3
$\begingroup$

I'm working on a blue painted wood, which looks like this:

As you can see, the growth rings are much rougher than the base which creates this interesting 'inversion' effect. I tried to replicate that in Blender and it turned out pretty well.

But the problem is in the 'inversion' point - I noticed a strange outline around the growth rings which shouldn't be there:

Material setup is very simple - principled bsdf with blue color and a roughness map plugged in it. What do you think causes this artifact? I tried both 8bit and 16bit roughness maps made in Photoshop just in case, although I'm sure that's not the case. Here's a .blend file (version 2.79): https://drive.google.com/open?id=1xUJojzKJBK-hIHhCrmSxztRWELIgGTKM

Improve this question
$\endgroup$

1 Answer 1

Reset to default
3
$\begingroup$

I could be wrong, but I think the halo is caused by the very same phenomenon as the inversion, that it can be expected, and it's made more obvious by the low resolution of the roughness map.

Let's divide the material into three regions for the sake of argument: glossy, semi-glossy, and rough.

Where the glossy is directly reflecting a dark backgound, the rough is lighter: it's reflecting contributions from many more directions, including the lighter areas.

Where the glossy is directly reflecting a light background, like the sky, the rough is darker - its contributions include scattered reflections from the ground.

In the grey zone, around the inversion, it's possible for the semi-glossy to be brighter than both the glossy and the rough: imagine the cone from which reflection-samples are being taken. It's possible for the medium cone to contain a bigger fraction of light areas than both the narrow and the wide, especially if there is variation along the horizon. The semi-glossy area is particularly prominent in your texture-map - all of the anti-aliased region between light and dark.

Maybe not strictly part of the problem, but there's also an optical illusion going on, like the checker-shadow illusion illustrated here. The 'dark' areas near the top of your rough regions of grain are actually lighter than the 'light' areas toward the bottom, although they don't look it.

You might consider playing with a procedural wood using something like this method.

enter image description here

.. which can get the same inversion, but with a less prominent, distinct, middle-zone. The material shown above is in this .blend, and can be tweaked.

Improve this answer
$\endgroup$
2
  • $\begingroup$ Makes sense, thank you for explanation. The reason I made such a simple roughness map is because I wanted to test the 'theory' about reflective (glossy) and rough regions. The line thickness and a small number of 'rough' parts just helped me to see how it works. Of course adding more thin growth rings and some noise makes it look a lot better and realistic. And gotta say, your procedural wood shader really comes in handy here. So thank you very much again $\endgroup$
    – user57600
    Commented Apr 5, 2020 at 9:58
  • $\begingroup$ @VyacheslavTalanov I think this is actually a really interesting question.. kind of hesitant with my answer, because I think, possibly, the way the Principled shader handles the interaction between Fresnel and Specularity has something to do with it? I've been trying to look that stuff up.. maybe someone will come along and expand on this answer. $\endgroup$
    – Robin Betts ♦
    Commented Apr 5, 2020 at 10:03

You must log in to answer this question.