10
$\begingroup$

I'm experiencing this weird behavior with a fluid domain.

enter image description here

It looks like something is occluding the object whilst there's nothing else in the scene.

Even if I apply the FluidSim modifier, I still having this issue. Here you have a Blend file to see it for yourself.

In addition, I've uploaded the original blend file plus the last 10 frames of the cached simulation. Anyway, you could make your own cache:

https://www.dropbox.com/sh/a3wplh6bo8iiz4n/AADafjpXM95iDiWZtLoGWRSqa?dl=0

EDIT: I've experienced this again even with a brand new simulation made in official Blender 2.79.

$\endgroup$
7
  • $\begingroup$ It would be more helpful to see the blend version before you applied the fluid, but clearly, there's a bounding box around it acting like an invisibility cloak. I'm wondering if it corresponds to a bound that you place on fluid before conversion. $\endgroup$
    – bertmoog
    Commented Aug 5, 2017 at 19:09
  • $\begingroup$ I do not know the cause, but if you open blender and then open the file without the UI for the file, then there is no error. $\endgroup$
    – user1853
    Commented Aug 5, 2017 at 19:18
  • $\begingroup$ @cegaton - Weird... I'm still seeing the problem even without loading the UI. $\endgroup$
    – bertmoog
    Commented Aug 5, 2017 at 19:32
  • $\begingroup$ It renders fine, just previews wrong $\endgroup$
    – bertmoog
    Commented Aug 5, 2017 at 19:58
  • $\begingroup$ Hi there! Thanks for your attention to my question. I've added the simulation file with some cached frames. I hope you'll find the cause of this error. $\endgroup$ Commented Aug 7, 2017 at 10:57

2 Answers 2

2
$\begingroup$

Solution - to bring Fluid Domain whole visible in a viewport:

  • under Render Properties > Performance > Dynamic BVH change to Static BVH

enter image description here enter image description here

  • or under Object Data Properties > Normals enable Auto Smooth

enter image description here enter image description here

I don't have an explanation for that, very precise pixel based cut looks like some kind of clipping, but I don't see any type active. Also I was thinking about some flipped normals that caused something similar to me once (that's why auto-smooth probably worked here), but since the cliping is straight, it can't be a faces of fluid sim rather than something that cross it, but there is nothing like that ... :) Did you used FlipFluid add-on for that?

So, mystery wasn't explained yet :) But probably future generations will find some useful string of meanings from my notes, that brings a light into this dark old 2.79x ... :)

Anyway I don't have this issue in 2.8x

$\endgroup$
2
  • $\begingroup$ Thanks for your answer! Anyway, that old file was made with 2.7x, internal fluid simulator and now we have 2.8, Mantaflow and FLIP Fluids. But I do have the same issue in 2.82 with my original file and your Auto Smooth solution works, so maybe it's useful for someone in the future. $\endgroup$ Commented Feb 23, 2020 at 12:14
  • $\begingroup$ After I wrote answer I noticed how old your Q is :) Community boot brigs it up. I didn't noticed that, but your answer surprised me even more :) so thanks for comment, BTW if you can reproduce it with a new file definitely try to post as a bug, but if its stick only with one specific file, no chance someone will care. I'm still curious what can cause something like this. $\endgroup$
    – vklidu
    Commented Feb 23, 2020 at 14:26
0
$\begingroup$

I have a similar issue but instead the fluid would disappear only from certain angles (although it seems like maybe yours also becomes visible when viewed from the top), as if there was an invisible clipping box, without top and bottom faces, around it. I had to apply the fluid modifier, export the mesh as an OBJ and import it back for the problem to go away.

$\endgroup$
2
  • $\begingroup$ But you only did a still frame of the simulation, right? $\endgroup$ Commented Jan 26, 2018 at 10:35
  • 1
    $\begingroup$ That's correct. It's all I needed. If you need an animation sequence then this won't work. $\endgroup$
    – virtualHC
    Commented Jan 27, 2018 at 3:40

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .