[bjorn.syse]

Wow, if someone out there knows a good analogy or simple way to explain how to think when creating SSS materials, please explain this for me. I find the settings found in the SSS component, and their explanation in the manual some of the hardest to understand within the maxwell material model.

Have a look at two of the preset materials for example: Silicone gel, and toothpaste gel.



In my eyes, these are two materials with a pretty similar look, except for the color. If you have a look at their respective SSS settings though, they differ quite a bit.

Silicon gel has an Absorption coeff. at 5 together with a pretty dark transmittance color (0,25,50)
Toothpaste gel has an Absorption coeff. at 400 together with a very light transmittance color (0,229,255)

This makes the whole thing very hard for me to get. I'm not a physicist, but i don't think I should have to be either. The presets are great though. Without them I would be more lost than I allready am..

What are your thoughts about this SSS model?

regards,

- Björn

[killian2828]

The toothpaste in Madrid is made out of Silicone.

[tom]

The toothpaste in Madrid is made out of Silicone.

Well, it's because the numbers are coefficients to colors. It means they are being multiplied by the R,G,B entries. For example, RGB 100,100,100 and Coef 10 is equal to RGB 50,50,50 and Coef 20 or RGB 200,200,200 and Coef 5 and so on. Final color is not an input as in BSDF model, so it's not possible to load silicone gel and turn it into toothpaste gel only with rotating the colors. In SSS model, the light enters, is being scattered and absorbed, then the final color appears as a result of this complicated interaction. So, it's very unpredictable taking a material as base and only changing its color. That's also because "color" is not a separate property of light. I suggest you making some more exercises in order to see the relation between colors and the balance between scattering and absorption.

[bjorn.syse]

Hmm, well, I'm sure all that makes sense in some way, but not in any user-based way.

I mean, even If I study the manual dead-slow, I might be able to understand what a particular setting actually means and does, but since there are a few, and they all work together in this complicated interaction - it is VERY hard to get a grasp of the total impression.

It is not intuitive in the users/human's eye or way to think about material properties. Look at the Flint preset for example. It is white in my eyes, but the colors set in it, are orange and turquoise. HOW would any normal user that set out to create a Flint material, ever come up with those settings unless they are either physicists or just doing it by trial and error.

- Björn

[Bubbaloo]

just doing it by trial and error

That's the best way to learn anything.

But I understand your point. It is complicated and unintuitive. I guess that's the nature of physically correct SSS. Imagine how it would be to use SSS without the presets!

[tom]

It is white in my eyes, but the colors set in it, are orange and turquoise.

But there's nothing "white" in reality.

[bjorn.syse]

No, but in practice, and in peoples way of thinking - there are a lot of white stuff hanging around

Bubbaloo - Right, without the presets SSS would be almost unusable. Which says something about the level of complexity.

What I'm saying is just that, it would be great if the interface could be simple and intuitive, and those values could transfer into this physically correct model. In my humble opinion, that is is what makes good software awesome. A layer of simple and intuitive controls translated into hidden complexity.

[marcotronic]

What I'm saying is just that, it would be great if the interface could be simple and intuitive, and those values could transfer into this physically correct model. In my humble opinion, that is is what makes good software awesome. A layer of simple and intuitive controls translated into hidden complexity.

QFA!

Marco

[tom]

What I'm saying is just that, it would be great if the interface could be simple and intuitive, and those values could transfer into this physically correct model. In my humble opinion, that is is what makes good software awesome. A layer of simple and intuitive controls translated into hidden complexity.

I think it's already working just like you said. Take a look:



Initially, we've only had the Advanced Panel which requires no internal conversion. Then we reduced this model to something more visual and the Intermediate Level born. Without intermediate level, you had to enter long decimal numbers which makes visually no sense. Furthermore, we decided to make something more straightforward and the presets born for the beginners. Today, we really have a good range of realistic presets.

About your specific question, you mean "white".. OK, have you tried setting both colors to white/gray? Does it render "red" then? No, I don't think it's that much unintuitive.

[Mihai]

First part of the confusion may come from not knowing if you should use high absorption and low scattering or vice versa. For example it may seem strange that the marble and porcelain presets have such high scattering but low absorption.

But when you think about it, if you were to have a very thin slice of marble, it would let a lot of light through it. It's just that it scatters that light so much it quickly looks opaque when it's thicker.

Sor for materials that should look opaque but allow the light to scattered to a larger depth inside the object, it's better to use a high scattering value and not so much absorption.

For a material that would 'block' the light quickly and not allow the light to be scattered deeply inside, use a higher absorption.


Second part of the confusion, which colors to set? How can milk have such a blue scattering color and yellow absorption, when mostly it turns out white. Well if you look closely at a glass of milk you will see that in the thinner parts it does have a blueish tint, and eventually turns yellowish/white. Why this happens in reality is how the light interacts with the different types of molecules that make up milk, they can absorb one wavelength, but can scatter another.

Too abstract I know, but in my mind I think of the scattering color, the color that will be visible even in thinner parts, it's like the 'first' dominant color, and absorption color is the color you get deeper inside the object.

Make a test for example to see the difference. Set abs/sca coefficients to the same amount, use grey for transmittance, red for scattering, then grey for scattering, red for transmittance and examine the differences.

[gadzooks]

This screams a video tutorial an the think site???!!!. Anyone who understands it up to making one?

[Asmithey]

Mihai,

A very understandable, and in my minds eye, visual explanation.

Thanks,

Aaron

[itsallgoode9]

Is there a way of determining any of these values or colors based on real world measurements of an object (like how index of refraction is)?

or we just need to tweak these values and colors based on our eye?

[Mihai]

You can use the Advanced SSS Wizard if you have real world measurements, just enter them there and the final colors and sca/abs coefficients will be calculated from those. But the problem is there isn't much of this data around....