viki164

Thanks. I have been having success making other float field work like temperature and heat ! I will try to isolate the velocity vdb to see if it produces cleaner result as compared to houdini volume.

Hello wonderful community, I wonder if this is a known issue when using retime SOP with VDB's ? I am experiencing some juddering in the velocity and pulsing effect in the beauty renders. I am retiming by 0.1 speed with no sub frames caching. I imagined the new Retime node is suppose to be smart and doesn't need any incremental sub frame caching ? On the other hand when I converted those vdbs back to Houdini Volume , added a volume resize node with a bound along with some padding and then used Retime SOP with blend mode set to Advected. To my surprise the velocity channels were properly interpolated and the renders were less flickering but at the expense of heavy cooking time and rendering. I did convert those houdini volumes back to vdb after retime sop. The Retime sop doesn't fully support vdb for velocity interpolation to my experience ? I will be happy to be proved wrong and I wonder if others also have similar issues when doing a retime with 1/10th of its speed ? Do you guys add a large static bound to your volumes before converting it to vdb or just volume resize node with some padding ? Thank you ! p.s : retime with interpolate velocity often produces empty vdbs in velocity channel when used with vdbs but works fine with houdini volumes.

Use Vdb advect node ? isolate the vel channel and connect it to the second input of vdb advect node. you can resample the incoming volume using vdb resample and then connect it to the first input of vdb advect node.

hi Charles , I exactly know what you mean and what you are trying to achieve. For ref have a look at this guys work :

Apologize I don't have anything handy at the moment. The test I did is at work . The setup is quite simple though. I don't have any complicated graph Just few nodes because Pop Grains alone itself does the trick. I made a small geometry patch on the ground plane. Then used pointsFromVolume & Added another AttrWrangle node for randomizing pscale & mass. f@pscale = chramp("pscale", rand(i@ptnum)) * 0.015; f@mass = f@pscale / 0.55; These points were fed to Pop Network as First Source Input. Now inside the Pop you put a Pop Grain Node. The settings for PopGrain to imitate Foam behaviour are : Particle Separation : 0.05 Constraint Iterations : 60 Behaviour Friction : With Colliders : 0.2 With Particles : 0.1 Accurate Friction Checked Static Threshold : 0.5 Scale Kinetic : 4.5 Internal Collision : Weight : 1 Stiffness : 1 Enable Mass Shocking Scale = Local Shock Scaling Power = 100 & Shock Axis : 0 1 0 Clumping : Weight : 2 Stiffness : 0.5 Explicit Constraints : Weight : 0 Stiffness : 100 Target Pins : Weight : 1 Stiffness : 0.5 SOLVER SETTINGS : Max Neighbors : 35 Neighbor Query Scale : 2.75 Max Accelaration : 5 Velcoity belnd : 5 Over Rlaxation : 2 & Open CL is checked ON Just by these settings it should give you the foam behaviour & clumpiness. Add the ground as static collider. I also Added a bit of Gravity : -9.80665 * 0.3 & Pop Wind : // Pass Through wind = wind; airresist = smooth(0, 1, fit(v@P.y, .02, .22, 0, 1)); Amplitude : 0.25 , Swirl Size : 0.5 , Pulse Length : 20.. Just by these few nodes..you might get the foam behaviour. Reference :

Hi Hernan , you got me.. yes its been Six months I did the switch. From SOUP to Houdini..life is much easier & nice. Still learning & exploring. I haven't used the Pop Spin yet ..will give a try. Thanks P.s : Not looking for Per-particle spin instead I would like to spin the clumps !

Hello folks, Newbie here. I am trying to make some foam particles with POP Grains. So far the motion is coming along well what I am missing is the particle spin based on clumps it has formed. In Maya Nparticles the particles rotation is computed dynamically. I wonder if there is a way to do the same with POP Grains & link the rotation or spinning speed based on the clumps. Any Idea or workaround would be great ! Thanks, Vik