cwalrus Posted April 13, 2016 Share Posted April 13, 2016 Hi - I made a particle-driven cloud sim of dust coming off tires - it is perfect except for my art director wants it to play back at though it's in slow motion... This is tricky because it is following an alembic source. So basically i just have to have the particles and the cloud sim evolve more slowly, but otherwise do the exact same thing. I tried the 'Timescale" and the "Scale Time" on the solvers but neither preserves the previous look of the animation...e.g the particles barely make it off the tire ... any tips? thanks! Quote Link to comment Share on other sites More sharing options...
rbowden Posted April 13, 2016 Share Posted April 13, 2016 Could you take what you have already run out and use a Timewarp SOP to do your slow motion?-Ryan Quote Link to comment Share on other sites More sharing options...
cwalrus Posted April 13, 2016 Author Share Posted April 13, 2016 No - as I said, it is locked to the alembic animation (the tires moving and spinning). Quote Link to comment Share on other sites More sharing options...
rbowden Posted April 13, 2016 Share Posted April 13, 2016 Right. I was talking about your cache, not the alembic. So you have your dust cache wrote out as a bgeo.sc, yes? How about taking that cache, and timewarping it? From your first post, it sounds like the alembic is not in slow motion (I could be wrong, that is what it sounds like that). It sounds like your art director has decided that he wants this shot to be slow motion when the original plan wasn't it to be. Whenever you mess with the scale time on the DOP solver, it is going to change the way your simulation looks a bit. That is just the way of things. Here are more topics you can read up about doing slow motion.http://forums.odforce.net/topic/24445-steam-train-slow-motion/ http://forums.odforce.net/topic/24013-upres-on-a-retimed-lowres-pyro-file/ http://forums.odforce.net/topic/18214-slow-motion-dop-smoke/ The last one has more information about doing slow motion smoke from Jeff Wagner. I suggest you give it a read. In the past I have used the Scale Time and have used the time warp method both with success. I am not the one doing your shot so, I can't tell you which one is the one that will be best for you. Good luck!-Ryan Quote Link to comment Share on other sites More sharing options...
cwalrus Posted April 13, 2016 Author Share Posted April 13, 2016 Thanks Ryan for all the info - Yes actually the shot is in slow motion - but the original dust sim I did doesn't look "slow motion enough" so to speak. He wants the same shape, translation, etc, but he wants the clouds to move more slowly... does that make sense? I suspect if I simply time warp the cache, it will no longer follow the tires' emitters? anyway I'll give all those links a read, thanks! Quote Link to comment Share on other sites More sharing options...
cwalrus Posted April 13, 2016 Author Share Posted April 13, 2016 Here's a question: If you render out a cache for your sim, and then bring it back in and time warp it down to half speed, does Houdini successfully interpolate between the cache frames? Or does it stutter-step? Quote Link to comment Share on other sites More sharing options...
rbowden Posted April 13, 2016 Share Posted April 13, 2016 If you have subframe information, yes it will interpolate between frames. If not, it will stutter step. -Ryan Quote Link to comment Share on other sites More sharing options...
cwalrus Posted April 13, 2016 Author Share Posted April 13, 2016 (edited) How do you generate that in your cache? It is not working for me.... I have a pyro sim, with substeps both on the pyro solver and the dop network node itself. Then I baked that out via a ROPOutput driver. Imported the cache via file node, appended TimeWarp, and it is stutter -stepping.... ? EDIT I think I found the solution here - render out the cache with subframe increments. gonna try this now: Cache Out Sub-Frames TechniqueJust run your sim at full speed with the look you want but in the ROP Output driver for both your emitter and smoke fields, render out the sub-frame data. Set your frame range but use a fraction for your increment value to render out subframes (which in itself may change the simulation results). This may change the simulation so best to set the DOP sample rate to your sub-frame increment value as you test.You can use $N (current frame being rendered) instead of $F (which will give you floating decimal values for frames = not good) in the name of your file. See the Help Card's Local Variables for the Geometry Output ROP. Read in the sequence from disk to see things slowed down with little change in the look. Edited April 13, 2016 by cwalrus Quote Link to comment Share on other sites More sharing options...
Atom Posted April 13, 2016 Share Posted April 13, 2016 Follow up with a TimeBlend to help smooth out interpolation errors. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.