Lucas Ridley

Unfortunately, the problem persists.

Changing the pad bounds fixes on one frame but introduces it on another. I can't figure out why this is happening or how to solve it.

Here's one pad bound value and then tweaking it




Then here's the same pad value between 2 different frames. The issue just pops in and out of existence on different frames even with different pad bounds.

Phlok
Looks like some broken data.

- Try exporting another FLIP sim, just a handful of frames and see if it is still broken.
- check the surface field (you can visualize it inside the particle fluid surface SOP Output > Surface VDB) does it look OK?

Thanks for the response.

I re-simmed everything with the same result. Visualizing with a surface VDB appears like everythings filled in correctly.

On playing with this further I think I found the issue, or at least a way to kind of resolve it. It was a mismatch between bounding boxes somehow. (I'm still new to houdini.)

Essentially the defaults that come off the shelf were not working.

Whatever is going on with the bounding box to cut off the sim to create this into a flat plane were grabbing thin edges at the bottom of the bounding box and extruding them along their axis like it was doing to extend the ocean surface. And the holes in the bounding box were changing every other frame which led to huge planes not parallel to the ocean surface appearing and shifting every frame.

So for now I deleted whatever expression the y bounding box value was and reduce it until it clipped those stray edges out.

It's still not clear to me why this was happening, and still the first frame of the sim has holes in it and is acting wierd which I can't figure out. But I think it's workable now.

I am using a guided ocean layer shelf setup.

It appears something is occurring on the fluid extension aspect.

The incoming cached fluid looks fine.


But at the point when that is getting particlefluidsurface somehow it's breaking.



On inspection the particlefluidsurface appears to be missing the entire mesh on some frames and then when it extends those edges it appears really broken.



The cached fluid that it's using via an objectmerge node doesn't appear to have this problem. So I'm not sure how to resolve it.