Corona Displacement Condensation Problem

[Omarb3d]

Hi all,

So this is an issue I have been plagued with for years.

Whenever I try to take the approach of using a masked out displacement (or even bump) for my condensation on a glass bottle, it completely destroys my refraction and reflections on the bottle itself. I use a 100% black and white opacity mask to create the areas, and a similar displacement or bump. This works on solid objects okay but is absolutely useless on glass and I do not understand why.

As you'll see in the screen shots provided when the mesh object with the displacement is over/above the glass the underlying glass object renders awfully, no idea. When I move the object away from the bottle you can see the difference between the two. I have spent time adjusting some of the refraction depth and such but the default settings are already so high this does nothing.

Please any help would be appreciated.

[GeorgeK]

Hey Omar, things to consider:

Do all your opaque materials share the same IOR(efraction)?
Does your liquid use Volumetric Scattering Distance?
What type of MASK png/tiff/tga is the alpha clear white 255? (check 3ds max filtering sometime messes up things)

Don't know if you can try have the droplets on a separate material layer, or even surface with on top of your bottle, where essentially only the droplets count as displaced geometry + shading, hopefully you will avoid problems. Also be sure to check shadow casting of certain objects.

[Omarb3d]

Thanks for the reply!

The screen shot you have is exactly as I have the material set up. It is a separate mesh on top of the bottle (see third screen shot). I will check on some of the things you mentioned and report back though.

Also, it looks great in the material preview and when it is alone, but the moment it is sitting above my bottle the refraction of the bottle goes crazy, doesnt even need to be touching the bottle to effect it (again see screen shot 3 for best explanation of this effect).

[Omarb3d]

1) Opaque Materials (glass and liquid inside) have different IOR because they need to, but when I turn one or the other off the effect from my masked condensation still persists.

2) Liquid doesn't use volumetric scattering, when I apply some to it the effect persists.

3) I am using but a 255 white/0 Black colored TIFF as well as the same White/Black alpha map, I have changed some of the settings on channel outputs for the corona bitmap to no avail. I have applied color correct and cranked values, nothing helps.

4) As in my last reply the condensation is as you described, separate object with a material/displacement on it. I have turned off shadows, and a ton of other settings in the object properties as an experiment and still it looks like this...

[romullus]

I think you've made mistake somwhere, because i tried your setup and it renders perfectly normally, no need to turn shadows, nor any other hacks. Just one small thing - you must make sure that glass and droplets meshes don't overlap. Simply add push modifier to droplets mesh and push it ever so slightly outwards.

P.S. sorry for suspiciously coloured liquid inside the vessel - let's pretend that it's lemonade :]

[GeorgeK]

1) Opaque Materials (glass and liquid inside) have different IOR because they need to, but when I turn one or the other off the effect from my masked condensation still persists.

2) Liquid doesn't use volumetric scattering, when I apply some to it the effect persists.

3) I am using but a 255 white/0 Black colored TIFF as well as the same White/Black alpha map, I have changed some of the settings on channel outputs for the corona bitmap to no avail. I have applied color correct and cranked values, nothing helps.

4) As in my last reply the condensation is as you described, separate object with a material/displacement on it. I have turned off shadows, and a ton of other settings in the object properties as an experiment and still it looks like this...

Hey again! glad to hear most things are set nicely, check your mesh surface normals if they are flipped, try x-form the object then back to editable. Check with the STL Check modifier for any sneaky vertices or some overlapping faces. Other than that would be great if you post some screenshot of your material setup.

[Omarb3d]

So I'm struggling here.

I have completely recreated a new scene, very simple.

Still running into this issue, flipping normals, doing all kinds of things. See attached screen shots. I am clearly doing something wrong here but I have no idea what.

Whenever I flip the normals it looks like it might be right, but then I apply my displacement and realize it's only working because the back of the condensation object is facing the camera now (displacement pushes the drops into the bottle)


------ Also I included two different objects because I made the first bottle shape using my normal methods of layering edit polys, shelling, etc, thinking that might be the issues. The other basic cylinder shape was made using none of that (except the liquid inside) and still having same issue -------

[GeorgeK]

Omar I've checked your scene, I have some stuff to suggest regarding your scene and issue:

• Try collapsing most of your modifiers on objects,3ds max is really unreliable when many geometry-altering modifiers are stacked

• When working with many overlapping surfaces that include refraction it's good the outer layer to be as "thin" as possible but not too thin like 1.0 IOR, 1.1 usually works like a charm, again depends on the preferred outcome.

• On the same note, removing shadow casting + atmospherics from the outer layer is the way to go.

I've included some screenshots and edited your scene, hopefully your refraction problem is resolved. Also on a personal note glossiness should be lowered a bit on the 2nd layer glass as it should be little bit "frosted", again depending on your preferred outcome.

[Omarb3d]

Ill take a look. Thanks!

Most of the settings on the scene provided were just kind of tossed on regarding refraction and such. I'm sure Ill have some Qs in a few minutes here!

[romullus]

@Omarb3d, actually you are right, it's not your fault, it's Corona. I was under impression, that Corona's refraction issues were solved in V4 with overlapping media fix, but apparently they are not. Basically, in Corona it's impossible to get correct result with your method. You need to do it with SINGLE mesh, multiple material IDs and SINGLE material. Either that, or scatter droplets as geometry. In attached rendering, on the left is droplets with your (duplicate mesh) method, on the center is single mesh and single material method and on the right are droplets scattered as geometry. Notice how on the left object, refraction is very wrong (not to mention that it's also the slowest to render). I'm attching the scene, so you can see how mesh and material are set-up. FresnelIOR and IOR does not neccessary have to be mapped. They make material slightly more physically acurate, but the difference is quite subtle.

By the way, if you want to do condensation on non refractive objects, then method with duplicating meshes, is perfectly valid.

Edit: and really, REALLY there's no need to turn off shadow casting or do any other non physical hacks.