snoot

Good shout! I'll give it a go! Thanks @davpe

Hi guys, I'm attempting to get the thickness of an object and use this to blend between two pbr layers - in this case a sharp refractive layer into a sss-y one. I could sample an sdf volume on disk and read the thickness from there. But I'm wondering is there a way to do this at render time. For example, catching the position of a ray entering and exiting a surface with lots of sample. Any ideas? Example scene much appreciated! Thanks.

Hi guys, I have two identical bone hierarchies - one is at rest pose and the other is animated motion capture skeleton. Is there a way to blend the rest skeleton to the animating skeleton using chops? I've looked at the constraint shelf. But it would be quite cumbersome to do this for every joint using these tools. Is there a way to bring in both the rest and animating skeleton channels procedural - using wildcards or something and then blend the rest to the animation? Anybody have any setups I could take a look at? Thanks for your help!

@Farmfield Thanks for the reply. Just to clarify, you were using different dop networks with different one-way relationships and then feeding the back through each other? Any extra info would be great.

Haha, habit of writing in google docs! Pastes in that way!

Hi guys, I’m trying to get POP object and a FEM object to have two-way interaction. The pops are colliding with the FEM, but the FEM object is not being affected by the impact of the POPs. What do people think is the best way to get this to work? Currently I’m playing around using a SOP solver on the finite FEM object to bring in “feedbacks” input. This seems to contain a the particles that hit with fininte element. The pt’s have hit attributes and, “impulse” and a normalised “normal” attribute which I assume is direction. Not sure the best way to apply these feedback points to the finite element. Which attr should I modify and what do we think is the best way to lookup the "feedback" data? Perhaps a point cloud lookup? And suggestions greatly appreciated!

Just out interest which version were you using? Thanks

Hmmm... What are you meshing with? A denser point cloud should help you out... creating more information to derive the surface. You could try creating a "vdb from particles" node, and then filtering using the vdb tools. It can give your more control especially in small scale effects like the one you are trying to achieve.

Hi Mossi, Flickering: Looks like you are a using a lot of filtering on your flip fluid to get the detail you want. Try turning all that off on your meshing - it should stop flickering then.Try reducing the particle separation in your Flip sim. This should allow you to create a more detailed point cloud. You shouldn't need so much filtering then and the filtering you do need should be more stable. Sticking to surface: You might need a sop solver to force them to stick. The other option is a nasty post sim SOP approach.

A useful way to think of it is that if you have a function(noise) and you put the same input into that function, then you will get the output. So if you manipulate the position to repeat the same input, you will get the same result. In f1480187 example, a point's position (-1,-1,0) gets remapped to (1,1,0)... so it will produces the same noise result as the point that has the original pos (1,1,0). Effectively a mirroring.

Currently, each loop in a for each loop works in serial - one after the other. You can now "compile" loops to improve things, however not all nodes are supported. My guess is the Assemble works in more parallel fashion. This video is very informative: https://www.sidefx.com/tutorials/houdini-16-masterclass-compiled-sops/