radurock

Thanks guys, both solutions work just perfectly!

Nope, just a point wrangle running on points

Hey, Probably stupid question but I'm trying to filter my impact data based on some attributes it has on it. for example i run this in a point wrangle if(@otherobjid == 0) { removepoint(0,@ptnum); } The erorr i get is "Attribute is hiding global ptnum" . Any particular reason why it's doing that? Thanks Radu

What you can try is creating a bent plane in the rough shape of a tsunami and modify the normals so that they point upwards, along the shape of the plane ; then transfer the normals to velocity and advect that velocity to your current sim so it gives you an extra push. You could also use a clip on the SOP plane and animate it from bottom to top so that you gradually feed in the velocity into your sim.

Thank you Andrew !

Hey guys, First I want to give a big thank you to this amazing community. I've learned so much from all you talented artists , and I would like to share something with you back, if I may : Thanks! Radu

If you display your collisions for the Egg you'll see the real geo that bullet deals with; at the moment it looks bad. You could lower the padding for the collision but that doesn't always give better results ; or you could scale up your scene so as not to be forced to lower the padding. However considering it's not a large scale sim you could get away with using concave to represent the egg and increase the substeps on the bullet solver cause your bullet moves pretty fast. Also play around with the glue strength . I've attached the .hip file EggYork2.rbd.hip

It has something to do with the subnet being parented to the null. If you unparent those two you get perfectly fine collisions. You are using object transform on the static objects but the tool tip says the "transform of the object containing the chosen SOP is applied to geometry" whereas in this case the object itself (the subnet) doesn't have direct transforms on it but it's rather parented to the null, so I'm not sure how the transform gets interpreted. IMO it would be easier if you could do the animation with transform at SOP level, if that is an option

Could you post a simple .hip file of the problem? Would be easier to figure out the issue

You edit the houdini environment : HOUDINI_OCL_DEVICETYPE=CPU HOUDINI_OCL_VENDOR=Intel® Corporation (or AMD) Install OpenCL from : https://software.intel.com/en-us/articles/opencl-drivers (OpenCL™ Runtime 14.2 for Intel® CPU and Intel® Xeon Phi™ coprocessors for Windows* (64-bit) , or Linux) or AMD : http://developer.amd.com/tools-and-sdks/opencl-zone/ Then you check Houdini -> Help -> About page -> Show Details and look for OpenCL Platform to be set to your CPU. All these are in the link above posted by Graham, although it's a bit confusing finding the OpenCL runtime-only download link.

Have you tried OpenCl on CPU not GPU? I get almost the same speed increase (1.5x-3x) wether I run OpenCL on CPU or GPU

My bad I haven't mentioned that the point count is not constant (it's debris emission) and when i export with frame range on the alembic, only the points that were existent on the first frame are exported along with the animations they have on them for the length of the export range.

I have @20k points on which i'm copy stamping 20 different objects. On the copy stamp i'm packing the geo and then i export it with an alembic rop. When I import in Maya I only have like 10% of the whole geometry, any idea why? A workaround I found would be to pack all the geometry stamped on points as a whole big mesh but that obviously gives me huge file sizes and doesn't take advantage of the whole instancing .

Yes i can edit them no problem!

Hey Michael, just found this thread ; was looking for someone who had the same issue. This morning I found out a way that worked for me....so I was importing the .obj geo which already had UV's (per vertex) on it and I had to create UV's for the inside faces (after voronoi fracture). I was experiencing the same problem of maya complaining about UV's being per vertex even if they actually were inside houdini. I ended up using an assemble sop to pack the geometry before exporting to alembic. Even if houdini can't show the uv's in the packed state, when I imported the geo in Maya it had perfect uv's with no issue. Hope this helps anyone that might experience the same problem !