timfagan

Have you tried using the 'Filter Above Resolution' on the ocean spectrum node? You can clear away those higher frequency waves by setting a lower value.

I can't see looking at it what's wrong, but I like how you used the foreach on the plank pieces to create the voronoi fracture. Saves the extra attribute trying to separate the wood pieces later...

Thanks Johnny, yeah the only problem with using those transforms is that it effects the glue network extracted from the voronoi fracture node, it's all shrunk down. Maybe there's a way to compute this separately though. Or just scale that network back up, I'll check this when I'm in front of it later.

Really cool, maybe some light/shadow projected from the lava to the environment and the steam doesn't seem to follow the lava as it pours down.

I ended up just loading an image on to a free image hosting service and linking with the URL.

Thanks guys, yeah I found the polyexpand2D to be pretty useful as it avoids the overlapping when extruding in a flat plane.

So I've been trying to figure out how to fracture panels of wood! I've got a method here but I'm hoping there's a better approach. I wanted a high strength/glue value within each panel piece and close to zero between panels. The basic idea is to pass an attribute which has a unique value for each panel and use this to create the clusters. Then delete the glue links between different clusters. It works but I'm kinda hoping someone has a better approach, as this one gets quite slow especially moving around in DOPS. See an example of the final glue network attached. Anyway I've set up the file so there's very few fractures so it loads quickly, but to really get it to work it needs at least around 3000 points. So if you check out the file have a look at it, then increase the number of of points on the scatter node. One problem is that I haven't been able to get the nice elongated 'wood' fractures via the transform before/after method - it changes the glue network taken from the voronoi fracture node.... Also, the cluster tab on the voronoi fracture node seems to have no effect? If this worked all this would probably be unnecessary. Wood_Panel_Fracture.hipnc

If I wanted to smooth that middle line/edge, can I resample just that line somehow? I've managed to isolate the line through either a group SOP, but can't seem to resample it. Or I can delete everything except those middle line points and resample them, but I'm not sure how to reinsert them back to the mesh... (The resample node has that option to treat polygons as Interpolating Curves, which smooths out the line as it inserts points.) I've fiddled around with the connect adjacent pieces and using that to try to form a polygon mesh. I also tried converting them to NURBS curves, and skinning them together, but they don't connect correctly. C_smooth.hipnc

Haha that's really cool, I haven't seen those sops used in that way before. thanks again.

Also with the polyextrude, what you say makes sense, but anytime an inset is applied with a high angle between adjacent edges, it seems there's going to be overlapping polygons, as in the ends of the 'c' above. Is that right? I'm sure there's a way to correct for it, I don't know what it is right now...

Yeah looks spot on, thanks for the tip Konstantin.

...A resample node before the extrusion cleans it up a bit. I'm still not sure how to clean up the edges.

I'm wondering if anyone has figured out how to solve this problem with the extrude sop? I have a trace of the letter C, and I would like to add an inset with the extrusion. However the result has overlapping polygons. Is there a way to correct this? I thought it might be to do with the normals but they're all in the same direction... Edit: What I'm trying to get is the extrusion meeting in the middle of the letter....

Hey guys, I'm relying on displacement to create procedural paneling and I'm having some issues. When making panels, would it be a wiser approach to build it procedurally at object level rather than rely on the shader? In the checkerboard pattern, light gets through between the positive and negative displacements. Is there a way to fix this in mantra? Also there's no back face... In the other image, the white area should all be displaced the same amount, but some sections seem just translated out.

Hey Atom, yeah it seems they need to be primitive attributes. This might help: https://www.sidefx.com/index.php?option=com_forum&Itemid=172&page=viewtopic&t=38882&highlight=stylesheets