[eric nixon]

I just tried to use a backface material with thin-sss, It works in the sense that it looks different from both sides and appears to cast a shadow which is influenced by both front and back materials, but as seen from the camera there is no effect.. so its not correct is it?

here is the rear map which should show through a little, but it doesnt at all



render..

[eric nixon]

Well, seeing as everyones too busy to answer immediately, I had another go and learnt something;



Because maxwell only appears to recognise the transmission of the mxm applied to the face that the camera is looking at. I decided to make 1 transparency map for both mxm's. (actually I made two maps with differ slightly) So basically its only the reflectance properties which differ between front and back faces.

This does seem logical in hindsight, and really needs to added to the online docs. It also means that getting a true to life result will require artistic judgement / some testing...

here is the same scene from the other side;

[photomg1]

nice work Eric , I also agree about the docs !

[rusteberg]

here's some tests with only two maps

Emitter:





(i don't know what's causing that little sliver on the corner there)



HDRI

[eric nixon]

Hullo again Rusty, good to see you.

here's some tests with only two maps

That looks perfect. Can you explain how its done? thaNKS..

EDIT; I'm guessing you used 1 map for reflectance and scatter, and the 2nd map just for trans, and then reversed that order for the backface material.
I have now tried that and its super quick to setup, but the problem I found was that it became inflexible to tweak the relationship between reflectance and trans using bsdf weight. but maybe I set it up differently to your mxm.

Anyhow heres my tweaked material sl21. 1hr..

[eric nixon]

Same mix with new maps;



I like how everything is mapped from the front, (I tend to flip horizontally the backface map in PS to simplify the mxm) thankyou NL, I think the logic is good here

And obviously I wish this money was real...

[rusteberg]

Looking good, Eric! For the example I posted I made a two layer material. The base for sss and an additive layer for reflectance. At the base SSS layer I used the map for the facing side of the bill in the scattering slot, and the backside map for the bill in the transmitance slot (flipped like you mentioned) with a low attenuation value and high coef & assym values. The additive layer just uses the front side of the bill in ref 0&90 and is weighted.... If that makes any sense....

[eric nixon]

Thats exactly what I tested as your supposed setup I had a null base layer thats the only difference, I am still observing the 100% power rule for thin-sss in mw3. Although there can be exceptions if there is a heavy texture, for example this frog, was around 110% layer weights due to the need for extremes in the texture.. just wanted to show this frog thingy really.. it doesnt have a backface material.



The grey bit under his chin are some dead flies, thats all.

[eric nixon]

And the opposite situation here with no texture and pure white colour, where going over 100% will create problems;

[seghier]

beautiful !
i try to create one:

[eric nixon]

Just needs a few tweaks, keep going.