Jump to content


  • Content count

  • Donations

    0.00 CAD 
  • Joined

  • Last visited

Community Reputation

0 Neutral

About mlmcgoogan

  • Rank

Personal Information

  • Name
    Matthew McGoogan
  1. I came across some really interesting procedural hard-surface topology here: https://www.instagram.com/mglhs_com/ He's definitely using Houdini for the workflow as seen in this post - I've been trying to figure out a workflow to generate fractures with workable topology but without success. When I try to push the boolean SOP with outputs from fracture SOPs anywhere near the kind of detail/shapes he's getting the geometry totally breaks down. It doesn't look like he's using any VDB workflow, based on the video it looks like it's purely mesh-based. Is anyone here able to get a better read on the approach he's using to get those kinds of clean fractures, gaps, and mesh separation?
  2. "square" scatter

    So this is a packing problem -- https://en.wikipedia.org/wiki/Packing_problems The bad news is there's no silver bullet here. This is a type of optimization problem where we really only have essentially brute-force algorithms that improve solutions only by trying more variations and spending more time. And even with a lot of time spent a "perfect" pack is not likely outside of ideal conditions. That said if you can find a way to stand an imperfect pack (gaps, scaling variation) then I've attached another approach using UV Layout. This is like what Entagma posted a few weeks ago (http://www.entagma.com/houdini-17-quicktip-packing-geometry-using-the-uv-layout-sop/) Using the Attribute Randomize SOP to vary pscale will improve results. If squares aren't a hard requirement you can do even better sometimes by varying a vector scale non-uniformly. Also the Island Rotation Step in the UV Layout SOP can have a dramatic impact at the cost of increased bake time and non-uniform square rotations. By varying the scaling, the number of squares, square rotation, or even making slight adjustments to the path you can improve the packing. The important takeaway is that this method does require massaging variables to achieve a better fit. Also UV Layout won't play nice if your path isn't planar. packing.hiplc