Phone Backpanel Shader

[fixeighted]

I did a couple of tests. Apparently the way a material like this is made is to laser cut a pattern into the surface.
So this shows just a basic glass shader, (though blinn because microfacet is broken a bit for stuff like this) with a b/w lines bump map.
It's a bit slow but has some interesting results. Not sure how to achieve the multi refractions as in that video but it's a reasonable start.
I'm testing the apparently real-world patterns derived here https://www.thorlabs.com/newgrouppag...tgroup_id=9026
and they are interesting but it is a bit of a rabbit hole

[fixeighted]

It seems that a bump doesn't really do the job well enough, so I created geometry of the necessary saw-tooth and other variations and it works great.
If anyone want to play then here is the small scene -use any hdr you like. https://www.dropbox.com/s/j4xwvuo1lo...METRY.max?dl=0
I still can't quite understand how the box pattern of offsets is done in reality, so am interested in theories, as the real surfaces can only have parallel grooves of whatever profile shape.

EDIT: So reading the paper that Antoine Toisoul wrote on his approach, he states that the effect is done, quote;

''using a single texture lookup with a diffraction lookup table without knowledge of the microgeometry''

So the question is, can this be usable with a Vray shader and if so then how?
The lens effects parameter seems to do exactly this, so seems tantalisingly close....

[fixeighted]

Another test, based on chatting with the guy who wrote the paper
and trying to figure out exactly what was happening in reality.
The settings used to achieve this image are unrealistic in terms of real-world values, though they still produce the result.
For instance, the refractive ior is 0.1
I've played with the various values for this plus the abbe number and the tiling of the displacement texture and this is the most pleasing outcome.
The shader is simply reflective glass with these values and the geometry is displaced with a pattern derived from a photo of a laptop screen
which was within the original paper, basically small dots. The light source is a small sphere light.
It matches, in a way, what the paper suggests, though is not exactly the same.
Whether that is dependent on any one factor is hard for me to determine and I'd wlecome more input, especially from someone from Chaos
who probably knows the detail.
I can share these files if it's of interest to anyone

[Slazzo]

You don't need displacement at all to represent the pattern, better use shading normals for each glare direction, this was my attempt few years ago, unfortunately not VRay

[fixeighted]

Interesting...can you elaborate more on how you achieved this please, in a way that might translate to Vray?