CoronaPhisMtl: anisotropy. Artifacts of stretched glare

[imaginary friend]

Hello! The goal is to achieve the same anisotropy effect as in the reference.



I made a midpoly model for medium- and long-range plans.


UVW unwrapping



I'm assigning the Corona PhysicalMtl with the brushed aluminum preset.



The glare is radial by default. This is not satisfactory.


Switching local axes to UVW coordinates


Even in the preview of the material, an artifact in the form of a ladder appears.


This was alarming, but the result seemed to be similar to the truth.


You can see a staircase of some sort in the glare. Upon closer inspection, I noticed some breaks in the reflections in the area of the seams.


I applied Turbosmooth. The effect is stronger now.



In general, I hope I have described the problem thoroughly. The goal is to achieve the same glare as in the reference. Perhaps someone who has encountered this problem before can share a solution, or maybe there is another way to achieve such an elongated glare. Thank you for your attention to my question.

I have attached a file containing a teaspoon model in FBX format.

[romullus]

I can't see nothing particularly wrong in your setup, but maybe roughness and anisotropy are set a bit too high - remember at higher roughness you need lower anisotropy and vice versa. Here's my take, basically the same, but with lower roughness and texture mapped anisotropy. I don't see any obvious artifacts, but as with all metals and especially the ones with anisotropic reflection, it all depends on the environment, so test it in your scene and see if it works there.

[imaginary friend]

I can't see nothing particularly wrong in your setup, but maybe roughness and anisotropy are set a bit too high - remember at higher roughness you need lower anisotropy and vice versa. Here's my take, basically the same, but with lower roughness and texture mapped anisotropy. I don't see any obvious artifacts, but as with all metals and especially the ones with anisotropic reflection, it all depends on the environment, so test it in your scene and see if it works there.

Thank you for taking the time to answer my question. Your test looks quite good. As you wrote, the level of anisotropy and roughness is important, but deviation from these values brings back the artifact. The noise map almost masks the artifact; if you remove the noise map, the artifact becomes visible. In any case, I concluded that this artifact is a problem with the CoronaMtl parameter. The lighting does not make a difference. I also suspect that models of a certain shape may exhibit this artifact more strongly. Sadly, I leave this issue unresolved and will use your customization option, as it best masks the problem. Thanks again.