I´m try to undestand in a good way materrials but I find two contradictory informations:

In the Maxwell documentation I found this talking about Common glass...

"Set the Nd to 1.51 which is a reasonable Nd for glass. Following the relationship between Nd and the Fresnel effect, we can use the Fresnel formulas to calculate that the reflectance of this material (with an Nd of 1.51) near incidence angle (looking straight on at the material) would be about 4%, which translates into a RGB around 11. This is the value you should use for the Refl 0° color. Of course you do not normally have to calculate this value exactly; just know that for materials with an Nd around 1.5, the Refl 0° color has a value of around 10-20. You can raise it if you want shinier looking glass. If you want to create a very realistic glass, you should look at (images of) real glass objects for references."

They said you take this 0.04 Reflectance value and translate to RGB 255*0.04=10.2 and they said 11 ... and put it on reflectance 0 and leave reflectance 90 in 255 like in the graphic:


But in the other hand I see all the values offered on the Materials produced from the material manager and the colors are reflectance 0 = 38 and reflectance 90 = 148, not only this ones is the same for bottle glass frosted glass,water,ice,diamond,ALL translucent materials... seems a constant...


And the question is why this values? some gamma correction on the reflectance translation? only to give people a more conservative value that produces less noise (but without a scientist explanation)? it´s only that the artist thinks is a better color for this parameters but withot a scientist base only artistic? and in the end..... will be right to use this king of conversion reflectance/RGB? or not?

Thanks in advance!
User avatar
By tom
Well, this is a very long story but I will try to explain what's going on briefly. First of all, that 4% frontal reflectance of common glass and fresnel falloff are only valid for perfectly specular surfaces (roughness=0). So, these values are completely experimental when it gets glossy (roughness>0). In early versions of Maxwell, we didn't have the Force Fresnel oprion next to Nd so it was user's responsibility to compute and set proper ref0 and ref90. This is what user manual is saying and it's only meaningful for zero rough materials. Today, you don't need to know and set reflectances for clear glass anymore. Make sure Force Fresnel is on and that's all. Doing so will disregard the intensity of given reflectance and only use its hue. On the other hand, when you increase roughness, Maxwell will gradually start to use the given reflectances instead of Forced Fresnel and blend it out while blending the given reflectances in. This is why wizards prefer experimentally invented moderate values. So, Force Fresnel still ensures the correct look for roughness 0 and the custom entries define the desired reflectances for the roughest state and they get blended in and out. As a sidenote, the RGB evaluation above is not accurate due to the missing gamma in the calculation so, please disregard it. We need to update this part of the manual, thanks for the heads up! :wink:

Is there any update regarding CPU engine ? Some […]

The manual is not explicitly clear, important for […]

Hello, I agree with Jasper ... I still have this […]

Squared looking surfaces while rendering

That's the typical "shadow terminator problem[…]