Jump to content

cwhite

Members
  • Content count

    157
  • Donations

    0.00 CAD 
  • Joined

  • Last visited

  • Days Won

    4

Everything posted by cwhite

  1. You can use Bullet's constraint networks to create constraints between agents and regular rbd objects - the PinnedRagdolls example in the docs shows a sample of this. There are some extra attributes you can use to attach the constraint anchor to an agent's joint with an offset (https://www.sidefx.com/docs/houdini18.0/nodes/dop/constraintnetwork#anchortypes)
  2. You should be able to use hou.PackedGeometry.setEmbeddedGeometry()
  3. Fuse SOP 2.0 No unique?

    I think the Facet SOP has an option to do this
  4. Clustering wood fractures

    You can do it in Voronoi Fracture by varying the Name Prefix based on the current iteration number. A more general approach that works well for multiple levels of fracturing is to set up names on the input pieces (for unfractured pieces, use e.g. the Assemble SOP), then change the Name Attribute parameter on Voronoi Fracture to Append and set the Name Prefix to whatever separator you want (e.g. "_"). This will name the sub-pieces by appending to the parent piece's name.
  5. Clustering wood fractures

    Your piece names aren't unique - each iteration of the for loop is producing pieces with the same names. If you drop an Extract Centroid SOP down, you'll see what the RBD Cluster SOP is working with (and why you get odd clusters)
  6. Clustering wood fractures

    I think RBD Cluster does exactly what you're looking for - what did you try (and what didn't work?)
  7. If you're using H17, there's a much easier approach that's also a lot faster than the other solutions - just turn on the Use Piece Attribute toggle on Attrib Promote example_fix_03.hip
  8. Compile block around a single node??

    Your thoughts about the performance characteristics are correct. I think in that launch example the TOP node expects a compiled block (very much like the existing Invoke SOP), which is why a single node would be wrapped up. The other consideration is cases where the compile block can be executed in-place on the input geometry, such as when using the option on SOP solvers to invoke a compiled block - in that the compile block is executed in-place on the DOP geometry instead of copying it back and forth between DOPs and SOPs
  9. If you want to do custom stuff with boolean fracturing (i.e. not the wood or concrete-style fracturing that rbdmaterialfracture provides), then you can just use the Boolean SOP or the new Boolean Fracture SOP (which is an HDA that provides more convenience features on top of Boolean). If you're trying to combine it with RBD Material Fracture, you likely need to merge together the constraint network produced by each
  10. Some of the fracturing methods in RBD Material Fracture use the Boolean SOP, if that's what you're asking about
  11. Panel From Curves Node?

    I believe it was renamed to Planar Patch from Curves before the release
  12. You can freely merge the transition graphs for each agent type, as long as you've set up different agentname values for each. The transition graph geometry has an 'agentname' point attribute to disambiguate between different agents that have the same clip names.
  13. This was fixed in 16.5.546 - https://www.sidefx.com/changelog/?journal=&categories=48&body=&version=16.5&build_0=545&build_1=546&show_versions=on&show_compatibility=on&items_per_page=
  14. Glue Constraint Relationship parameters questions

    Half life controls how quickly the 'impact' attribute on the glue bond decays back to zero (assuming the glue didn't break). So a longer half-life means that successive smaller impacts can break the glue The propagation rate is a value between 0 and 1 that specifies how much of the impulse is spread across the glue bond to the neighbouring object when an impact occurs - 0 means that nothing happens, and 1 means that the impulse is evenly split with the neighbour. If e.g. there is a glue bond between object A and B with a propagation rate of 0.5, a glue bond between A and C with a propagation rate of 1, and A has an impulse of 5000, then after 1 iteration the impulses are: A: 2000 B: 1000 C: 2000 The number of propagation iterations specifies how many rounds of this diffusion to perform, and therefore controls how far an impact spreads through the glue network in a single timestep.
  15. Try reducing the minimum bone length parameter at the top of the viewport
  16. Control crowd source randomization

    I should also mention - the input points to the Crowd Source SOP can have point attributes like state, agentcurrentlayer, etc that will override the Crowd Source SOP's parameters
  17. Control crowd source randomization

    If you need more precise control over how each agent's layer is selected, I'd just use the VEX setagentcurrentlayer() function. The Attribute Randomize SOP has a mode that is similar to the Crowd Source SOP's options (the Custom Discrete mode with Type set to String), which might be useful for doing the actual randomizatoin
  18. Crowd Sliding Backwards?

    If changing to in-place fixes it, that sounds extremely suspicious (e.g. perhaps the locomotion data and/or clip frame range is incorrect).
  19. Finding activation frame

    This thread has the simplest solution:
  20. Crowd agent orient problem

    For the Locomotion Orient parameter on the Agent ROP, you should specify another joint (like the upper leg) - the orientation is extracted based on the vector from the locomotion joint (hips) to that joint. If you preview the clip on the Agent SOP with Apply Clip Locomotion turned off, you want the character to face in a consistent direction so that you can blend between clips correctly. Also, the number of blend frames for the transition was set to 20, which is far too high given how short the turning clip is.
  21. Crowd agent orient problem

    You should be able to blend between locomotive and in-place clips. If you're able to post a scene file, that might make it easier to debug
  22. Crowd agent orient problem

    Have you set up the turning clip as a locomotive clip? it should work fine if its locomotion channel includes the overall orientation.
  23. Ragdolls explode

    I think it would be helpful to see a hip file, since I'd otherwise be randomly guessing as to what went wrong
  24. Ragdolls explode

    What do you see if you turn on guide geometry for the crowd object to see the rigid bodies? Does it look like the constraints aren't active (e.g. the bodies just fall apart), or is the ragdoll blowing up?
  25. Ragdolls explode

    If you've used the ragdoll test sim asset and it behaves fine, I'd first check that the animation clip on the crowd state is valid (if you're doing a partial ragdoll or motor-driven ragdoll)
×