Jump to content

Search the Community

Showing results for tags 'retime'.



More search options

  • Search By Tags

    Type tags separated by commas.
  • Search By Author

Content Type


Forums

  • General
    • Lounge/General chat
    • Education
    • Jobs
  • Houdini
    • General Houdini Questions
    • Effects
    • Modeling
    • Animation & Rigging
    • Lighting & Rendering
    • Compositing
    • Games
  • Coders Corner
    • HDK : Houdini Development Kit
    • Scripting
    • Shaders
  • Art and Challenges
    • Finished Work
    • Work in Progress
    • VFX Challenge
    • Effects Challenge Archive
  • Systems and Other Applications
    • Other 3d Packages
    • Operating Systems
    • Hardware
    • Pipeline
  • od|force
    • Feedback, Suggestions, Bugs

Found 7 results

  1. Retime slow motion points

    Hello all, I have a sand set up based on points being advected by vdb vel. The results come out nice enough but the motion is too fast. I need the speed sim/solver speed to remain the same though to retain the kind of details I'm getting so I can't just multiply the fps down in the solver to cheap a slow-mo effect. What I'm looking for is to simply retime the points post-cache and make my cache slow-mo, which actually works using a timewarp (I just stretched the output range an extra few hundred frames) -- The slo mo effect is achieved but there is obvious stepping in between frames where the cache will freeze and play ever other frame (due to the stretching) To avoid this kind of stepping I tried using a timeblendql, which seems to have almost worked but because my points are dying and being deleted based on age, the timewarp breaks. Timewarpql has a neat feature where you can use the v of the point to blend between subframes, but since I'm inheriting my point v from the vdb vel its not as accurate is if you were using a trail sop, and again the trail sop breaks in my scenario because of the points being deleted perhaps based on age.. I can smooth the points together in post with nuke using a frameblend and you don't see the stepping... But i'm sure there is a proper and procedural way to interpolate between the substep frames of the point between A + B... just not sure how to go about this and I am having trouble finding a reference on the net. It would also be nice to have this setup on hands in the future. Thanks for all your time!
  2. Re-timing Caches

    Hey, I have a question about re-timing a cache. I don't need interpolation to slow it down, but rather I want the cache to play at normal speed, then gradually speed up. then go back to normal speed. I looked at time warp, but it only seems to remap the whole input and output rather than letting you gradually speed up or gradually slow down. How do I do this? Thanks
  3. HELP! Slow motion shot approach

    Hey guys. how would you approach a slow motion shot? I have to do a fluid sim, which has to be slowed down drastically. so here is the information I have available: I have a raw plate which is 1600 frames long super slow motion. The plate is being retimed to 129 frames in nuke. After the retime the first 35 frames is normal speed, and then it slows down until it hits the slowest point available with the fps it was shot at (speed of the raw plate). So far my initial thought is to Sim the whole thing to the raw plate and then apply the retimed curve afterwards, but I have such a hard time getting my sim to look correctly when I sim it this slow, and also that means I have to sim 1600 frames. The other thing is that I have currently slowed down the timescale on the Dopnet for my sim to get it as slow as I need, and as soon as I do the look on my sim totally change. I figured it is because the timescale is not truly a timescale that can be used for this, but this is as far as my knowledge goes with regards to slowmotion stuff. So is there anyway I can sim at regular speed, so I maintain proper force appliance etc (timescale 1 on my dopnet) and then extract some kind of information (I am thinking substeps maybe?) that can give me the in between frames information (maybe @P?) and then slow it down that way? let me know what you guys think, or which approach you would use for something like this. Cheers
  4. I've found a few threads on this topic, but I haven't seen any concrete solutions or explanations that I can understand. I've followed the DT course on how to control your fractures to start, but now I'm getting pretty stuck with what I'm trying to accomplish (explained below). I still have a lot to learn about Houdini, so please let me know if my process if Goal: To cache out a sim of a solid object that fractures a little bit when in the air, then shatters even more when it impacts the ground. I'd then like to take this cached out sim and retime it. My Process: - Make a base object called geo_orig and fracture it with a voronoi fracture sop - Make the geo_orig a packed "RBD fractured object" - Use "glue adjacent" on geo_orig so I can control where the in-air shattering occurs by deleting unwanted constraints. - Then I go into the RBD packed object and hit "make breakable" and cache out the sim with a ROP output driver. - I make a new geometry called geo_read_sim and drop a file node within to read the cached sim - I append the time blend and time shift sops within the geo_read_sim node and that's when the problems start The Problems: - When I play the retimed sim, the geometry freaks out the moment a shatter happens. I think this is due to new points being generated on impact?) - Shattering related, retime unrelated: I can see shattered geometry coming out of unshattered geometry. Like the original geometry stays behind while a duplicate geometry shatters off and penetrates through the old. Any thoughts on why this might be occuring? Research: I've looked at the hip on here, but this deals with retiming a simulation, not a cached one. http://www.sidefx.com/index.php?option=com_forum&Itemid=172&page=viewtopic&t=22492&view=previous&sid=89f08de40a51517484edd04e850202a6 This one deals mostly with volume retiming. http://www.sidefx.com/index.php?option=com_forum&Itemid=172&page=viewtopic&p=118869&sid=b2c7bc6e940c50c42208fa380d53f641 This is the best one. It says on the 1st post of the 2nd page that changing the rbd solver from bullet to rbd and messing with the FPS works... but every time i turn the rbd solver on, I lose the fracturing on impact and glue constraint work. https://www.sidefx.com/index.php?option=com_forum&Itemid=172&page=viewtopic&t=25640&postdays=0&postorder=asc&start=25&sid=353a937f6b8da5898b0a29f9eb1ec0c4 Any help or pointers would really be appreciated. fracture_retime.hiplc
  5. Hi Guys I have a Maya scene exported out as Alembic cache. It has a very fast moving characters and camera. I want to retime the cache and slow it down. It was easy to retime the characters with Timewarp but how do I retime the camera Please Help.Also what would be the best solution to sim fast moving objects Substeps or retiming the shot. Thanks
  6. Bullet Timescale

    Hey Folks, So i seem to be running into an issue of scaling time with bullet. When i use the scale time parameter on the dopnet it gives some weird results. At normal speed it works fine in terms of behaviour its just too fast for a slow mo shot. I could do it post sim using timeblend and time warping the bullet points but this wont give me accurate collisions with moving characters in the scene. I could attempt to retime the collision character and run the sim at normal speed then time warp it back to the normal speed i guess, but this would be much easier if Timescale worked. Any suggestions for making the timescale work live. Cheers Saqib
  7. slow/retime RBD sim

    Hey Guys, I know this has been asked a few times, and I did some searching, but I'm still unclear on a few things. I need to retime a large RBD sim, with many objects. The time scale parm on the dop network doesn't seem to be much use (?), and I've tried messing around with using a dop import to fetch my sim, and send that into a timeshift sop, but then how do I go about caching everything out? Is there a better way to do this? with more control? less iterations? Thanks!
×