Jump to content


  • Content count

  • Donations

    0.00 CAD 
  • Joined

  • Last visited

Everything posted by pgrochola

  1. Python install version?

    I tired to install these on Windows a while back, I couldnt get it working in the end so gave up on windows. Install Ubuntu, much easier and works great! Linux is a much better environment for this kind of stuff.
  2. feedback loop in COPS

    Hi, Can someone here help me out with a cops feedback example. I'm sure it's possible but haven't managed to set it up. Cheers Pav
  3. hey, Anyone else notice when using RBD packed objects, the convex hull generated changes typology thorough out the simulation? Causing jitter and general strange behaviour. This is particularly evident on small objects because it happens at the edges (or when collision padding is high) Check out below file, inside the DOP i have the same object as packed or regular rbd fractured object. Apart from the typology changing on the packed objects, the behaviour of the RBD fractured object looks much more natural to me. Not sure if this is related to typology change, but there is something odd about the pivot location of the packed object. Thanks Pav packed_vs_standard.hipnc
  4. Hey smart people, Im not sure if the topic title describes my problem very well but,... I have a math problem that's been holding me up. Maybe someone out there can help. Say I have a bunch of points with normalized values sorted by highest to lowest like... @ptnum @value 0 .9 1 .4 2 .3 3 .01 ect.. How do I now remap "value" for each point so that the total adds up to one? BUT this new distributed remapped value is weighted by the original value. The obvious thing to do would be: (@value/sum of @value) but the result is: @ptnum @value @simple normalized 0 .9 .55 1 .4 .24 2 .3 .186 3 .01 .00621 ect.. This kind of remapping is not ideal because point 0 is has a very high value which gets lost. In the case above I could keep point 0 at .9 the distribute the remaining .1 across the rest of the points. Which I know how to do but what happens when you have a case like: @ptnum @value 0 .6 1 .58 2 .52 3 .1 4 .1 ect.. In this case I would need to average the first three points then distribute the remainder across the rest of the points. Is there a really smart solution out there that doesnt involve setting up test conditions then adapting the code? I've attached a hip file that demonstrated the problem.. Thx! mathProb.hiplc
  5. thx Anim! Hadn't thought to do that and it helps.
  6. Hey, Seem like there should be an easy way for the "locomotion" velocity of the current time sample (from that clip) to influence particle velocity. I dont want to "Lock Particle Speed to Mocap" because I lose the ability to simulate the crowd. (correct me if Im wrong here) With the "inplace" method I need the particles velocity to automatically adjust to the clips locomotive velocity at that time step. I know there is a "Gate Speed" but what if the character changes locomotive velocity over the course of the clip? Without this, it would be difficult to correctly blend between different states while correctly adjusting the particle velocity. The "inplace gate speed" should be linked to the correct sample of the locomotive velocity rather than being a hard number. This seems like a basic requirement.. am I missing something? Anyone found a work around? Thx
  7. Bullet: small objects not stopping to move

    try lowering your collision padding. If you switch on display collision geo in dops I bet the small pieces look very rounded (because collision padding is too low) Will change the look of your sim though best to avoid siming really small pieces.
  8. RBD packed primitive convex hull jitter

    hey, any info from SESI on this?
  9. RBD packed primitive convex hull jitter

    On my system I see the "display geometry" (original geo from sops) flickering. That's what I think gl problem. the convex hull collision geo changes typology, which is not a gl issue. Anyway.. Thanks for submitting bug. If we don't hear from them soon ill submit one also.
  10. RBD packed primitive convex hull jitter

    thanks for looking. I noticed the geo flickering also but put it down to GL bug. If you look at it in wireframe its fine. I also scattered some points on the primitive and they don't seem to flicker around so put that one down to display. The convex hull flicker gets worse at small scale objects or large collision padding.
  11. Glue System Problem

    The reason its happening is because you have two sets of name attributes feeding into the constraint network, both prefixed with "piece". Result is that you have multiple pieces named the same thing. this is a problem for the constraint network because you need unique names so it knows how to build the constraints. In the "Glue" subnet.. see how the polygons are all crazy. Easy fix is to use a different name attribute inside of "create_fractured_groups" .. like "static_pieces" or something. This will then make glue for correctly for balcony. To glue balcony to static building you'd bring in those pieces in before the "connectadjacentpieces1" node. Hope that helps
  12. Testing bullet: why is everything popping/blowing up.

    yes! I just posted a new thread about this.
  13. Hey, I have a very simple curve rendering question.. Inside a shader.. how do you get the equivalent of parametric s? When building a shader, if i plug the parametric t into the surface color i get a nice black and white ramp running ALONG the curve. How do I get a ramp that runs ACROSS the curve? Pluging the parametric s into color doesn't return anything. (which makes sense because curve has no geometric thickness) What am I missing? thanks!
  14. curve rendering parametric s?

    ok got it. just tick on "shade curves as surfaces" in the dicing tab.
  15. BulletSOP 2.0 beta_v3 for Win64

    ok, I get it. A and B stand for the two bullet indexes that collide. makes sense!
  16. BulletSOP 2.0 beta_v3 for Win64

    Hi Milan, Great job! Love the speed and simplicity. I'm particularly interested in the "voronoi_impact_break" example. Can you give me some more info about the point attributes that come in from impacts ("btLoader_IMPACTS") For starters, there seems to be an A and B version of the same data. Why is that? What is "bullet_indexA" and "bullet_indexB"? Thanks Pav
  17. Methods of fracture geometry

    You should play around with squashing and stretching your geo before you feed it into the voronoi fracture. You then store this position change, fracture the geo and then stretch the geo back to the original shape (using your stored position change attr) That way you'll get much more interesting shapes! Works really well for wood splinters, glass shards ect. Avoids the whole crappy Voronoi cell look.
  18. Compute normal

    great example. Thanks SYmek!
  19. cheap curve collisions

    Hi, I have a bunch of curves stuck on a surface (for fur groom). I'm trying to avoid going through dops to do geometry collisions. I'm developing an idea of using the velocity of the collision object to calculate the rotation of a curve at the base. At the moment I'm stuck on how you define the pivot in the rotate node in vops. The attached example illustrates what I'm trying to achieve. Has anyone done something similar? Thanks cheapCollisions.hip
  20. cheap curve collisions

    Here is a quick and dirty hack using the sop solver. (no dynamics) Could be useful if speed is more important than accuracy. Pav furDynamicsV05.hip
  21. simple wire solve question

    Hi, I've got a bunch of curve stuck onto a surface that moves around pretty quick. I'm then running a wire solve on these curves. What I'm trying to do is have the dop curves follow the sop curves and not become dynamic until they are hit by a collision object. Reason I'm doing this is I don't want to effect the groom of the curves, I just want to add collision interaction. I tried increasing the "angular spring constant" (which forces the dops curves to follow the sops curves much better) but it messes with the collisions (because curves are too stiff). Can anyone shed some light on how to solve this? Is there a better way? Thanks
  22. cheap curve collisions

    thanks Macha. I've also looking into using a sop solver for this. Seems really promising. I'll post my finding if I come with anything decent. Cheers
  23. Houdini 12 Sneak Peek

    Brilliant work SESI! Massive improvements, congratulations to everyone involved.
  24. Posted an example of something similar here: Does that help?
  25. I've just finished a job where I also struggled with this problem. In the end I had to export my geo as an obj sequence and read them back into maya using a plugin. Luckily I didn't need motion blur! (fracturing was happening in slow motion) It was also very slow to read a new file each frame in maya, but at least it worked! As far as I can tell Maya has big problems reading radically changing typology. Sounds like the best solution would be the rayfire approach mentioned in a previous post... basically you need to eliminate the "changing point count factor" through some sort of baking process. It's an awesome tools, just needs that one last step. anim: great suggestion... Hadn't thought to try that! Would the bin file store velocity correctly and would maya/vray render the motion blur? Would love to know if anyone has had any success with this! Cheers Pav