Issue with Reconstructing Beauty Pass: SSS Thinwalled Data Missing

[antoine_guerin]

Hi everyone,

We are working on a project with a limited budget, so we need to have maximum control during post-production. When trying to recompose the beauty pass using the direct diff and direct specular passes, one piece of information is always missing: the SSS (in Thinwalled mode). Of course, we have checked all the passes to try and recover this information, even by compositing through subtraction, but nothing works.

We absolutely need a solution; otherwise, we’ll be forced to switch to another render engine.

Thank you in advance for your responses!

versions: 3ds Max 2025 - V-Ray 6

[hermit.crab]

Just tested this on a simple scene and I don't see any difference in the result when comparing the BtB composition vs regular RGB. The thin-walled SSS info should be in the GI pass. Mind if you attach/send a simple scene so we can investigate?

[antoine_guerin]

Thank you for your reply, and sorry for the late response!

I can’t send the original scene because of a confidentiality agreement, but I recreated a basic scene to show the issue.

I’ll post a screenshot where I tried to rebuild the beauty pass using LightSelect, with DirectRaw, DirectSpecular, IndirectDiffuse, and IndirectSpecular to get the RawGI and RawLight.
Even with this setup, the Thinwalled translucency doesn’t appear in the final result, especially in the GI pass.

Thanks again for your help !

[hermit.crab]

The only thing I noticed wrong in your setup is the fact that the Reflection and Specular use multiply rather than Plus. However, there is indeed an issue, particularly with the VRayLightSelect's DirectRaw element, which seems to ignore the Thin-Walled option. I've logged it (internal bug-tracker id: VMAX-14585) for dev investigation.

[antoine_guerin]

Thanks for your reply!
in the image, it says "multiply," but it’s actually addition, i’m sorry for the confusion.
I look forward to finding a solution and if you find one I’d love to know if there’s a way to start using it in production.
Thanks a lot for your help !

[hermit.crab]

Replicating your setup, namely: (Direct Diffuse / Diffuse Filter + Indirect Diffuse / Diffuse Filter) x Diffuse Filter + Reflection + Specular, I get identical results on a simple scene. The issue lies only in the Direct Raw VRayLightSelect. If there is a difference, however, something else must be happening. Mind if you attach the scene from the screencap, so I can troubleshoot further?

[antoine_guerin]

Thanks for your reply! I might have found a solution.
I was using Direct Raw, thinking it would recompose like VRayRawLight, but the final result didn’t work. So, I tried dividing the DirectDiffuse by DiffuseFilter and recomposing everything, and with this method, I finally got the beauty pass back.

However, I don’t understand why adding the different DirectRaw passes doesn’t give me a standard RawLight result.

Attached is a quick diagram showing the recomposed beauty with DirectDiffuse and another with DirectRaw (the version that had the issue).

[hermit.crab]

So, I tried dividing the DirectDiffuse by DiffuseFilter and recomposing everything, and with this method, I finally got the beauty pass back.

This is basically the way V-Ray extracts the raw information as well, hence why for the raw element to work properly, you need to add the DiffuseFilter and regular element as well.

However, I don’t understand why adding the different DirectRaw passes doesn’t give me a standard RawLight result.

If you continue using thin-walled translucent objects, then the issue is the one already discussed and logged above. Otherwise, the sum of the raw lightselects should yield the raw lighting pass.