

Not even factory compounds like stick to surface and stick to emit location works on my mesh. WTF?

I need help and lots of sleep. Anyone?

Softimage 2013 SP1 on Win 7

Regards
Werner

[Zybrand]

Hi Werner,

Not sure if this is any help because I'm on a 2011 machine at work. But Getting EmitLocation seems to work fine for me on a Simulated ICE tree but not in the Modeling or Animation stack with a enveloped mesh.

[Werner]

Hi Zybrand,

That is our problem as well. We did manage to find a work around though.

There is a cool little compound called something like "deform hair" in Kristinka 3 that works. Check the Kristinka sample scenes. Not sure if it uses uv's but everything stays in the modelling stack and moves with the deformed object without swimming around on the surface of the enveloped mesh.
Small down side is that you need a static duplicate of your mesh, but that can be hidden, so it's not a problem.

[Helli]

You tried to emit particles in the modelling stack and used the emit location to set the point position ?
All in the same ice tree ?

This won't work.
You need to emit the particles and then set the position in a simulated ice tree through the emit location, otherwise the emit location would change on every frame cause the particles are emitted over and over again.

If you only want the particles to follow put the 2nd ice tree in the post simulation stack so u can scrub threw the timeline and you don't have to simulate.

I guess, cause u mentioned the static duplicate, its using the "reinterprete location on new geometry"-node.

[Mathaeus]

Hi Zybrand,

That is our problem as well. We did manage to find a work around though.

There is a cool little compound called something like "deform hair" in Kristinka 3 that works. Check the Kristinka sample scenes. Not sure if it uses uv's but everything stays in the modelling stack and moves with the deformed object without swimming around on the surface of the enveloped mesh.
Small down side is that you need a static duplicate of your mesh, but that can be hidden, so it's not a problem.

Yeah it's using "reinterpret location to new geometry", together with usual inverse of reference pose (that is, emit location), multiplied by moved pose. But it also takes the actual point position, for purpose of some kH compounds, able to move the point position out of emit location. So, in simulated ICE tree, it could output something unwanted, like some kind of orbiting. In un-simulated one, should work OK.
I hope there is some more generic node of this kind, for all cases, somewhere.