Jump to content


  • Content count

  • Donations

    10.00 CAD 
  • Joined

  • Last visited

  • Days Won


Everything posted by Andrea

  1. Houdini 18 Wishlist

    Would be cool to be able to add easily comments to flipbooks in MPlay and be able to save the images with them. I know is possible to use OpenGL camera comments but sometimes they get cropped and for change them you have to launch the flipbook again.
  2. Solaris Reveal

    As they said you can take objects from /obj context and import (and probably also convert automatically) for LOPS. So I don't think is going to be a problem at all to have different file formats.
  3. Ocean Spectrum Heigh Map

    There are several ways to do it. I think the easiest one is to go in the ocean evaluate, in the tab "save to texture" you can save the ocean displacement as texture aka height map. If you are exporting the geometry from houdini as alembic you can save the height information with the geometry and read it in maya.
  4. Is this from my hip file? Here it works If is from yours, be sure you unpacked the spheres
  5. I did nothing else than what I wrote above. I just went in the DOP network that is already inside the vellum solver SOP and changed the emission type. By default this parameter is not exposed so you have to do this or create your own DOP network as we usually do.
  6. Pro tips for flip meshing?

    What it does is simulating the forces of attraction that are between molecules of water. This is particularly visible in small scale/macro water simulations, where you can clearly see that the water tries to stay together. A great CG example is this https://vimeo.com/336861798 . It doesn't look bad to me. I would listen to the advices of the guys before and maybe try to reduce and randomize the pscale of the particles that are flying by their velocity maybe or by the droplet attribute. It coul help
  7. Ocean Spectrum Heigh Map

    Sorry can't understand what you need to do. What do you mean by the deepest map? Filtering the ocean spectrum based on the frequencies?
  8. Hello! First unpack your geometry before feeding it into vellum nodes, they see just one point. Second go inside the DOP network of the vellum solver. In the vellumsource change "Emission Type" from Only Once to Continuous, otherwise is going to take just the first frame. sphere_vellum_fix.hip
  9. Ocean Foam Camera Emit H17.5

    I don't think is intentional, that's a bug and should not happen. Maybe some nodes changed a little bit and caused this
  10. Ocean Foam Camera Emit H17.5

    Hi, I just tried with your hipfile and everything seems to be regular, so you did nothing wrong. The issue is inside the oceanfoam node. In particular how is calculating the density of the foam. In a wrangle (calc_camera_density) is calculated an attribute called @density. If this attribute is zero, also the foam generated in the following nodes will be zero. So why the density is zero? Without bothering you will all the mathematical operations... everything is because when you have your camera really low in ocean there are some "degenerated" polygons generated internally that have an area of 0. To fix this, just put a clean node inside the ocean foam tool, after "clip at near". See the hip file I have attached. Strangely this doesn't happens in H17. ocean_foam_fix.hipnc
  11. Whitewater h17.5 Not to emit bubbles

    In the new whitewater solver there isn't anymore a distinct separation between foam, bubbles and spray as it was before. Or better, there is, but is more a fluid transition. If you activate the option "Add State Attributes" on the white water solver you will see that you'll have some extra attributes called foam, spray and bubbles. The states are defined "simply" by the depth of the particle in the surface SDF, which is generated by the fluid simulation. This depth is not random at all, it's a range that goes between two values. It's specified by the parameters called Foam Location and Depth Range (always in the whitewater solver). A particle is going to be fully bubble if is at the depth of Foam Location - Depth Range, so at the minimum of the depth range or below. A particles is going to be fully spray if is at least at the depth of Foam Location + Depth Range. Said this, if you want to avoid to have bubbles in your simulation, you can do it in several ways. I think the cleanest solution would be to act to the whitewater source, so before the data gets to the whitewater solver. The first option in the whitewater source is "Limit by Depth". You should increase the minimum value from the default of -1 to something closer to the zero. Definitely look at your emission volume from the whitewater source and also try to give a look at your surface volume, if it has something strange there. The particles seem like they are floating on a surface, they are spawned there or moving on a specific direction? If they spawn at this position is probably your emit volume, otherwise could be the surfaceSDF or the velocity volume if they get shot in the air for example. Hard to be more specific without see it.
  12. vdb vel combine

    What you are doing wrong is how you connect the two inputs. Apparently is not possible to connect both the A input and B input to the same merge node. The node than don#t know anymore what is what. Also you don't need to specify Group A and Group B vdb_velocity_combine2_fix.hip
  13. vdb vel combine

    Hi, you should just use VDB combine and not VDB vector merge. Just plug two volumes at the time in the input A and B , then switch Operation to "Add" VDB vector merge is for merging vector volumes that have been split for some reasons, for example when you have 3 volumes that are named @vel.x, @vel.y, @vel.z
  14. Shade volume attribute based on Volume Size

    It doesn't have polygons so no, you can't calculate the perimeter of a volume. You can store the perimeter of the volume bounds, but it doesn't make much sense.
  15. Mantra UV Pass

    Connect the bind uv with a bind export, and of course export always the uv attribute. You also need to add an "aov" or Extra image plane in the mantra node. Set it to vector and VEX variable to be uv.
  16. Hair & Fur Problem - Skin Attribute not paintable

    If you use the shelf tool for creating the guides you can first select the polygons you want to create the guides on and then clicking on the shelf button. If you create the node guide groom manually you can basically do the same thing as above by selecting the polygon's group under "rest skin source". If you do this your guides should grow just on this polygons and not where, for example, you are going to have intersections. For the skin attribute lookup first you need to paint somehow your attribute on the skin. Then you plug your guides in the first input and the skin you used to grow your hairs on the second one. When you grow guides, they take the uv attribute from the geometry. This node just uses this "skin uvs" to transfer the attribute from a certain location on the skin geometry to the correct guides.
  17. Earthworms Jim in Houdini?

    I need to give to this a proper try but that's funny Here's an attempt. I generated the constraints during the vellum simulation to attach the worms to some points moving with the geometry. The movement of the worms is done with pop wind noise. Next step is to make them follow the moving character
  18. Hi, sorry that's my fault. When I gave to you the project file I assumed that there was just one point in the scene. In the wrangle I wrote int prev_active = point(1,"active",0); But this was importing the active attribute from the only point in the scene, which is 0. Instead if you have multiple points (and only if the point count is constant during the timeline) it should be: int prev_active = point(1,"active",@ptnum);
  19. It's nothing special, just keep in mind that the solver is a.. solver. Every frames depends from the previous one starting from the first frame. I banged my head on the wall many times trying to understand how they work Inside the solver you can know the value in the current frame (white colored node) and the value of the previous frame ( violet colored node). Everything I have done is adding the wrangle and writing a short expression. What it says is: step 1. import the value of the previous frame from the second input. step 2. If in the previous frame the value of the active attribute was zero and the current height (@P.y) is above a certain value -> set active attribute to 1 step 3. If one of the two previous conditions aren't met, so @active is alreay 1 or @P.y less than the threshold, it keeps the value from the previous frame. 0 if it was zero and 1 if it was 1. That's why this concept is usually called attribute retention. Hope it helps a little
  20. Here is a quick example I had saved. The solver takes the attribute active from the previous frame if the condition is not met. If the condition is true than it will assign a value of 1 to the attribute. attribute_retention.hip
  21. Hair & Fur Problem - Skin Attribute not paintable

    Sorry I can't open the project right now but if you already have hair guides and you want to delete them you can assign an attribute to the skin. Then transfer this attribute to the guides/hairs using the node called "guide skin attribute lookup". After that you can delete them. If at the start of the project you already know where the object will be, why you don't avoid completely to grow the guides there? You can choose the primitives by grouping or selecting them manually
  22. RBD does not fracture ?

    On the assemble SOP, click on "create packed geometry" for packing your geo. Rbdpackedobject can't work with regular geometry since it requires intrinsic attributes like packedfulltransform, pivot and similar
  23. Pyro trail and quick interaction

    Hi! 1. The difference is that by using the global substeps option on the DOP network you are always using this many substeps. If you use the min and max substeps on the pyro solver the substeps, the substeps are not constant and you will get them when you need them. This decision is not random but based on a condition which is the CFL condition. I can't realy tell you if one is better than the other for the results/quality but using the min&max should be faster in the same conditions because you are not always forcing the simulation to run the maximum amount of substeps. 2. Not really sure if I pictured it correctly in my mind, I should test or at least see it, but try this. Decrease the Ignition Temperature really really low (0.001), burn rate to 0.999/1, fuel inefficiency to 0 and reduce the Flame height which changes how long the flames will stay on the simulation.
  24. The Beauty

    Love it! And +1 for being able to work on a project for so long, I just can't
  25. house collapsing

    Try give a look at the hipfile attached. There are some issues in the way you built your constraints and try to break them. Of course is good that you separated them in strong, weak for the frames etc... but glue constraints are not going to break if there aren't any forces applied to them that are stronger than their strength, gravity is not enough. If you have everything connected to everything, nothing is going to move. That's why we need to use metaballs or some other magic tricks to initialize the simulation, breaking some constraints so the simulation can start and propagate. If you just delete all constraints in a certain area, of course the whole structure is going to collapse and you are not going to have any big pieces. In scenarios like that is useful to give some conditions (for example using the name of the constraints) and break just some of them. That's why preparing as much as possible in SOP is going to make it easier later on. If you want to break a different type of constraint, just change the name in the SOP solver to match it. testtOd_00.hipnc