RE_Lighting black with FastSSS2?

Hi,

I set up a simple scene (one sphere, one Rect Light, Indirect Illumination enabled with solid color for GI and Reflection) to test assembling Render Elements to match the RGB Beauty. I also created a Light_Select RE associated to the Rect Light.

First, I assign the sphere a VrayMtl. The result is good. Lighting + GI + Reflection = Beauty
And the Light_Select = Lighting. This is what I expect to see.

Then, I assign the FastSSS2 to the sphere. I got some unexpected results.
1. The Lighting RE is empty.

2. Diffuse RE x Raw_Light RE is not equal to the Light_Select RE, which is also supposed to equal to the Lighting RE (though it is empty now).
Diffuse x Raw_Light Light_Select difference between the two

3. Interestingly, now the Light_Select + GI + Reflection = Beauty

Is the black Lighting RE an intended result as the FastSSS2 is used?
Why the Diffuse_RE multiplied by Raw_Light RE does not match the Light_Select when FastSSS2 is used?
If a SSS_RE is generated, what is the composite formula such that the composite result will match the Beauty? I tried to simply “add” the SSS_RE into the comp tree but the result is incorrect.

Thanks,
Jason

scene file attached.




vray_multipass_test.zip (59.8 KB)

The Lighting element contains the diffuse lighting of the surface; since the sss2 material uses a different material model (that’s what the whole sss thing is about, otherwise it would be just a plain diffuse material), it produces a black Lighting element. Instead, the dedicated sss2 element contains the equivalent information, and it is typically “added” to the composition.

Best regards,
Vlado

Thanks Vlado. It makes a lot more sense to me now.

But what confuses me now is:
I compare the SSS_RE and the Light_Select_RE and they are different. How come?
When a regular lambertian model is involved (eg. a VrayMtl is used), Lighting_RE = Light Select_RE (these is only one Rect Light in the scene).
Whereas when the FastSSS2 is involved, the SSS_RE is the replacement of the Lighting_RE and should be equal to the Light_Select_RE, shouldn’t they?

If I chose to involve the SSS_RE into the comp tree, how should I do the get the missing elements such that the SSS_RE + GI_RE + Reflection_RE + Specular_RE + Missing_RE = Beauty?
SSS_RE + GI_RE + Reflection_RE + Specular_RE
the difference to Beauty

If I If I chose to involve the Light_Select_RE into the comp tree, how should I do the get the missing elements such that the Light_Select_RE + GI_RE + Reflection_RE + Missing_RE = Beauty?
the difference to Beauty.

Thanks,
Jason



I’m curious as to whether you found the correct formula. I have a similar situation now.

V Miller

See my post here: http://www.chaosgroup.com/forums/vbulletin/showthread.php?69928-Render-Elements-gt-Compositing&p=560743#post560743

It contains an explanation how to combine the render elements, including the sss2 one, into the final image.

Best regards,
Vlado

Not sure if it’s just me, but I don’t have the privilege to view those jpg attachments and I can’t reply in Softimage forums. I am wondering if it’s only viewable for Vray for Softimage owner?

Same for me, i can see nothing from XSI forums.

Same here. I was checking the thread out before you posted the link here Vlado and I don’t have priveleges to any of the images of your nuke script.

V Miller

Hm, that’s odd. Anyways, will see if I can re-post them here.

Best regards,
Vlado

I think one needs to divide sss by diffuse to get raw sss first then add that raw sss to raw lighting otherwise the sss contains parts of diffuse map slso which is undesireble if you use raw passes.

Just plus one-ing that, i can’t see them either…

Chris

You should be able to see the attachments now (you might have to log out and log in on the forum again).

Best regards,
Vlado