attribute transfer / paint

[Dee]

Hi! I am pulling my hair out over this problem! I basically want to accumulate/paint a value onto an object based on another object. Like having a ball roll down a hill and have the ball paint the surface as it rolls. With that, I can then use that value to displace the surface or signal a shader change/blend. I just can't figure out how to accumulate it. When I do an attribute transfer, it looks like there's a spotlight panning across the ground. And paint only has one input option, so I can't use that to paint it. Also, if I wanted to use attribute transfer to "paint" displacement on, it would be useful to have the max velocity of the colliding object available. So, the same thing would apply. Just paint the max and keep the max as a point attribute. Can anyone tell me how to do this? It's driving me nuts!

[aracid]

Hey there

I would definately take a look at the Feedback SOP

Friday, December 19, 2008

Houdini 9.5.325: Added a new proto-installable SOP, the Feedback SOP. This SOP allows its output to be used as an input to compute the next frame's geometry, in a feedback loop. Normally networks connected in this manner would cause an infinite recursion error.

Another solution could be the sop DOP solver,

because u can accululate what ever u have from the previous frame,

although they use a magnet to deform a surface, in the exact same token, you can use the attribute transfer and accumulate the Cd value,

just like you need.

you'd have too dive into the example of the SOP DOP /SOP SOLVER and take it from there, but its easily achievable.

hope this helps

brian

[Dee]

Thank you! I'll do that! Awesome! That seems to work!

Hey there

I would definately take a look at the Feedback SOP

Friday, December 19, 2008

Houdini 9.5.325: Added a new proto-installable SOP, the Feedback SOP. This SOP allows its output to be used as an input to compute the next frame's geometry, in a feedback loop. Normally networks connected in this manner would cause an infinite recursion error.

Another solution could be the sop DOP solver,

because u can accululate what ever u have from the previous frame,

although they use a magnet to deform a surface, in the exact same token, you can use the attribute transfer and accumulate the Cd value,

just like you need.

you'd have too dive into the example of the SOP DOP /SOP SOLVER and take it from there, but its easily achievable.

hope this helps

brian

[dbukovec]

hi!

and another would be using CHOPS on your attribute, and just the LAG CHOP, then get it back with channel SOP, it will accumulate this way.

[johner]

Thank you! I'll do that! Awesome! That seems to work!

Another pretty straightforward way to do this is with particles. Basically you emit motionless particles from the part of the sphere touching the underlying surface as it rolls, leaving a trail of particles in its wake. From there you can attribute transfer from the particles to the terrain, copy metaballs to the particles and make a trail or feed the metaballs into the Magnet SOP to deform the terrain, etc.

The only tricky part is determining where to emit from the sphere. Few ways to do this. One I like is to use the Ray SOP to put points within a certain distance from the terrain into a "hit" group, create an "emit" attribute set to 1 on just those points, then AttributePromote to the primitive level with the "Average" setting. Then feed the sphere into a POPNet as a particles Source and use "Surfaces (attribute)" as the Emission type. So basically you get particles based on the value of the emit attribute on the primitives. Make sure to set the various initial attributes on the particles such that there's no velocity.

Edit: I should add the attribute promote stuff is only relevant if you need randomly placed particles. You could also just emit a single particle from each point in the "hit" group if you're just going to attribute Transfer from them or something.

If you find the particles aren't lying directly on the surface after emission, you can Ray SOP them down the terrain again at the SOP level using "Minimum Distance".

Edit: see attached example.

Good luck

emit_trail.hip

Edited March 27, 2009 by johner

[Dee]

Thank you! That's very helpful!

~Dee

Another pretty straightforward way to do this is with particles. Basically you emit motionless particles from the part of the sphere touching the underlying surface as it rolls, leaving a trail of particles in its wake. From there you can attribute transfer from the particles to the terrain, copy metaballs to the particles and make a trail or feed the metaballs into the Magnet SOP to deform the terrain, etc.

The only tricky part is determining where to emit from the sphere. Few ways to do this. One I like is to use the Ray SOP to put points within a certain distance from the terrain into a "hit" group, create an "emit" attribute set to 1 on just those points, then AttributePromote to the primitive level with the "Average" setting. Then feed the sphere into a POPNet as a particles Source and use "Surfaces (attribute)" as the Emission type. So basically you get particles based on the value of the emit attribute on the primitives. Make sure to set the various initial attributes on the particles such that there's no velocity.

Edit: I should add the attribute promote stuff is only relevant if you need randomly placed particles. You could also just emit a single particle from each point in the "hit" group if you're just going to attribute Transfer from them or something.

If you find the particles aren't lying directly on the surface after emission, you can Ray SOP them down the terrain again at the SOP level using "Minimum Distance".

Edit: see attached example.

Good luck