Satin/matte car paint

[jackgriffirths]

Hi,

I’ve been looking all over the internet for a satin car paint and no seems to have made it at least not publicly. I’ll add a few screen shots below of the sort of paint I’m trying to go for.

If anyone could give any advice on how to create this that would be fantastic!

Cheers!

[maru]

Have you tried creating a simple chrome-like material (Diffuse level 0, Reflection level 1, Fresnel IOR 10 or more) and just lowering reflection glossiness?
The material is probably more complex in reality, so you could also try mapping reflection glossiness with a Falloff map to have sharper glossiness on the grazing angles, but it *might* be pretty simple. In the reference photos it even looks a big CG-ish. :)

[John.McWaters]

Have you tried creating a simple chrome-like material (Diffuse level 0, Reflection level 1, Fresnel IOR 10 or more) and just lowering reflection glossiness?
The material is probably more complex in reality, so you could also try mapping reflection glossiness with a Falloff map to have sharper glossiness on the grazing angles, but it *might* be pretty simple. In the reference photos it even looks a big CG-ish. :)

Can you please explain why I often see a Falloff map plugged into the reflection color input for metals? Usually, for a colored element like gold, it will be set up for a white reflection on the grazing angles. Is this just an preferred aesthetic adjustment, or are there actual physical properties that cause this to occur?

[Juraj]

Have you tried creating a simple chrome-like material (Diffuse level 0, Reflection level 1, Fresnel IOR 10 or more) and just lowering reflection glossiness?
The material is probably more complex in reality, so you could also try mapping reflection glossiness with a Falloff map to have sharper glossiness on the grazing angles, but it *might* be pretty simple. In the reference photos it even looks a big CG-ish. :)

Can you please explain why I often see a Falloff map plugged into the reflection color input for metals? Usually, for a colored element like gold, it will be set up for a white reflection on the grazing angles. Is this just an preferred aesthetic adjustment, or are there actual physical properties that cause this to occur?

Metals can reach very high total reflection at their grazing angle (towards the edge). So for gold for example to be physical, simply using gold color alone will clamp some of that reflection.

The difference for most metals isn't very high since metals are already so very reflective to begin with, but it can make stronger visual difference when the metals get rough (lower glosiness) and naturally start loosing some reflection strength and their edges will start darkening. Using that fallof in reflection slot will prevent excessive energy loss, simulating the real metal fresnel much more closer, so it is a physical thing.

Of course, it's possible to use plugins like ComplexFresnel to simulate this exactly, but the difference between that and simple fallof curve is 99.999 visually identical. No need to sweat it.

[John.McWaters]

Have you tried creating a simple chrome-like material (Diffuse level 0, Reflection level 1, Fresnel IOR 10 or more) and just lowering reflection glossiness?
The material is probably more complex in reality, so you could also try mapping reflection glossiness with a Falloff map to have sharper glossiness on the grazing angles, but it *might* be pretty simple. In the reference photos it even looks a big CG-ish. :)

Can you please explain why I often see a Falloff map plugged into the reflection color input for metals? Usually, for a colored element like gold, it will be set up for a white reflection on the grazing angles. Is this just an preferred aesthetic adjustment, or are there actual physical properties that cause this to occur?

Metals can reach very high total reflection at their grazing angle (towards the edge). So for gold for example to be physical, simply using gold color alone will clamp some of that reflection.

The difference for most metals isn't very high since metals are already so very reflective to begin with, but it can make stronger visual difference when the metals get rough (lower glosiness) and naturally start loosing some reflection strength and their edges will start darkening. Using that fallof in reflection slot will prevent excessive energy loss, simulating the real metal fresnel much more closer, so it is a physical thing.

Of course, it's possible to use plugins like ComplexFresnel to simulate this exactly, but the difference between that and simple fallof curve is 99.999 visually identical. No need to sweat it.

Thanks for the input! When you say 'total reflection at their grazing angle', are you referring to glossiness or reflection color? Most of the time, I see a falloff map plugged into reflection color so at the edges, the reflection will go more white. Additionally, are you saying the phenomenon occurs similarly to non-glossy metals, say satin bronze or brass? Or do they lose their reflection color at their edges (going darker)?

[Juraj]

In this case I am talking strictly just reflection, not glossiness.

Whether metal is glossy or not, it's the very same material with the same behavior, only that the benefits of having reflection setup with grazing angle intensity shows its benefits more in rough versions. With mirror-polish (0.9-1.0) metals the difference is there...but hard to see :- ). But with rough material (<0.9), you will start to see edge darkening if you didn't setup the metal with full grazing angle intensity.

This setup shouldn't be very noticeable, you can setup the fallof with IOR 1.6, you shouldn't be able to distinguish the metal color suddenly changing from gold to white, it's more to prevent edge darkening.

Fallof in glossiness slot creates completely different phenomenon, that is to simulate microfacets, the surface appearing more mirrory(less rough) at grazing angle, something the current shader doesn't do perfectly by itself.

Fun fact for car shader tests: While metals don't darken along edges, for car paints only the flake treatment is metallic. But what does cause edge darkening is the last coating, which can be micrometers thick only but will do both refraction and inter-reflection that will cause loss of energy.
Coating can currently be simulated in shader only with multi-material but that one only blends between mirror at edges and paint underneath. It doesn't physically simulate these layers actually being beneath each other on shader level. But once we actually copied the car shell, applied push modifier to it so that the "coat" became physically another object on top and applied the coat material not as mirror but actual glassy/plastic shader, the edge darkening become just like in reality. It was super cool.

[John.McWaters]

In this case I am talking strictly just reflection, not glossiness.

Whether metal is glossy or not, it's the very same material with the same behavior, only that the benefits of having reflection setup with grazing angle intensity shows its benefits more in rough versions. With mirror-polish (0.9-1.0) metals the difference is there...but hard to see :- ). But with rough material (<0.9), you will start to see edge darkening if you didn't setup the metal with full grazing angle intensity.

This setup shouldn't be very noticeable, you can setup the fallof with IOR 1.6, you shouldn't be able to distinguish the metal color suddenly changing from gold to white, it's more to prevent edge darkening.

Fallof in glossiness slot creates completely different phenomenon, that is to simulate microfacets, the surface appearing more mirrory(less rough) at grazing angle, something the current shader doesn't do perfectly by itself.

Fun fact for car shader tests: While metals don't darken along edges, for car paints only the flake treatment is metallic. But what does cause edge darkening is the last coating, which can be micrometers thick only but will do both refraction and inter-reflection that will cause loss of energy.
Coating can currently be simulated in shader only with multi-material but that one only blends between mirror at edges and paint underneath. It doesn't physically simulate these layers actually being beneath each other on shader level. But once we actually copied the car shell, applied push modifier to it so that the "coat" became physically another object on top and applied the coat material not as mirror but actual glassy/plastic shader, the edge darkening become just like in reality. It was super cool.

Can you explain why I'm seeing some metal materials setup with a corona color set to pure white being plugged into the Fresnel IOR slot? A lot of time, when I see metals setup like this, they're reflection glossiness is set to something around .7. Is this to essentially disable the Fresnel IOR and use glossiness to create the surface condition?

I really appreciate your input on these topics. I'm still trying to wrap my head around the physics of it all.

[Juraj]

Can you explain why I'm seeing some metal materials setup with a corona color set to pure white being plugged into the Fresnel IOR slot?

This is bit unnecessary hack for those who want absolute (even if it's impossible to see the difference 99.99perc. of times) disabling of Fresnel.
By default, you disable Fresnel by using 999, but even that is not 'total' disabling but....you almost cannot tell the difference. Even Fresnel 40 is almost fully flat line.

You disable Fresnel if you want to fully recreate it manually by using curve in Reflection Slot.

Physically making material like chrome could look following:

Disabled Fresnel (999). Fallof curve in reflection slot, IOR 1.6 or something low, the "black" slot would be the frontal reflection. Chrome has reflection albedo of roughly 60perc. or something like that. So you would put in color that is 0.6 in linear value. But since Chrome like every metal gets almost fully reflective at grazing (edge) angle, you would put pure white color into the "white" slot of fallof map. This will be very reality-mimicking shader setup. This takes like 2-3 minutes at least to setup, bit faster in Slate editor, and lot more if you use textures instead of clean colors (metals are not pure).

And this is what "hack" solution would look, it will be 95perc. identically looking, but not absolutely real:

You don't disable Fresnel, instead you just lower it to simulate the curve, using any low number like 4 to 12 for example.  Instead of few minutes, this just took 5 seconds to do.

For some metal materials, the difference between "real" and "hack" shader setups will be none, nothing perceptible by eye, esp. highly reflective materials like Aluminium, Silver, etc..
But for others, like blackened steel, brass, bronze, gold, etc... the differences will start to become more obvious.

It depends how important the shader is. If it's chrome chair legs, use hack. If it's visualization of golden bracelet in studio setup, use as real as possible setup because it needs to look like photography. Horses for courses :- ).

{Disclaimer: There is *third* way of using Fresnel, you can do it for example in Vray or Fstorm, where you can use negative values (less than 1) to simulate the extinction coefficient ('K'). I don't think you can do this in Corona, which only uses the simplified Fresnel (only using 'n' - refractive index number). RefractiveIndex is serious business :- )

[Njen]

Just to back up what Juraj has said, in earlier versions of Corona, it wasn't possible to enter in a very high number in the Fresnel IOR value to 'disable' it, so the easy work around was to use a Corona colour node. Now you can enter in really high numbers, there is practically no need to use a colour node anymore.

[Njen]

{Disclaimer: There is *third* way of using Fresnel, you can do it for example in Vray or Fstorm, where you can use negative values (less than 1) to simulate the extinction coefficient ('K'). I don't think you can do this in Corona, which only uses the simplified Fresnel (only using 'n' - refractive index number). RefractiveIndex is serious business :- )

Numbers between 0 and 1 are exactly the mirror value of numbers from 1 to n. For example, both of these values will give the same Fresnel look in your shader: 0.657 and 1.52. The math to convert a Fresnel value above 1 to below one and vice versa is essentially 1/n (1 / 1.52 = 0.657). The other way is true too: 1 / 0.657 = 1.522

This is why IOR maps in substance appear to always be between 0 and 1 and still look like a surface that has IOR values above 1.

[John.McWaters]

Can you explain why I'm seeing some metal materials setup with a corona color set to pure white being plugged into the Fresnel IOR slot?

This is bit unnecessary hack for those who want absolute (even if it's impossible to see the difference 99.99perc. of times) disabling of Fresnel.
By default, you disable Fresnel by using 999, but even that is not 'total' disabling but....you almost cannot tell the difference. Even Fresnel 40 is almost fully flat line.

You disable Fresnel if you want to fully recreate it manually by using curve in Reflection Slot.

Physically making material like chrome could look following:

Disabled Fresnel (999). Fallof curve in reflection slot, IOR 1.6 or something low, the "black" slot would be the frontal reflection. Chrome has reflection albedo of roughly 60perc. or something like that. So you would put in color that is 0.6 in linear value. But since Chrome like every metal gets almost fully reflective at grazing (edge) angle, you would put pure white color into the "white" slot of fallof map. This will be very reality-mimicking shader setup. This takes like 2-3 minutes at least to setup, bit faster in Slate editor, and lot more if you use textures instead of clean colors (metals are not pure).

And this is what "hack" solution would look, it will be 95perc. identically looking, but not absolutely real:

You don't disable Fresnel, instead you just lower it to simulate the curve, using any low number like 4 to 12 for example.  Instead of few minutes, this just took 5 seconds to do.

For some metal materials, the difference between "real" and "hack" shader setups will be none, nothing perceptible by eye, esp. highly reflective materials like Aluminium, Silver, etc..
But for others, like blackened steel, brass, bronze, gold, etc... the differences will start to become more obvious.

It depends how important the shader is. If it's chrome chair legs, use hack. If it's visualization of golden bracelet in studio setup, use as real as possible setup because it needs to look like photography. Horses for courses :- ).

{Disclaimer: There is *third* way of using Fresnel, you can do it for example in Vray or Fstorm, where you can use negative values (less than 1) to simulate the extinction coefficient ('K'). I don't think you can do this in Corona, which only uses the simplified Fresnel (only using 'n' - refractive index number). RefractiveIndex is serious business :- )

This is a great breakdown, and I'm able to follow the majority of it. Thanks for taking the time to type it out.

For the more realistic setup, would the curve in the fall-off map (the one plugged into the reflection color slot) need to be edited much from the default curve in order to get realistic reflections?

[Juraj]

I use Fresnel curve (1.6), with no editing. This is not a curve that resembles the actual fresnel curve of the metal because my "black" slot (incidence/frontal angle) is already the specular albedo.

There is second approach, where you leave "black" slot black, and you will make the full curve manually, this was popularized few years ago by Grant Warwick, but it's largely unnecessary now.

You can look at this graph that shows the fresnel curve for different metals but also non-metals. Iron is not so far from chrome, it starts at 55perc. specular albedo (0.55 Linear value grey color), goes mostly flat, has slight dip and then turns towards full 100perc. specular reflection (1.0 linear white color).
Making curve like this would be annoying for grey metals, but for colored metals you would make three curves. So it's much easier to simply place the specular albedo (basically the color of the metal) into "black" slot of fallof (incidence angle) and keep default fresnel. The result is incredibly close.

[John.McWaters]

When you refer to setting the fresnel curve to 1.6, are you referring the the grayed out setting under "mode specific parameters"?

edit: I just realized I need to change the fall off type from perpendicular/parallel to fresnel. I uploaded some screenshots of the chrome material setup you describe.

[romullus]

Diffuse level should be 0, reflection level should be 1. Glossiness level could (should) be lower than 1 or mapped. Everything else looks about correct.