Jump to content


  • Content count

  • Donations

    0.00 CAD 
  • Joined

  • Last visited

Community Reputation

2 Neutral

About Graeme

  • Rank

Personal Information

  • Name
  • Location
    Melbourne, Australia

Recent Profile Visitors

1,392 profile views
  1. I've seen a similar looking issue in Mantra with inconsistent point numbering ptnum/id have you checked that?
  2. Constraints, custom velocity and active attribute

    It explodes when they are 100% active because within the first frame or timestep your object is colliding with the ground and breaking the constraints. Turn off visibility on the rbd packed obj and you should see what your constraints are doing. Then look at the geo spreadsheet for the constraint geo (prims) you will see some usefull info
  3. Boat reaction to large ocean sim

    Hello there! This is from the manual under particle fluid tips... http://www.sidefx.com/docs/houdini/fluid/tips.html To make particle fluids affect the motion of objects they collide with, turn up Feedback Scale on the FLIP solver’s Volume Motion ▸ Projection sub-tab. This value needs to be scaled according to the mass of the objects you need to push around and the mass of the particles, so values in the 100s may be necessary. Ahhhh I may have read that wrong... Not a sim, but the ocean spectrum...? Try the supplied files on this link... https://www.sidefx.com/tutorials/houdini-16-ocean-tools/ 1:12:30 into the vid John talks about the technique.
  4. Constraints, custom velocity and active attribute

    you could sidestep the problem by adding this code to a wrangle before your constraint output int pt = findattribval(1, "point", "name", s@name); int a = point(1, "active", pt); if(a == 1){ removepoint(0, @ptnum); } Feed the output of make_half_active into the second input
  5. No problem. I actually did try it myself and had it working with colour. Peter only looks up his wetness value and feeds one float into the RGB which is all he needs. But if you look up Cd as a vector and then do a vector to float then feed the outputs of that to the rgb it’ll work with colour
  6. Hello there! Maybe try this tutorial from Peter Quint. The relevant texture stuff is from around 28 mins in. Super cool technique.
  7. Hi guys, I have a little test I've been working on to learn more about matrices and rotations using vex wrangles (and avoiding the copy sop). I've attached a file that shows where I am at, I have an object being translated by an input P and it's rotation is is being controlled by a rotation matrix calculated using the dihedral function working on @N and @v. It all seems to work almost properly. What I am wondering is how I could use an up vector within the wrangle to stabilize the rotations, just like you can achieve using the copy sop, but keeping it all in vex. Also any tips on how to do this all better, optimisations, etc. would be ace! cheers G the file is H16 MatrixRotations.hipnc
  8. FLIP Fluid Compress artifacts

    Hi Thomas, thanks for the response! I used the wavetank shelf tool, but then I rearranged a few nodes and removed the wavetank ocean source node and went for my own setup and I did remove the boundary layer. Based on what you're saying, it must be because I removed too many pieces of the default setup. I thought I'd tweaked the cull bandwidth too, but I just went back and had a look and it turns out I hadn't. What I did in the end was just leave out the compress process, I ended up with around 170 million particles arrr! Lucky I have a lot of disk space. So now my meshing is being written out via the command line to save memory. I'll try running the default setup next time and I'll have a play with the cull bandwidth. Thanks again for taking the time to reply. cheers Graeme
  9. FLIP Fluid Compress artifacts

    Hi there, I have been toying with bringing custom vel fields into my FLIP sim, the file attached shows a whirlpool type of effect (which is in it's early stages and needs a lot of work). I have a question I am hoping someone may know the answer to... The artifacts you can see in lower left corner which appear below my surface, are these caused by the fluid compress process? They aren't present when I bypass this process, but it would save me a lot of disk space if I could figure this out. I'm thinking it is because I'm churning up the particles below the surface..?? thanks heaps! untitled.mp4
  10. Hi there Szymon! Thanks heaps for the response and insight! I did mean metaball like or vdb. I am literally only trying to accomplish this as a means to learning new things in Houdini and grow my understanding of how this could work in a big pipeline and I find this kind of optimisation quite interesting. Render time surfacing of SDF sounds promising, would you just accomplish this through the VEX volume procedural? thanks! G
  11. Hi everyone! I have posted this on SideFX forums also.. I'm hoping some of you procedural shader genius's might be able to shed some light on the topic of creating Geometry Procedurals for rendering in Mantra. I've spent some time researching online, particularly reading through the HDK documentation, and what I can ascertain thus far is that the process would be: code the proc shader in C -> Compile via Houdini terminal(Mac) -> get the .dylib(Mac) files into the houdini/dso/mantra folder -> create a new operator type, of type SHOP-Geometry/Procedural Shader and in the Node tab change the Shader Name to be the same as your Shader Name in your shader -> apply the shader to the geometry's render/geo tabs Procedural Shader parameter. This is the basics of what I did with the VRAY_DemoBox and it seemed to work fine. I am wondering if this is the most efficient process…? Is it correct? My end goal is to try and create a procedural that will render a surface blobby at render time by scattering points on my mesh (like the mesh in my video below) in Houdini, feed mantra the points from Houdini and have Mantra surface it via my procedural shader. Perhaps there might be a method to do this using Metaballs or VDB..? Maybe even isosurface..? Thanks heaps for any input, comments and insight into this process! Cheers G
  12. simple cloth tearing for fun

    Awesome! great to know, thanks!
  13. simple cloth tearing for fun

    ha that was my cheeky celebration at about 1am when it finally worked.
  14. simple cloth tearing for fun

    Thanks for your comments! I think it is currently really slow because I turned the collision steps up way too far. I wanted nice self interaction, but it is running way to slow as a result. It would be interesting to try the pcopen method though, I might give that a shot. cheers. G
  15. simple cloth tearing for fun

    Hi there!! I have been using odforce for a while now, but this is my first post! I have just been playing with cloth tearing using the SOP Solver to calculate stress by analysing the distance between points and removing them based on a threshold. Just did it for fun! Any thoughts, comments or opinions would be great! Or maybe it will help someone... HIP file is attached. cheers G stressTest_GraemePhillips.hipnc