VrayNext : issue with fog color (object with opacity map inside a refractive one)

[S3xyKwala]

Hello everyone,


I'm working on emeralds, and I have an issue with the fog (fog color in vrayMtl).
Natural emeralds often have bubbles, inclusions and fractures in it. To achieve that, I planned to place objects inside the emerald.

It works for bubbles and inclusions (rocky object inside), but it doesn't for the fractures ("simples" objects inside the stone (vraymtl + diffuse, reflect, bump, opacity etc.)
The fog color doesn't seem to react very well with the opacity of those objects.

As you can see on the following image, even 0% opacity zones still appear lighter.
The bad result is visible in the beauty and refraction pass only, it is not a specular or relexion. And I guess it is an issue with the way the fog color is calculated, because with default settings in fog color, the inside object remains transparent.

(clic to enlarge)





Those images have been rendered with vrayNext GPU RTX (update 3.1), but it is the same with CUDA, or VrayNext (cpu).

I tried many things (almost everything I think) :

- inside object with or without a shell modifier
- GI on/off
- simpliest material for the inside object (default vraymtl with only opacity map)
- opacity slot : 8/16/32 bits opacity map, GradientRamp, map inside a vrayHDRi, vraycolor full black, etc.
- trace reflexion off
- textures in 8/half/float in the Perf section
- vraylight (dome + hdr), simple vraylight, standard lights (all types of shadows)
- normal/clip/stochastic for opacity mode (vraynext cpu only)
- energy preservation RGB/Monochrome
- fog unit scaling on/off
- object properties : visible to cam or not, etc.
- with object properties > visibiliy set to 0 (only), of course I dont get the lighter effect, but as far as I know I can't control it with a map, right ?
- vray properties : everything (matte, alpha contribution -1, etc.)
etc.


Any idea to deal with it ?

Regards,

[milva.perkovich]

Hi,

Thank you for your feedback. I have tested the situation you mentioned but I cannot replicate the problem here. Could you send over a scene to milva.perkovich@chaosgroup.com for further investigation?

Best regards,
Milva

[S3xyKwala]

Hi milva,

Thank you for testing.

Strange you can't replicate that. I get the same issue even with a really really simple setup.

Here is a really simple scene, same result. (attached scene in .zip)

edit: the issue seems to be the fog bias, if set to 0, no issue (of course, but I loose color variations)


But if you need, I can send you my original file by email.

[milva.perkovich]

Hi,

Indeed there seems to be an issue and it has been logged in our bug tracker under VMAX-9537.

However, I can propose to you some workarounds. You can disable the fog bias and map the diffuse color of the material with falloff map to get color variations or you can try scattering particles with the map you use for opacity.
I hope this was helpful and will get you results close to what you are aiming for.

Best regards,
Milva

[S3xyKwala]

Hi Milva,

Thank your for the investigation. I hope it will be corrected one day or another

I'll try what you proposed, not sure for the diffuse color trick, but particles should work if the falloff trick doesn't.

Regards,