Material with custom falloff

[zeitmeister]

Galinette, sensational material!!!!
Carpaint!!!!!

[Polinalkrimizei]

I want to make love to that material

[Zom-B]

I'm also still testing with my "maxCosTheta()" function, here some example:

pure diffuse:

diffuse + maxCosTheta() material (50/50):

The Key here was to set some Bump / Normal Mapping on the material so the transition of the maxCosTheta material get smoothed out nicly

[CTZn]

Very nice Zom-B, it's great that you are fiddling with ISL !

Besides using bump/normal maps, wich is quite mandatory anyways for this kind material, you can also add the "maxCosTheta()" function to a somewhat normalized, dimmed fbm.

On a different note I would like to remind that what Ono has posted so far was rendered with Indigo, while Etienne is using a private software. The results are gorgeous though not as close to us Indigo users...

I'm saying this to make a little justice to Ono's work on the matter wich is a great addendum already

[OnoSendai]

Another velvet material test.

[OnoSendai]

red velvet again.

[Zom-B]

looks great Ono, as usual
Any reason why these renderings are so slow compared to the one before?

When do you thing do we get a beta to test out this new material, the Indigo competition would be great to test this material on some of the couches and seats

Very nice Zom-B, it's great that you are fiddling with ISL !

Actually I just grabbed some stuff from the Forum and did a lot of trial & error to tweak it. More like a caveman who get some "swiss army knife" in his hands... there are more possibilities hidden in there

Besides using bump/normal maps, wich is quite mandatory anyways for this kind material, you can also add the "maxCosTheta()" function to a somewhat normalized, dimmed fbm.

I would like to repeat the caveman analogy here

... while Etienne is using a private software ...

Something I would like to hear more about

[Pibuz]

@ Ono
Red velvet material is GREAT! Complete new shader or modding of existing one?

@Zom-B
No doubts that cosThetaMAT is way better than previous simple diffuse mat.
Seems like you sort of modified a little the lighting, or is it just a material response effect?

[CTZn]

Down the page 2

Yup, it's a new material type.

Zom-B: fbm is the caveman's mace ! The idea is to introduce different maxCosTheta responses for equivalent surface orientations. That would simulate bumpiness or better said, clumpiness.

[OnoSendai]

@ Ono
Red velvet material is GREAT! Complete new shader or modding of existing one?

Hi Pibuz,
It's kind of a modification of Phong, although it's almost a complete new material.

[Zom-B]

No doubts that cosThetaMAT is way better than previous simple diffuse mat.
Seems like you sort of modified a little the lighting, or is it just a material response effect?

Lighting is the same, here are some example of the function with bumpmap and without:

Zom-B: fbm is the caveman's mace ! The idea is to introduce different maxCosTheta responses for equivalent surface orientations. That would simulate bumpiness or better said, clumpiness.

As you see I only introduced some bumpiness trough the bump channel of the material, I have no clue how to include some noise function in the code for the color itself -.-

Here is the material code if somebody wan't to play with the shader:

Code: Select all

def eval(vec3 pos) vec3 :
                  lerp(
                    vec3(1.0, 1.0, 1.0),
 					vec3(-0.9, -0.95, -1.0),
                    maxCosTheta()
                   )

[CTZn]

Try

Code: Select all

def eval(vec3 pos) vec3 :
	lerp(
		vec3(1.0, 1.0, 1.0),
		vec3(-0.9, -0.95, -1.0),
		maxCosTheta() * (fbm(pos * 500.0, 4) * 0.1 + 0.95)
	 )

fbm: the noise function
pos * 500.0: the world coordinates, preferred to UV coordinates for consistency of the spatial distribution of details over different meshes.
* 0.1: scaling of the noise signal down for a subtler effect
+ 0.95: the dimmed signal is brought closer to 1.0 (maxCosTheta() * 1.0 = maxCosTheta()).

Here we are in fact using a "noisy 1.0", oscillating between 0.9 and 1.0 to be multiplied with maxCosTheta(). That is, 0.1 distributed around the threshold of 0.95 (0.1/2.0 = 0.05, below and above the said threshold).

In effect, 0.95 and 0.1 are choosen concurrently so that the signal does not exceed 1.0. Anything below 0.0 or above 1.0 would have no effect as a lerp mixing parameter (Indigo would clamp it between 0.0 and 1.0 anyways).

For instance, if you want twice a stronger effect replace 0.95 with 0.9 and 0.1 with 0.2. The signal would then oscillate between 0.8 and 1.0 times maxCosTheta().

PS: 4 is just the level of complexity or intrication of the noise function. Four octaves.

[Zom-B]

thanks a lot for the explanation and code CTZn, I'll dive deeper into it next week
Just had a first test done with 8 octaves and a 0.65 strength of the noise.
Seems it needs more testing ^^

[CTZn]

eeek. XD

[pixie]

thanks a lot for the explanation and code CTZn, I'll dive deeper into it next week
Just had a first test done with 8 octaves and a 0.65 strength of the noise.
Seems it needs more testing ^^

fmb.jpg

Do you mind sharing the material Zom-B?