Jump to content


  • Content count

  • Donations

    0.00 CAD 
  • Joined

  • Last visited

  • Days Won


Everything posted by amaviztime

  1. Dear Houdiniers, I am doing a heavy FX shot, mostly FLIP colliding with geometry and whitewater from that FLIP. I need to cache multiple .sim and .bgeo.sc sequences for various different SOP and DOP nets. It WAS..... all working fine, but now caching anything it seems, whether "save to disk" or "save to disk in background", doesn't work properly. In save to disk it seems to hang and when in background randomly decides to finish early and jump to 100%. There seems to be no pattern to the amount of $SF it gets to, at least none I can discern. I have tried down rezing it and allowing caching to disk to no sucess. I hope someone can help. This is a shot that may land me my first job in the industry, fingers crossed! Cheers 5D PS: haven't included .hip because it has so much cache all over the place. Should I upload it anyway??
  2. Dear Houdini Peeps, I am advecting and spinning packed objects from FLIP data and it is working well. However I need to stop any of the objects that collide with a separate rbdobject from being advected or spun anymore AFTER they collide with the object. There is a "group to be affected" option in the advect and spin nodes, so I am planning to put all packed objects into a group and then remove them when they collide with the rbdobject. Is this a good plan? NB: rigidbodysolver is using bullet I think I should use the popgroup node and popcollisiondetect node but want to know if collisions between a packed object and an rbdobject transfer to any given particle of that packed object? Otherwise setting up the popcollisiondetect node seems simple but I don't know about the VEXpression I need in the popgroup node (am a VEX noob). something like: [put all objects in said "GROUP" except any with i@hittotal>0] ??? My other bit of poss confusion is where to place the nodes? I'm guessing before the advect and spin nodes? Hope you can help Thanks 5D
  3. Transfer v from cached FLIP to BULLET rigids???

    i managed to get it to work with your advice thanks
  4. Dear Houdini genii, Most of my question is in title. Here is more specificity: I want some rigids (broken glass of aquarium) to move outwards as the water (the FLIP I'm going to .sim cache) flows down and out and around the table its on. I'm guessing a good way to do this is replace the velocity of the rigids with the velocity from the FLIP (that'd be volume???). Does this make sense? And do I achieve this? I'm guessing that i should add a little noise too. Attribute wrangle the velocity or VOP it??? Then I'm going to cache the rigids and feed back into a new FLIP that runs slightly faster and collides with the rigids so the water flows a bit in between the glass rigids (will look cool if not totally realistic???). Yh not entirely sure about the realism level (it's very very slo mo), because I cannot find reference, but looking kinda real and kinda just what I think looks cool is what im going for. Thanks in advance. 5D
  5. Transfer v from cached FLIP to BULLET rigids???

    I'm sorry I don't understand your explaination.
  6. Multiple RBD/FLIP Mutual Sim Probs

    Dear Forumers, So I am attempting to learn some deep dynamic stuff by trying a difficult shot: a bullet going in slo-mo into a fishtank that then shatters [SEE VIDEO ATTACHED] As you can see there are MANY MANY issues with this Mplay, but for now i just wanted to focus on at least the two most glaring I can see atm. 1. (and this has been a big problem throughout iterations) can you see the two frames where the RBD tank disappears near the beginning? what could be causing this? this has happened a few times in the last two weeks and less often the flip sim disappears for a frame or two??????? 2. i have successfully got the mutual FLIP/RBD interaction but it looks rubbish as you can see. it's like not all the shape of the glass fragments are being taken into consideration? (i would got through the exhaustive list of ideas I have for the problem/things I have tried but i thought this would get too long and someone may have the glaring solution without having to read through my noob mind spidermap!!!) NB: i'm using RBD not bullet atm Thanks in advance Nick TANKBULLETTESTiter98.mp4
  7. Ways to deal with RBD "popping"(?)

    Dear Forumers, What are ways to deal with (i think it's called) popping in rbd sims(?) What i mean is when all the fractured bits or objects are moving smoothly as they probably should but then one or two of them just randomly move alot faster like they have been impacted, but when they haven't? Currently going through loads of learning iterations and just increased constraint iterations to 20 (to see if this works?) i have no constraints but it says it does collisions too (NB: nothing is colliding fast with a piece to create the popping but it is colliding with the other fractured pieces next to it probably) Thanks Nick
  8. Ways to deal with RBD "popping"(?)

    Ah thanks for advice. What scale smallness are you referring to roughly?
  9. Ways to deal with RBD "popping"(?)

  10. RBD Fracture/FLIP Interaction

    My friend Dan just helped me: big ups to him!!!!! So you can simply pipe rigidbodysolver and flipsolver into a merge (rigidbody HAS to be on left, even with mutual relationship), and then increase feedback scale in the solver tab of the flipsolver above zero. AND BOB'S YOUR UNCLE woop woop Simpler than I thought it was going to be!
  11. RBD Fracture/FLIP Interaction

    Dear Forumers, Been struggling this morning to get a FLIP fluid to "push" a fractured RBD object I want in a [breaking glass tank of water scene] the water to push the glass outwards. Researched alot and tried: 1. Just piping rigidbodysolver and flipsolver into merge inside DOPNET with left affects right, and FLIP on left. 2. fieldforce with force attr set to vel, with node fetching vel from flipfluidobject 3. geometryvop with volumesamplevfile sampling vel, piped into multisolver with bulletrbdsolver 4. popadvectbyvolumes with vel from flipfluidobject piped into rigidbodysolver presolve Can't get any method to work. Sad me! Attached is my .hip file Any help much appreciated, Nick NB: the bullet is not meant to do anything yet, but wanna pipe into FLIP after I get this first problem solved!!! NB2: not worrying about the constraints for the bottom of the tank yet either! tanksetup06.hiplc
  12. Rendering slices of pyro

    Dear Forumers, I have a project where I am combining slo-mo explosions (from Houdini Pyro) with a few different parts of exploded sci-fi drones (that are comic book style, illustrated by someone). I want there to be depth to the final output so I want to slice each pyro into 3 or 4 segments (perpendicular to the camera's LOS), so I can put the segments on top of each other with the drone wreckage in between, so different bits will be obscured by different amounts of smoke and fire. I do not want to light the bits from the pyro (it doesn't have to be photo real/or even close, going for heavily stylized result). I reckon I will be able to figure out the slow mo but I'm not sure how to render different depth slices of the pyro. Thanks in advance for any help Nick PS: Want to use Redshift
  13. Dear Forumers, I am planning and continuing to construct my own pet project short film. The vast majority of the 5 or 10 minutes it will fill is CG, for various obvious mainly ZERO budget considerations (i am a believer in as much practical as possible but you need to have budget obvs). The plan is I am going to do practically everything and it's a TALL order but one I hope to achieve within maybe 5 years of spare-time. Anyway I'm not going to explain the idea here, but one major set and it's construction has been bugging me, mainly because of the title of this post. I intend to use Houdini and Redshift, for the 3d. I currently have a very good home PC (18 Xeons, 128GB RAM, RTX 3090, possible need more SSDs though). Anyway the set is in creation atm (an "underground science lab" that will contain 2 scientists that are not CG and only them). There is a central core with loads of stuff coming out of it (think AKIRA cold storage, ref attached) and inside loads more equipment. I have no idea what the polycount will be, but it's gunna be HUGE. I think the problem is that most of the runtime is taken up by one single long take shot covering all different details in the lab and transiting outside eventually (I do plan hidden cuts but not sure many or maybe any poss inside lab). So I am not a 3D expert, but not a noob either. So to actual questions: 1. How does one approach planning levels of detail and specific geo poly counts for shots and projects? 2. Displacement maps are obviously essential but how far can you go with them. Does using more displacement maps and less polys help the memory situation vastly? 3. What is normal memory used for if the GPU is holding the scene in it's memory? 4. I'm going to use SUBSTANCE PAINTER for textures. How important is the resolutions of these textures towards memory usage? 5. Any other general advice? (NB: going to do previz so I can see how much lack of detail I can get away with) PS: Set is going be quite dimly lit (Obvs normal labs are bright and whitish but not going for this for artistic reasons), this should help hide some details missing??? Thanks for any advice, 5D
  14. Rendering slices of pyro

    I abandoned it due to artistic considerations, but I imagine ryew's second suggestion would be best. "use the VDB combine node to clip the cached volume to the intersection with a VDB box you have created". Hope it works well.

    Thanks, will try. =)
  16. Dear Forumers, Yh the title and attachment explains all. What is this please? Thanks 5D
  17. Hey guys and gals, Trying to cache 10 frames with 20 subframes each, then bring them back in across 200 frames on slider. I'm using a file cache after a pyrosolver (I used $FF AND turned off integer frame values). I hoped to bring them back in after caching and "stretch" with timewarp SOP. Is this the right way to do this? Also my first problem atm is that the $FF subframes don't collect nicely into one sequence to choose in filecache (because of 2.1000000000000001 etc???) Any help much appreciated, 5D SLOMOMANTRABOOM02.hiplc
  18. Initial SIM frame in all subsequent frames

    it maybe VRAM problem. i have 24GB on a 3090 so thought it would not be a problem but decreasing voxel size massively solved it. so i reckon makes sense its VRAM prob yh? gunna try to find sweet spot now.
  19. Dear ODForce people, I've fed a frame from a pyro sim I did into another sim to be the initial data. I then run that and cached it but it seems to still contain the initial data frame. I've attached .hiplc file but not cache files. Can you see or do you know the problem? Do I NEED to upload the cache files too? Cheers Nick SLOMOPYROexplode05.hiplc
  20. Initial SIM frame in all subsequent frames

    this only seems to happen when attempting cache. i'm trying to troubleshoot atm
  21. Initial SIM frame in all subsequent frames

    getting this error now: Traceback (most recent call last): File "Sop/filecache/execute", line 1, in <module> File "C:/PROGRA~1/SIDEEF~1/HOUDIN~1.462/houdini/python3.7libs\houpythonportion\ui.py", line 1081, in decorator return func(*args, **kwargs) File "C:/PROGRA~1/SIDEEF~1/HOUDIN~1.462/houdini/python3.7libs\hou.py", line 69068, in render return _hou.RopNode_render(self, *args, **kwargs) hou.OperationFailed: The attempted operation failed. Error: Cook error in input: /obj/explosion/filecache2/cast_attributes Error: Invalid source /obj/explosion/pyrosolver2/dopimportfield1/load_each_field_end Error: The sub-network output operator failed to cook: /obj/explosion/pyrosolver2/dopimportfield1/switch3 Error: Invalid source /obj/explosion/pyrosolver2/dopimportfield1/load_a_field Error: pyro - /obj/explosion/pyrosolver2/dopnet1/pyro_solver/add_turbulence/add_forcvel_temp_to_vel: OpenCL Exception: clEnqueueUnMapMemObject (-30).... WTF?
  22. Initial SIM frame in all subsequent frames

    cool. i see what you did. thanks. so i am going to proceed by siming 50 sub frames and then putting these into the normal frame time range right? not entirely sure how to do this but going to look into it this is the way to do it isn't it? i understand timescale breaks the physics equations under the hood? navier-stokes? i'm intermediate at maths.
  23. Initial SIM frame in all subsequent frames

    Thanks alot bunker. Yh i bet its not easy!!!!! I always dive in the deep end with my projects, even if they don't get finished sadly! BUT NOT THIS TIME!!!!!!
  24. Initial SIM frame in all subsequent frames

    oh ok. here: SLOMOPYROexplode07.hiplc
  25. Initial SIM frame in all subsequent frames

    here is short clip shadingproblem.mp4