Jump to content

groblus

Members
  • Content count

    11
  • Donations

    0.00 CAD 
  • Joined

  • Last visited

Community Reputation

0 Neutral

About groblus

  • Rank
    Peon

Personal Information

  • Name
    Jakub
  1. Bullet packed primitive - inherit color on collision

    Thanks for some tips pisces and Amaya. I'll have to explore both ways but I thing I'll strart with the impact data route since it promises having real collision data (no particle proxy). If I manage to get some results - I'll try to answer my own question. Regards, groblus
  2. Hi, I'm looking for a way of inheriting color on collision while using bullet and packed primitives. Let's say I have an object colored white - and I would like every single object that it touched to be colored white as well. Prefereabli those new "infected" object would infect others on collision. I did some search, found some topics on POPs using "hit count" - but nothing regarding bullet. Could you just point me in the right direction? Some use of collision data? Regards
  3. Hi, I'll reply to my own topic so that the solution stays. I'm not really sure if groups were not updating and nearpoints() failed to give proper results, but filtering results on my own managed to solve the problem: New code, that works looks like this int branches = chi("branches"); float radius = chf("radius"); int seeds[] = expandpointgroup(0,"seed"); int taken[]= {}; foreach(int s_ptnum; seeds) { vector s_P = point(0,"P",s_ptnum); int nearpts[] = nearpoints(0,"!used !seed",s_P,radius,branches); setpointgroup(0,"seed", s_ptnum, 0); setpointgroup(0,"used", s_ptnum, 1); int dests[] = {}; int index = -1; foreach (int n_pt; nearpts) { index = find(taken,n_pt); if (index<0) { push(taken,n_pt); addprim(0,"polyline",s_ptnum,n_pt); setpointgroup(0,"seed", n_pt, 1); } } }
  4. Hi, I'm trying to find a couple of nearest points, the problem is I want to avoid duplicates - so let's say point no.1 finds their nearest points (lets say no. 17, and no. 39) I don't want them to be included into nearest search anymore. To be honest I can't tell what am I doing wrong ( I suspect some stupid error or groups not updating until code is finished). No matter what I do - I still have points included in my search. Here is my piece of code - that I use in "Detail Wrangle". Then it's used inside a for loop (node based) to control number of iterations (it simulates growth). Scene file also included. Any clues on that? int branches = 2; int seeds[] = expandpointgroup(0,"seed"); foreach( int seed; seeds) { setpointgroup(0,"seed", seed, 0); //remove point from seeds setpointgroup(0,"used", seed, 1); //mark point as used vector pos = pointattrib(0,"P",seed,1); //get seed position int nearpts[] = nearpoints(0,"!used !seed",pos,2,branches); //search within not used points, //this should prevent finding duplicates foreach (int nearpt; nearpts) { addprim(0,"polyline",seed,nearpt); //draw a line setpointgroup(0,"seed", nearpt, 1);//make new point as a seed } } duplicates.hipnc
  5. Thank you very much for your file. It's really nice to see some other approach to the same problem.
  6. Hi. After cleaning my scene from obsolete nodes (just to post it here) I managed do get further (past step 3). I used "Add" node, just to delete anything but points from the mesh (it caused problems) - ignoring it made a lot of stuff work down the pipe. The point - vertex relation is still not 100% clear for me (like having point's without having vertexes - it's a new concept for a maya user ). So at the moment I have step 3) working (I think) but I'm still stuck with 5) because I get some nasty artifacts during attrib transfer at the final step. I want to explicitly copy UV values from source point to the all vertices of the copied brick - now I use attribute transfer and it gives some artifacts. I attached my scene just to show you how I managed to achieve it, any kind of help is really appreciated. My goal is to get geometry which is brickified, and each brick has a UV collapsed to a single point on a texture map. I has to use simple UV's since it's going back to maya (so no houdini extras stuff) EDIT: I think I solved it, in the copySOP, I simply "set" uv values to points (then promoted to vertex and voila it seems to work - for now at least) uvTransferExample.zip
  7. Hi/ I'm working on a brickify effect (just as in one of the tutorials available on sidefx.com). The thing that I need is transfering UV data from the base mesh to the final brickyfi mesh. It's how it works now: 1) model comes from maya with uv already (uv as vertex data) - (works) 2) points are generated as a polygon mesh, and only points within tha base model are left (works) 3) transfering uv data from base mesh to point from step 2, i need vertex data -> point data cast (can't get it done - attribute transfer ??) 4) copying geometry with copy sop (works) 5) transfering UV data from the points (step 2) to the vertex data of the newly generated mesh (bricks) - (not yet completed since I can't get step 3 to work correctly) Result: brickyfied geometry with UV's, and UV of every single brick is folded into a point - taking single point colour from the texture I have a strange feeling that the solution is simple. If it was Maya - than I'd use nearestPointOnMesh or something like that to get the uv data from model to points, but with houdini I'm stuck. I can't get attribute transfer to work. If I use attribute promote and change vertex to point than attribute transfer starts to work but the uv map is damaged during attrib promotion (which is natural, since uv should be vertex not point data). Can as ask for some guidance with step 3 and some help with step 5 (which at the moment I have no idea how to accomplish) ? Regards groblus
  8. Max # of particles - H13

    The trick is I emit particles over time and use age to control more stuff happening down the pipe. The case is I come form maya world (I used it for years) - where emission and particle count a treated separately (just like in the old pop's - max # of particles). Houdini for me is some sort of terra incognita - the first steps are painful since some of the workflow's I'm used to don't go well with houdini. Your "cache sim" solution just resolved all the issues I had - thanks.
  9. Max # of particles - H13

    Thanks - unchecking cache sim works. Is a bug or a feature ? Regards
  10. Max # of particles - H13

    Hi everyone. I have a noob question. While using the legacy POPs there is a paremater on the popnet " Max # of particles" and it's missing while using "the new" pop workflow. As I'm a maya user I'd like to have the "max count" parameter. To achieve that effect I used "popkill" node with simple vex code: dead = (i@id > 80); It worked great, so I tried to make a user controled parameter (to be used in OTL's in the future), so I change the code to something like that (I added maxcount parameter to may popkill node): dead = (i@id > ch("../mypopkill/maxcount") ); and now the magic is happening - it works or it doesn't - purely random. It works till frame 75 but then - boom, the limit stops working, I tried to change the value of the "maxcount" paramater I added, and i works unless it doesn't - randomly. What am I doing wrong ? I suppose that's something obvious that I as a maya user can't see. Or perhaps there is some smarter way to limit particle count (but please - not but calculating emission time ) Thanks for any clues. Groblus
  11. Perhaps it's a stupid question and the anwser is really simple (I hope so), but as a houdini newbie I'd like to ask a question: Is there any way of having an attribute work up the graph ? I have a simple setup a grid and a duplicate node. I have cumulitavie transformation giving me a nice set of copies. But now I'd like to use $CY variable to control grids rows and columns parameters. Is there any way to post $CY information upwards (to the grid node) ? Thanks for any info
×