Jump to content


  • Content count

  • Donations

    0.00 CAD 
  • Joined

  • Last visited

Community Reputation

6 Neutral

About JJ FX

  • Rank

Personal Information

  • Name
    JJ FX
  1. update vel,P in popwrangle [SOLVED]

    Actually I just tried simple attrib transfer @v in sopsolver (which is besically the same thing) and Im getting the same issue... So.... Wait...... Am I just emitting on a rest position? Yes I am! So the code is working, its just my chaos on the scene
  2. update vel,P in popwrangle [SOLVED]

    No no, the code is not pasted. Because Of security I cannot copy/paste the code. So I mistyped it here Cannot share anything, later tonight when I come home I will recreate the issue and post the file
  3. Hi, So I Want for my particles to stick to the surface. So besically I just need to tell the solver to use @v from the geometry. Sounds easy. This is the code I used in popwrangle: float radius = 2; int maxPoints = 10; int handle = pcopen(1,"P",@P,radius,maxPoints); vector vel = pcfilter(handle, "v"); @v = vel I spawn points on the surface. And then I want them just to follow the surface (for now). And works nice, as long the geometry from popsource (I use surface emission) is not moved too far from the rest postion. Once its way too far, the points spawn on the rest position... And I can increase the radius, and then they do follow surface. But still the spawn in a wrong place What I want to achieve is the same resault if I just used point deform on my points to the animated geometry. So why does this code fails in popwrangle? Tried to do it on sopsolver in pointwrangle, but does not work at all
  4. Building a new station

    Thx Actually getting TR 1900x for future purpose make sense. And when thinking about upg I dont have to buy all the best parts right now. So eg I definatelly can work on gtc1050ti for some time and 32gb ram. This is the build (~2100CAD, ~1600USD): https://ca.pcpartpicker.com/list/gPf3yX
  5. Building a new station

    Hi there, So finally, after 3 years of working on a laptop Im goin back to good old proper PC Im planning to build it on my own, because I dont have a big budget. Sth around 2000-2300 for the whole case. And ofcourse I would love to build it with Threadripper. But whatever I do the budget always reaches 3k. So I guess it has to be Ryzen7 1800x. Maybe next time... This is supposed to be 100% Houdini rig. So definatelly Linux, no crappy windows. I might work on some unity game projects too, so probably I need some proper distribution. I think only Ubuntu is fully supported by unity, but im no expert on that. So few questions: 1) Does the frequency of RAM matter for simulation/rendering? Do I really need 3200mhz ram instread of some cheaper 2400mhz? 2) Is it worth to get a cheaper AMD instead of GTX? Because if Im going to experiance viewport glitches for $100 less price, than its not worth it 3) And ofcourse... is it possible to build it with Threadripper within this low budget. Any any help very welcome. Because Im going to buy the components somewhere next week. And dont want to make some stupid mistake Thank you!
  6. Curvature field

    Hm... U are right. It looks like both are doing similar thing... However now I can see that what I have is not what I wanted. However the resault is nice too curv_gradient_2.hipnc
  7. Curvature field

    Yeah yeah, something more specific would be nice Actually my friend helped me with that. Cannot share the file or screenshot because of securiity, but will try to show how its done. So.... 1. VDB Analysys (closest point, name = curvature) 2. VDB Analysys (normalize, name = cp) 3. VDB Analysys (curl, name = curl) 4. merge (2,3) 5. VolumeVOP or wrangle: curl = cross(cp, curl) 6. blast cp 7. vdbAdvectFromPoints (your points, 6, 1) - these are the inputs And looks like its doing something that resambles what I wanted....
  8. FLIP - sourcing from points

    Hi, I used shelf tool FLIP fluid from object. Then in sops I chaged mu source to just some points generated by popnet, removed all attribs from source. Then I just in FLIP object changed the input to particle field. And everything is nice and dandy, except when I go highes. I noticed that when I add more resolution (add particles, so decrease particle separation) my particles start do dissappear. Every couple of frames all particles are beeing removed. I thought that its because I use popnet as a source, but actually I removed all attribs. So I guess FLIP does not like to be sourced like that. The way I "fixed" it is more particles for source (like 1000x more!). Which is not an actual fix, because will be super slow when I really go super heavy highres. So how do I properly setup to source FLIP from points? But not using fluidsource
  9. Yellow! So how can I convert a curvature calculated from VDB Analysys (curvature) into a vector field, where the vector will be always pointing to the higher values curvature?
  10. update corresponding parameters tab

    yeah this method is nice as long u use one network and one params tab. Will dive into setCurrentNode() and PathBasedPaneTab....
  11. update corresponding parameters tab

    Does the same as .setCurrent()
  12. Hi, So in my script im using hou.Node.setCurrent() for selecting nodes. But I noticed that my parameters tab does not update when I select nodes like that. How do I update the corresponding parameters tab to the selected node with .setCurrent() ?
  13. [python] get active network editor

    paneTabUnderCursor() Yesss. Thank you :)
  14. [python] get active network editor

    P.S. Hope there is no typo, cannot paste the code
  15. Hi, Trying to get current network editor. So I though that using isPosInside() is the way to go. But does not work properly. Using this code: import hou print "\n" for p in hou.ui.paneTabs(): if p.type().name() = "NetworkEditor": print "Pane " + str(p) + "is: " + str(p.isPosInside(p.cursorPosition())) I get weird stuff. Sometimes I get double True (two active) or just printing out the wrong active network... ?