All Activity

Hi ! You can use the .12 icon to the right of the viewport to see your point numbers. As you can see, your points are properly increasing. In other words - your sort node works as you intend. You can bypass your sort node to see what the point numbers would be otherwise. The issue is with the logic in your split2 node. 0-`npoints(opinputpath(".",0))`:2 You're saying to keep all points from 0 to 999, but take one point out of two. This does give half the amount of points, but it takes 0, 2, 4, 6, 8, ..., 990, 992, 994, 996, 998 If you look at where 990, 992... are in the above screenshot, they're on "the wrong side". Instead, you want to take the points from 0 to 499, like so 0-`npoints(opinputpath(".",0))*0.5` You can simplify this like so 0-`npoints(0)*0.5` Then coloring your points will give the expected result. sortingbyX_fixed.hiplc

I've got these points that don't seem to want to listen: I wanted to sort them in the x axis, so the 0 is to the left and the highest point number is on the right of the x axis, and the same applies to the first half of points and 2nd half of points. I have placed some colored spheres to visualise and realised my lovely points don't seem care about the sort node (Sort, point By X, couldn't be clearler), or am I doing something wrong that confused them? I would have expected the left side is all red and the right side is all green. sortingbyX.hiplc

Have you come across the weird situation where with the very same file, very same settings, absolutely nothing changed, sometimes it renders the displaced ocean and sometimes it gives that mirror-like surface? that's on the same computer with H21.0.512. Then I move the file to another computer with H21.0.559 and the same none sense starts, this time the first computer starts displacing the second doesn't feel like it. If I kept getting consistently the same mirror-like render I would have thought I have messed something up, but getting randomly one or the other seems rather un-computer like behaviour, is my my hip-file haunted or is it a bug in both H21.0.512 and H21.0.559? displaced mirror-like

For distant ocean areas, it’s usually better to fade out true displacement and rely on a bump or normal map. Active displacement far from the camera adds cost without visible benefit. Using a bump map preserves wave detail, prevents overly smooth mirror-like reflections, and keeps the horizon looking more natural while remaining render-efficient.

This is the Feedback loop I was talking about: If you sort the prims by increasing rotate attribute, you can pack them tight against each others like this, by repeatedly Ray SOP'ing the latest prims against the prims already in place, and translating by the minimal distance This is an approximation to keep the Feedback look simple, there are a few intersections, but it's acceptable. If you want to pack exactly, you would have to ray SOP the points of the prims that are already in place against the candidate primitive, and handle the possible problematic cases, which I'm sure exist If the candidate book is taller than the rightmost book that's already in place, it will not lean against the top right corner of the rightmost book,there is some slight collision. If the book is shorter however, this is always exact. book_shelf.hipnc

thank you - actually im looking read from the details per wedge - i was succeeded to use " TOP Geometry Import " - but i can read just one wedge

First youy need to understand why you need substeps. In 99.9% of cases you dont, and if you save caches with substeps you can use usual timeline. Houdini get access to subframes in simulations or when you retime youe caches.

Look to TOP attribute create. You can create frame attribute threre and use it with wedge

Hi! I 'm faced with a simple problem, but I don't know how to solve it. We need to make sure that a crowd of people runs after the ball. When they catch up with him, they start attacking. And when the ball moves, the crowd catches up with it again. Here are my attempts at doing this, but I didn't do a very good job: Ask1.mp4

The UV layout-node lines up the books. Just disable the axis alignment and pack into UDIM tiles. book_shelf.hip books_shelf_3d_od.hip

If you're still looking for a solution, I can post something; Click the Like reaction on this message It's actually quite simple, but you have to be proficient in For loop with feedback SOP None of the distribute / align tools that ship with Houdini compute the collisions on the full geometry, AFAIK, you have to do it yourself with rays; there are a few different cases to test though, depending on how complex you want you book stack.

Hello everyone I have three spheres with different animation timelines want to read per-piece attributes from a detail attribute based on each wedge to feed them into the ROP Cache node (or something else ) but I don’t know how to do . i have drop a simple setup for my problem Thanks in advance for any help TOP Cache - Base on Per Piece Frame Range.hip

The interesting thing in the video is indeed the fact that adding ramp points adds geometry. I found a quick way of doing this, maybe a bit dirty but it looks like it's working ! chrampedgeloop.hipnc

Standard guided ocean setup. Have a big cargoship going thru som waves. everything works fine. As soon as i lower the particle separation on the simulation i get artifacts on the extended area on the ocean. Both visible in viewport and in render. Tried a lot of different stuf like bigger flattening distance but nothing seems to help. Anyone have a solution?

I have a geometry sequence that has substeps (saved as x number of substeps per frame). Would you have a standalone lops network graph or even a stand alone hip file just for that animated geometry with enable playback at fractional frames? or would you just mix it with other assets and simulations caches that are not saved with substeps? and if you mix it do you then enable playback at fractional frames? and if you do enable the fractional frames would that impact the other non substep elements during rendering?

I found some funny stuff, more inspired by floraform and the blender addon for differential growth ,it's mesh based and can create a variety of shapes, depending on the "weight" attribute used, and the direction. This method is not really the same as the one described in the paper I sent above. In the second hipfile, I try to replicate it 1:1. If you don't want to read i all : the main idea is using what they call an "intermediate" representation of the mesh to calculate the forces (eg, a point cloud). I'm still trying to figure out the different variations and methods used. I would love to hear any ideas, advice or anything ! TL_growth_mesh.hipnc unified_approach_grown_structures.hipnc test_edgedist.mp4 test_orientedvolume.mp4 testcurvature.mp4

This is amazing Antoine, thank you. Would you be open to create an asset that could automate this for me and handle different situations? I'll shoot you a private msg.

Totally possible using this neat trick Basically, polyslice -> clean up -> UVLayout on some input grids -> export as OBJ of DXF I attached the resulting OBJ and DXF files, which is a collection of curves laid out on 0.61m x 1.2m rectangles ( like a standard plywood sheet) The tricky part is to prepare the result of Labs polyslice for UVLayout consumption.. Other than that it's straightforward. UVLayout is not a world class bin packer, but it should do the job. You can always repack the curves afterwards once you've left H and you're in your CNC software. cutList.dxf cutList.obj Also, OBJ supports groups. Whatever groups are in the SOP you export, you will have them in OBJ as groups, or as LAYERS in DXF (I know less about that format though) That should be more than enough if you maybe need to split them further in a shell (a script in a terminal), or maybe the CNC software can deal with that.

Hi, I have a pretty simple setup. A file node to a polyslice. I am slicing the file into N curves (200 in this case, but could be many more). I need each curve to be saved out as a single curve DXF file (which I will use in a laser cutter to cut the shape). If I try to just save the geo as is it will export all curves on camera viewport. I need a single curve on a specific axis (top down). Can't figure out a way to do this - clearly I'm no expert as there must be an easy solution. Thanks for the help!

Hey, I’ve got a problem with the FLIP solver. These small spheres move when I rotate the camera. Sometimes it’s subtle, but other times it’s very noticeable and makes working almost impossible. Recording 2025-12-28 155911.mp4

Manually I can make a 4x4 flipbook of the viewport in the flip book settings, but I can't see in opengl node (ropnet) any option to make a flipbook of the viewport, I see only camera, and it chooses only 1. any idea how to make a flipbook with 4x4 grid in an automated way (I trigger the opengl with a TOPs rop fetch)?

Hey all! Im making a donut tearing simulation with vellum grains. I use the vertexsplit node to promote the uv's from my incoming donut model, from verts to points. I then transfer those point uv's from the original mesh to the vellum grains And after simming and meshing the grains, i transfer the uv's from simmed grains to the meshed donut, but when i smooth the uv's, i get these weird seam lines. Does anyone know how i can fix this?

Nothing I am aware of and it doesn't even make sense. What you can do is to use Edit Material Network LOP if you see the material in Solaris. But I recommend to edit the material on one place only.

Thank you so much. But this method doesn`t solve my problem

First up apologies for digging into a very old post here.. I've recently been looking for something similar, so it's great to find this thanks @Atom ! This somewhat works in the viewport comment by displaying the current time, but for me it does not update per frame when flipbooking or dragging through the timeline. @nikosgr00 Your solver method is working well and does update per frame when dragging the timeline, as well as when flipbooking. Between these two versions I'd love to know if it would be possible to instead of showing the current time of day, it could show elapsed time starting from zero seconds. Any ideas on calculating the math behind that?