[eric nixon]

This is a single bsdf with thin-sss, using the latest beta 3.0.95. I wanted to use it for a moving flag with motion blur, so it has to render quickly, and it does
Also using a single bsdf ensures realism (although I dont know what this material actually is, other than its a woven fabric thingy)



Thanks N.L. and also cool that existing thin-sss materials retain their appearance, even though I will probably update (simplify) them for the speed-up.

[tom]

Yes, Single-Sided SSS received several serious fixes.

[eric nixon]

I have been trying to remake this hot-air balloon fabric mxm, originally the mw2 file had an extra diffuse bsdf, I removed that, anyway I got a nice result by trial and error, but I dont understand the controls at all...

thin-sss r2 bsdf + spec bsdf


Hoping for some kind of examples/docs/explains soon, fumbling in the dark..

Sunlight test at scale,.. needs some work.

[tom]

You seem to be using it nicely even without the instructions. OK, we're going to cover it soon...

[eric nixon]

Thanks, just scratching the surface really, much room for improve...

Heres the latest; a kind of glowing version (just 1 layer but 2 thin-sss bsdfs, and some other bsdf's too) I have used some extreme values, thats probably the cause of the 'glow', but no noise so perhaps legal?

[eric nixon]

Attempt at paper;

[AlexP]

Eric, I think 2 bsdfs with thinsss don't work, it seems only sss from one bsdf works and other doesn't. You can check it just by duplicating one bsdf, you should get same look as with one with 100/100 weights right?

[eric nixon]

Why dont you try the same logic, duplicating 2 standard bsdf's, ofcourse the appearance wont change, if you clone identical bsdf's, and weight them 50/50 or 70/30.... The effect is exactly the same as having one bsdf. Just rendering a bit slower.

So for the record, you can use as many thin-bsdf as you need, they will all be working.

[eric nixon]

Attempt at white balloons, using a backface material (more matte, without the coating)

[eric nixon]

Pink version with thickness mapping.



Would like to get a more transparent look but that would need a third thin-bsdf which would result in a darker material overall.

[eric nixon]

Found a problem with having the SATURATED diffuse bsdf being thin-sss, (the inverted colour of trans). So now using 1 diffuse bsdf and 2 thin-sss, one of those being very transparent (assymetry -.98)

Solution isn't elegant IMO, but looks believable;

[rusteberg]

Nice balloons

[AlexP]

Eric, but with 2 same thin-sss bsdfs look changes, that's what I mean. Material looks darker.

[eric nixon]

Material looks darker.

Yes thats true, I dont know why.

[AlexP]

Thanx - hope I've detected bug
Anyway, I found also that now when ref0 is too bright it causes lots of fireflies around and materials start to glow. I think now we have to check if ref0 is not brighter then trans and scattering but as scattering uses multiplier (scattering value) I'm not sure what to do.
I've also found that using additive bsdf work really nice (additive fix in v3) with lower layer with thinsss bsdf getting darker.
BTW really cool balloon mat especially lower part when it gets thicker.