konstantin magnus

Nowadays you would rather use the group expand node, though: https://www.sidefx.com/docs/houdini/nodes/sop/groupexpand.html group_expand.hip

Small update on minimal surfaces using the poisson equation and some more modern nodes. minimal_surface.hip

Here is an approach combining resample nodes with sweep, noise and polycut. The VEXpression offsets the noise for each line. morse.hip

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

Hi rev, you could paint on the large terrain and transfer its mask to the smaller tiles using a volume wrangle. write_mult_volumes.hip

Maybe this helps: https://procegen.konstantinmagnus.de/wires-connecting-objects

https://procegen.konstantinmagnus.de/circuit-networks

Or you use VEX to apply UV coordinates onto the roof: // VERTEX WRANGLE float s = ch('uv_scale'); vector up = {0,1,0}; vector nml = normalize(prim_normal(0, i@primnum, 0.0, 0.0)); vector tang = normalize(up - dot(nml, up) * nml); vector bitang = normalize(cross(-nml, tang)); vector pos_rel = v@P - prim(0, 'P', i@primnum); float u = dot(pos_rel, bitang); float v = dot(pos_rel, tang); v@uv = set(u, v, 0.0) * s; Align each polygon to the bottom left corner of the UV space: // VERTEX WRANGLE string geo = geounwrap(0, 'uv'); vector size = getbbox_min(geo, itoa(i@primnum)); v@uv -= size; roof.hip

Hi @lacer8, you may want to look into @animatrix' setup here: https://www.sidefx.com/forum/topic/84345/#post-364636

@millsy Check this to take it further: https://procegen.konstantinmagnus.de/wicker-basket

Hi @millsy, maybe start with sweeps as they provide output attributes like primcol to later create patterns. straw_weave.hip

Did you model these procedurally? If so, there are probably ways of assigning UVs while modeling and not after.

It's usually best to create UVs on the fly and not just at the end. For this, you can use the sweep node for pretty much any building part and enable its UV outputs. Additionally other output attributes such as primrow and ptcol can be useful for selecting patterns next, ie. i@primcol%3==1

Hi @GlennimusPrime, you could measure and over time remap the surface distance from the mask. attribute_spread_KM.hiplc

This could be due to the internal order of the point group. try shuffling starting points by randomly picking a point number out of the array: int seed = chi('seed'); int pts[] = expandpointgroup(0, 'start'); int iter = detail(1, 'iteration', 0); int pick = pts[int(rand(iter, seed) * len(pts))]; i@group_boundary = i@ptnum == pick; infection_multiple_01.hip