Jump to content


  • Content count

  • Donations

    0.00 CAD 
  • Joined

  • Last visited

  • Days Won


Everything posted by haggi

  1. Render a simulation

    In my file everything is already setup for rendering. Sorry if I did not make it clear. In your scene you can set the display flag to the GUIDE null. And then you can render the geometry, or you simply open my file and render it.
  2. Render a simulation

    Click on the fluidsource node with the right mousebutton and then a popupmenu should appear which looks like the one in the image. There is a menupoint "Allow editing of Contents" at least in my Houdini. BTW. did you have a look at the attached hipfile?
  3. Render a simulation

    I think in this case the easiest way is to click on the fluidsource sop with the right mouse button and select "Allow editing of contents". Then dive into the network and set the display/render flag to the "GUIDE" null sop. fluid_.hipnc
  4. Render a simulation

    Very first result if I search for "render velocity trail houdini" on google: https://entagma.com/colored-trails-point-advection-rendering/
  5. Render a simulation

    What you see in the lower image is only a visualisation of the velocity field, not the final geometry. A fluidsource node is usually used to create volumes for fluid sims. If you want to render the trails, you can scatter points in the volume and use a volumetrail sop to create trails which can be rendered.
  6. Render a simulation

    If I undertand it correctly you want to render the velocity visualisation in the lower image, not the volume is that correct?
  7. Thanks, indeed using only the houdini.env file seems to work. But I hoped to solve it with json package files. The main problem seems to be that the Renderman paths somehow modify the Arnold paths and adding "prepend" or "append" methods to the path variables seems not to work. Well it seems that Houdini is reading the package files in an alphabetical order and so I renamed Arnold.json to ZArnold.json and now it seems to work .
  8. Hi, I try to install Renderman and Arnold at the same time in my Houdini Indie 18.5.351 using package json files. Unfortunately I cannot use both together. I attached the json files if anyone want to have a look. If I only use Renderman, it works fine, if I only use Arnold it works fine, as soon as I try to use both, I get errors like this if I create a new geometry: And the render rop node is missing a lot of necessary settings. Maybe anyone has some experience with this and can give me a hint what I can do? Arnold.json Renderman.json
  9. My resample follows exactly the nurbs curve with default settings. How did you manage to get the difference?
  10. FLIP FLUID Mesh penetrates surface

    As much as I know, this problem cannot be solved by simulation setting. Your surface field is not the accurate representation of the final flip surface. So what I usually do is meshing the fluid and then substract the collision surface from the liquid surface. There is even an option for this in the particleSurface SOP.
  11. Cross Product not returning perpendicular axis

    Instead of using @tan, you should use v@tan. Houdini only recognizes default attributes correctly. So if you use @tan, it is interpreted as one single float for all three axis. Ah.... juse a few minutes too late
  12. Motion vector not working

    I suppose that's because the optransform() returns identity matrices only in your case. What about using a trail sop to calculate the velocity automatically? Or use a timeshift to calculate the previous point position.
  13. Editing "Geometry File" parameter of Point Vop

    Usually the geometry file input is used to select the source from which you are reading data. So op:`opinputpath(".", 0)` means the very first connected input, `opinputpath(".", 1)` is the second connected input. But there is no need to edit this value. You can simply connect your inputs and select the correct one inside the points vop. The default GeometryVopGlobal node has four slots called OpInput1 - OpInput4. So if you connect something to the very first input you can read its data from OpInput1 the second one from OpInput2 and so on.
  14. Colorattribut will not read by ramp

    The RampFromAttribute mode in the color node uses an attribute to map it onto the ramp. The attribute should be normalized, what means it should go from 0-1. e.g. you can create an float attribute called "pscale" with an attributeRandomize sop. Then you have a attribute and then you can use this attribute as input for the ramp. In the color node you could then put "pscale" into the attribute name and the color should appear correctly. In your current setup you set a color with value (1,1,1) as input for the ramp and since the value 1 is the most right value of the ramp, the result is.... well 1. Attached is a simple scene where you can see how an attribute can be used as ramp input : colorfromattr.hiplc
  15. foreach- index vs iteration of all points

    It seems you are mixing some principles. First of all the wrangle node iterates over all elements. The setpointattib() function does exactly what you tell it to do. It adds 1 for every point in your array, but it does it 25 times because the wrangle node is exectued 25 times. The point() function gives you exactly what you ask it to give you. Since you loop over your three connected points, and write the result in an attribute, is always contains the last position of the array. This is executed 25 times. And you always ask for the same point in the point function v@pos=point(0, "P", pt);
  16. foreach- index vs iteration of all points

    Maybe it helps if you try to follow the instructions step by step to understand the process. There is an invisible outer loop which loops over every point is the mode is set to points. That means the next instructions are called for every point in your case 25 times. Get the indices of the connected points here you have 3 indices. For every index, add 1 to y. Go to the next point, what means go to step 2. and 3 and so on until the last point is done. Since you have three indices here [2,4,8], you add 25 times 1 to the y position of these points what results in a y value of 25 for these points.
  17. foreach- index vs iteration of all points

    In a attribute wrangle running over points you iterate over all points. What means one iteration for every point. So for every point you get the neighbours of point number 3 and add 1 to the y position. In your first example, you iterate over all 25 points, get the number of connected points of point id 3 and set a variable namend "pos" to the position of the last point id. This is done 25 times as well because it runs over points. If you want to execute it only once you could set the mode of the attribute wrangle to "Detail" so it is executed only once.
  18. Hi, I feel quite stupid, but does anyone know how I can make the surface vdb visible coming from a compressed fluid nde? The node contains the compressed points, a surface and a velocity vdb, but I cannot see them. I already tried the visibility SOP. Thanks for any hints. .... okay I found it. In the primitive SOP I can switch visibility for VDB's to isosurface what is exactly what I need.
  19. liquid steel

    Did you already try something? I'd start with a simple flip setup and increase viscosity. Then I'd try to play with velocity smoothing in the flip solver. A higher velocity smoothing gives you these nice tendrils (or was it less smoothing...?), just try it.
  20. Hmmm... to me it seems that you try to modify the activation after the simulation is doene what does not make sense. The dopimport imports the geometry from the simulation and any modification here has no effect at all to the sim. You have to deactivate the objects before you start the simulation and if you want them to be activated, you have to manipulate the rbd elements during the simulation what means in the dop network. One way is to use a sop solver. I tried to do a simple setup in the attached file. simple.hipnc
  21. How to execute shelf button via python?

    I just set the kwargs with: kwargs={} and it seems to work
  22. Hi, I'd like to execute a shelf button from python? Is this possible?
  23. How to execute shelf button via python?

    Thanks, that was fast. Unfortunately I have problems with both methods. The first method gives me (with a simple test tool which creates a sphere): >>> st = hou.shelves.tool("converter") >>> s=st.script() >>> exec(s) Traceback (most recent call last): File "<console>", line 1, in <module> File "<string>", line 12, in <module> NameError: name 'kwargs' is not defined >>> And the other one seems only to work from within the UI, but I'd like to execute the shelf button from hython commandline.
  24. Hi, maybe anyone knows an easy solution for this. I usually use the sleep mode for my machine if I do not use it for some time. Unfortunately it goes to sleep mode even if a hbatch process is running on all cores. I now turned the sleep mode off, but does anyone know a simple solution for this problem? For Windows7 I once used a standalone tool where you can select a process and it keeps Windows awake as long as this process is running. Is this still necessary? Or are there any Windows10 built in options? I even do not understand why it goes to sleep at all if it is working.
  25. Alembic export 3 floats vs vector

    Okay, I found it in the Houdini plugin in the alembic repository.