Jump to content

tmdag

Members
  • Content count

    156
  • Donations

    0.00 CAD 
  • Joined

  • Last visited

  • Days Won

    6

Posts posted by tmdag


  1. just create 'null' object inside the loop (just after 'repeat_begin1') and reference it instead of referencing 'attribcreate1'.

    you 'attribcreate1' has same value on each iteration because it is not within a loop.


  2. in situation described above, you could use 2 edge group nodes.

    First one: selecting all edges

    Second one: (with same name) selecting "unshared edges" and set its merge to "substract from existing" - this way you will get automatically selected shared edges. which later on you could pass to dissolve.

     

    if you are able to select your half edges in vex already, you probably have their points so you could use function like:

    setedgegroup(geoself(), "myedges", srcPT, dstPT, 1); 
    

    to create edge group based on your points.


  3. hard to guess just by pictures but inside "Engine" (sop level) you have obj space and "outside" (obj level) you are in world space. You could move your obj (sop level) to 0,0,0 as a one solution.


  4. nothing is happening as you are trying to "scale" each point of your curve, not the box.

    Even if we fix your scene, logic is wrong. By "scaling" point - you are just changing it's position relatively to the centroid. That's all.

     

    Therefore it is very hard to understand what were you trying to do. If you are trying to grow cube, what is that curve for ?


  5. Like I've said - I cannot speak for game industry but in vfx, for eg. modeling a chair is just modeling a chair.The fact that you can model it in 20x different ways within 5min might not matter if studio does not have Houdini licenses for modeling department. Also in many cases, companies after years of work have big, vast libraries of assets. If something needs to be custom done, it's usually because it needs to be unique and special and in such cases procedural modeling is not always an answer - if it is, it is going to FX/Env dept.

    Also, modeling is being pushed to India in some big VFX studios to cut the costs.

    I'm not trying to discourage you from anything, but i think it's important for you to understand the reality of today industry (vfx).

    • Like 1

  6. Modeling is unfortunately most over-saturated market. You would be much easier hired as FX TD / Environment Artist / Rigging Artist than modeler/texture artist (and same goes with $$$$).

    I can't speak for Game industry as I do not have much experience in that field but I can say few things reg VFX.

    1. No. Modeling is 99% maya + sometimes zbrush/mudbox etc. Modeling is dealing with assets like vehicles/buildings/characters. There is only 1% need of "procedural modeling"

    2. As far as I've heard, yeah substance painter is used around

    3. Each studio could use something different. If you want just to model, then it does no matter - you will use whatever studio supports - vray/arnold/maya/clarisse/renderman/mantra/3dlight

     

    It would be also good if you would specify what do you mean by "procedural modeling". In DNEG studios for eg they do use Houdini for Environments where procedural approach is need to scatter/create assets like trees or generate terrains etc.

    You should do whatever makes you happy ofc. but be aware that you might find it hard to find job as a modeler in general - it's just too many of them (like musicians).

    As need in VFX for "procedural models" is almost never existing, whenever such situation happens - usually it goes to FX TD to generate some procedural shape and then it goes back to modelling department to work on top of it.

     

    Hope it helps, Cheers!

     


  7. It could be so _many_ things!

    and unfortunately from the video there is no way we could debug anything. We even don't know what algorithm you have used for RBD simulation, what type of geometry, nothing.

     

    But I will take a long shot and say that this is a bullet bug in H17.0.4xx . Try another/newer build.


  8. Not a problem, pleasure to help.

    Yes, I've made that setup, feel free to use it.

     

    As of 'sagging offshoots' - yes you could play around with simulations. I've created just a line (which later is meshed) but that line could be used in a simple wire solver.

    You could also have full procedural approach - for eg. working in lines UV space. As U spans across an offshoot (from root to tip), you could displace points in -Y based on U, which could be done using a simple ramp.


  9. Would be easier to debug with an example scene.

    but few questions:

    On 3/16/2019 at 3:57 AM, minzy said:

    glass material have keep changing

    - keep changing how ? ( i only guess you meant that jittering at the end of animation?)

    On 3/16/2019 at 3:57 AM, minzy said:

    I tried to get the value to 0

    - get value of 0 of what? what value ?


  10. If you 'middle click' on any node, there is a lot of info, also a center position which you can strip out in many different ways"

    - like an answer from sidefx forums: $CEX/$CEY/$CEZ

    - or expression centroid("../node/path", D_X), centroid("../node/path", D_Y), centroid("../node/path", D_Z)

    - or vex "getbbox_center()" function. https://www.sidefx.com/docs/houdini/vex/functions/getbbox_center.html

    - or retrieve prim intrinsic bounds, substract min from max and divide by two.

     

    ... and probably 5 more other ideas ;]


  11. Unfortunately your example does not work. Something might have broken when you stripped it out (color sop ends up empty)

    But I could give you some idea:

    To remove by "number of neighbors" - you can do that noise not on Cd but on P, then with PcOpen you can check 'pcNumFound", fit that number and contrast it if needed (power). Remove point by threshold. 


  12. You should normalize your crossed vectors as they are not unit size.

    Also, order of your operations matters. If you want to 'replicate' your tree on the left, then probably you would like to:

    vector pos_b = minpos(1, v@P);
    
    vector dir = rint(normalize(pos_b - v@P));
    vector rect = normalize(cross(v@N, dir));
    vector nml = normalize(cross(dir, rect));
    matrix3 rot = set(dir, nml, rect);
    
    vector freq = {8,1, 40};
    vector pos = (rot * v@P) * freq;
    
    float deform = noise(pos);
    v@P += v@N * deform * 0.05;

     

    Although that probably still would not solve your initial issue


  13. Centos is used in most VFX studoios. It's very stable because it uses stable versions of libraries. That often means - "old" libraries. So things like Nuke/Houdini will run stable as expected.

    BUT

    as most of those libraries are "old" many new, small apps won't work that easily, some of them even impossible to get working. I've tried Centos myself for 2 years for workstation and laptop and got frustrated that i was not able to install some new apps (like new  version 'darktable' for RAW camera photo editing) - I would have to recompile half of OS for that and would probably break it doing so after.

    SO

    I've switched to Fedora, which is another branch of Redhat, so same family. Fedora is totally opposite - It's get updated all the time! which is also not that great as you are loosing stability (comparing to Centos). But as that was my home computer and not a corporation that requires to be stable during delivery, I am fine with it and very happy that i did that switch.


  14. 1 hour ago, Shalinar said:

    Hey Albert, thanks for the helpful response! I'm not sure about the first part. I was following one of Steve Knipping's particle tutorials, at the end he just adds a geo light and points it to his particle sim, and everything worked fine without instancing actual geometry onto the points or changing anything in the geo light. Also I don't really know what kind of material to add to a geo light... I tried creating a principled shader with Ce set to Use Point Color, but it didn't make a difference.

    In the end, controlling the Ce value on the particles' shader did the trick, so thank you for that! I would still like to understand why the geo light doesn't behave like I would expect it to, though.

    If you're not sure about the first part, here is an example (check before and after copy sop). Also take a look what objects i am rendering under mantra rop.

    Your geometry will become like small area lights. In case of polygons - each polygon is a small area light. 

    As for material - same, Ce is fine (I've created new one). Not sure what tutorial you have seen and what render engine it was (raytracing?micropoly?pbr?) but I'm using PBR in for this

    particle_lights_geo.hipnc

×