skin shader a la Alsurface pleaseeee

[AndresAhumada]

Yes, I will add the bump delta scale.

Best regards,
Vlado

hi Vladoo!

Is it possible to do some tutorial about the material?


Cheers

And tThanks!

[sharktacos]

hi Vladoo!

Is it possible to do some tutorial about the material?


Cheers

And tThanks!

TA DA!
http://www.anderslanglands.com/alshaders/tut_skin1.html
http://www.anderslanglands.com/alshaders/alSurface.html

[vlado]

Ok but why is the VRayScatterVolumeMtl so slower and most of all why can't I get blurred refraction with it (when I look at the silhouette of an object seen through it it doesn't seem to be blurred at all, it stays sharp)?

For the glossy refraction, the volume material doesn't do refraction by itself, but you can put it in a blend material as the base, and a glossy refractive material as the coat. Then you will get glossy refraction.

As for the slowness, it is probably due to the different way the scattering is implemented in FStorm. Attached are three renders of a scattering material with the same scatter/absorption colors. Barring the different internal color space (sRGB vs wide RGB) that causes differences for V-Ray vs Corona, they are at least somewhat in agreement, with a blueish fringe that goes to orange/red as light goes deeper into the material. The FStorm material, even though it has the same settings as the Corona material, renders very differently and light doesn't seem to penetrate as deep into the material, even though I've set the bounces to 25 and increased the GI clamp (without those modifications, the result was even darker). All images rendered for 1 minute (single i7-4471 @ 3.50 GHz for V-Ray and Corona, 2x GTX 980 for FStorm). After playing a bit more with the FStorm implementation, it doesn't seem to be entirely physically accurate.

Best regards,
Vlado

[arnol2x]

Hi Vlado,

I'm super stoked you took the time implementing the shader into Vray! And I'm sure many character artist and alike will appreciate that. Meanwhile, testing out the maya shader on some linux machines I got this interesting behavior of mixed patches, this happens both on adaptive and progressive. Widows version of the shader works like a charm though.



Keep the great work coming!

[vlado]

Meanwhile, testing out the maya shader on some linux machines I got this interesting behavior of mixed patches, this happens both on adaptive and progressive. Widows version of the shader works like a charm though.

Can you show me the settings or post the scene?

Best regards,
Vlado

[AndresAhumada]

Jajajajajajajajajaja Many Many Thanks!!!!

[arnol2x]

Can you show me the settings or post the scene?

Attaching the scene file.

Cheers!

[sharktacos]

So here's something to consider for the Vray version of the ALsurface shader:

Would folks prefer to have the "roughness" parameters on the two spec lobes be inverted so they are "glossiness," which us Vray folks are more familiar with? This would make the ALsurface shader fit in better with all the other Vray shaders, so I would vote "yes" personally. YMMV.

[vlado]

So here's something to consider for the Vray version of the ALsurface shader:

Would folks prefer to have the "roughness" parameters on the two spec lobes be inverted so they are "glossiness," which us Vray folks are more familiar with? This would make the ALsurface shader fit in better with all the other Vray shaders, so I would vote "yes" personally. YMMV.

I thought about it, but I decided not to do it for the moment - I don't want to add more confusion or overcomplicate the shader. Also, I can brag about how the shader is a PBR shader because it uses roughness instead of glossiness, and people would go "oooh, wow" (ok, that was sarcasm...)

Best regards,
Vlado

[jstrob]

For the glossy refraction, the volume material doesn't do refraction by itself, but you can put it in a blend material as the base, and a glossy refractive material as the coat. Then you will get glossy refraction.

As for the slowness, it is probably due to the different way the scattering is implemented in FStorm. Attached are three renders of a scattering material with the same scatter/absorption colors. Barring the different internal color space (sRGB vs wide RGB) that causes differences for V-Ray vs Corona, they are at least somewhat in agreement, with a blueish fringe that goes to orange/red as light goes deeper into the material. The FStorm material, even though it has the same settings as the Corona material, renders very differently and light doesn't seem to penetrate as deep into the material, even though I've set the bounces to 25 and increased the GI clamp (without those modifications, the result was even darker). All images rendered for 1 minute (single i7-4471 @ 3.50 GHz for V-Ray and Corona, 2x GTX 980 for FStorm). After playing a bit more with the FStorm implementation, it doesn't seem to be entirely physically accurate.

Best regards,
Vlado

Very interesting comparison, thanks Vlado for investigating this. For me that blueish fringe (if that's what I think) may be more physically acurate but it's most of the time blocking me to get the result I want to get. The colors get diffcult to control when there is the complimentary color inviting itself to the party. My guess is that the alsurface SSS shader is probably more like the FStorm SSS don't you think so?

About the blend material to get a blurred refraction with the vrayscattervolume I was suspecting it but didn't have time to try it, I will try this soon I hope.

[jstrob]

And by the way with fstorm I had to play with the translucence together with the refraction and volume settings to get a nice SSS effect.

[sharktacos]

I thought about it, but I decided not to do it for the moment - I don't want to add more confusion or overcomplicate the shader.

makes sense.

Also, I can brag about how the shader is a PBR shader because it uses roughness instead of glossiness, and people would go "oooh, wow" (ok, that was sarcasm...)

LOL, I get that comment from my students all the time!

Still, there are a lot of nice PBR-ish things about the shader, like microfacet Fresnel for instance. Have you considered integrating stuff like this into the VrayMtl?

[vlado]

Yes, the glossy Fresnel will be included in the VRayMtl material.

Best regards,
Vlado

[Midiaeffects]

Vlado, +1 for adding VrayALMtl, or at least the scattering models in it. They are very good, and vastly superior to VrayFastSSS2 at least in the few comparisons I did.

I think VrayScatterVolume to be the best general purpose SSS method for most materials except skin, so it's important for RT GPU to support it. Other renderers like Octane or FStorm do it, and it's very fast. How about adding it to VrayMtl as a translucency method? This way we don't need to create a VrayBlendMtl.

I had this old model/shader from the 2.4x days which was kind of a pain to get even something I'm ok with, so I decided to make a comparison of different SSS methods we have in Vray today. Also was a nice test for the new VrayALMtl. I know neither the model or the render are realistic or pretty (it's a test anyway) but it's good to look at something else other than skin.

-Eugenio

[sharktacos]

Midiaeffects,

Mmmmmm, yummy cookies! I'd be interested to see them without spec so we can observe just the difference in sss.