[MrC11111]

Hello

I have tried to create one's with a wax material base from the library, but I'm really not a master ...
The very subtile transparency/reflection/sparkling is so difficult to achieve!
Some example of my goal (and the wax base, but nothing worked from it...)

[MrC11111]

No answer... I think that it is too simple for you.
I have the V2, and the material editor is not so easy for a guy like me, I saw that in the V3 Maxwell has made a new way to build materials, more intuitive and friendly, like me

[Mihai]

Show what you got so far, motivates people to help Usually with marble you need very high scattering and pretty big attenuation distance. Try to start with the milk preset, then decrease the saturation of the blue scattering a bit, and increase coefficient to about 2000 or more. You can play with the asymmetry to make it deeper scattering (negative asymmetry, up to -0.85).

[MrC11111]

Thank you Mihai. I will try from the sss milk and I'll follow your idea,
thank you

[eric nixon]

I had a go, It is a tricky material to make because its inherently slow, achieving its lightness from the 'depths'.

I'm getting a benchmark of around 170 on my machine for the simball which is pretty slow

I think negative attenuation is the key here (I used -.4). and a very high coeff (I used 20000, it doesn't affect rendertime, assuming you compensate by making the sss darker) keeps it solid loking in thin areas. Otherwise its a basic glass with nd1.48 ... these values are just guesses though.





I was wondering why there is discussion sometimes about adjusting attenuation depending on the size of an object?, surely marble is just marble whatever size it ends up. Personally I haven't been worrying about the scale of objects when applying sss materials and that's been working so far..... so why is it mentioned?, and I've noticed its actually a function of the translucent wizard.. Is it just a convenient way to change the look without worrying about the reality?

[eric nixon]

Simball image for reference, and some volume geometry tests, volume geo shown in the 'wire' is non intersecting with just a coating@nd1 for the material.




The volume ones render very slow, prob not a good idea to do this...

[Hervé]

the budda looks more like porcelain.. spec too tight..

[eric nixon]

Yes it looks a bit wrong. Roughness was set to zero, (it appears to render the quickest) but it can be whatever you want, although keeping the roughness fairly low is preferable speedwise.

[eric nixon]

Update, with coeff 10000 and noise mapped roughness.

[Hervé]

I think you could try a nice fine noise bump to get the roughness just right.. (?).. if I look at the photo close up marble face..

[eric nixon]

Yeah, thats exactly what I did.

noise mapped roughness.

Ah, but you mean scale it up large enough to see actually see the noise, yes that could help.

This 'marble' material looks very similar to cake icing, and a few other things probably..

[MrC11111]

Hello

Eric what you have got is 1 billion better than mine, in fact I used a basic thing without any effect, very poor at the end...
Do you think that you could share with me your final mxm?

[tom]

Try Translucent Material Assistant and begin with Flint preset. Decrease scale to 0.2-0.3 or the way you like. Eric, yes the scale is just a practical factor for achieving the desired look in different object sizes while keeping the same combinaion of parameters.

[MrC11111]

Thank you Tom.
I try with your way... But where to "decrease scale to 0.2...", I don't understand this point
And how/where to map the marble texture to achieve this ???

http://img.archiexpo.fr/images_ae/photo ... 465497.jpg

[eric nixon]

model is 48cm high

http://www.mediafire.com/download/e7enk ... marble.mxm

This is a mw3 file. I followed Toms advice, to tweak the flint preset. I tweaked it for maximum brightness.
Gave up on my previous attempt, the 'soapy' marble, it was really slow, good for porcelain maybe..