Disney vs matte

[Sharlybg]

hi guys !
Just have an issue with the disney material not operating as expected. Removing reflection component to make it diffuse break the shader behaviour.
Not only it start to badly converge (lot of fireflies everywhere). But also there is bright rim all arrpund the ball.

Blendfile

Disney vs Matte.blend

(995.82 KiB) Downloaded 179 times

[DionXein]

I checked this, it's really too noisy and bright at same time. And matte material is more timeconsuming (almost +50% to render time)

Also, what's for this hole was done? Camera got parameter know as clippin plane, so you can skip walls in rendering.

[Sharlybg]

I checked this, it's really too noisy and bright at same time. And matte material is more timeconsuming (almost +50% to render time)

rendertime doesn't really matter here as even in 3min you can't reach same noise level with disney.

[Dade]

Your parameters for matte rendering with Disney material are wrong (you are assuming that setting everything to 0.0 mean "matte", it doesn't). You can use the default Cycles parameters:

Code: Select all

scene.materials.shell.type = disney
scene.materials.shell.basecolor = 0.6 0.6 0.6
scene.materials.shell.subsurface = 0.0
scene.materials.shell.roughness = 0.5
scene.materials.shell.metallic = 0.0
scene.materials.shell.specular = 0.5
scene.materials.shell.speculartint = 0.5
scene.materials.shell.clearcoat = 0.0
scene.materials.shell.clearcoatgloss = 0.03
scene.materials.shell.anisotropic = 0.0
scene.materials.shell.sheen = 0.0
scene.materials.shell.sheentint = 0.5

and the result is:

For cross reference, check the default parameters of Cycles Principled (aka Disney) BRDF (and how it works with yours).

[B.Y.O.B.]

Note that Cycles uses a non-standard disney implementation where specular = 0 means indeed that the material is pure matte and not specular at all.
However, this is not how it's planned to be in the disney material specification.
See https://github.com/LuxCoreRender/LuxCor ... -507023023

[Sharlybg]

So according to this shader ball table what are the parameter of the last blue rough ball down right ?



Is it Roughness 1 / Specular 1 ? (and maybe metallic 1 )
Or how to read this table ?

[B.Y.O.B.]

It looks like specular = 1 and roughness = 1. No metallic. The rest probably default parameters.
But I guess they chose specular = 1 to have a more pronounced reflection for the roughness showcase, you can use lower specular values.

[Dade]

Or how to read this table ?

By row by row: each row is not related to the other ones. It is not a matrix. For instance the last metallic ball doesn't have subsurface 1.0 or roughness 1.0, etc.

[Sharlybg]

Or how to read this table ?

By row by row: each row is not related to the other ones. It is not a matrix. For instance the last metallic ball doesn't have subsurface 1.0 or roughness 1.0, etc.

So does it mean that My first shader even if not matte is physically correct ?

[daros]

Hi, here is one more test and some questions:
Given that the color is applied to the entire box (same scene as Sharlybg, only shorter aperture time) and that the light intensity is constant, how it comes that the disney shader, which is aligned to the matte shader at material color 0.7,0.7,0.7 , delivers so much brighter results starting from color 0.8,0.8,0.8 ?
If the disney material is aligned to the matte material at 0.7,0.7,0.7 should it not be aligned also at 0.8,0,8,0.8? If not from where comes the light in excess?