Anisotropy

I’m preparing a satin metal, I find this tread:

on resolved bug.

Using about the same material showed in that tread (with or without a map for anisotropy), I still have bad results.
In the captured video image, you can see a quick set with some primitive and a modeled sink.
In the primitive geometry, you can see a problem on the top of the chamfered box. Just on the top and bottom faces. If I remove any smooth group from top faces (the box is divided 3x3x3) the problem disappears.
The material seems to work ok on the other primitive.
Same story in the sink, all around the borders it create a “false edge”. I’ve tried to change the geometry, to simplify or to increase subdivision, to change smoothing group, to add a a smooth modifier, to remove all smoothing group, but none has changed. I mean, the problem still remains.
I attach also the sink without anisotropy, the geometry works fine. And the box with top faces removed from smooth group.
Is it me? Any suggest?
Thx.



what build are you using? Dailies have new anisotropy which obeys local space

Sept 24, as you can see in the frame buffer, on max 2016

I download the last build and retry immediately.

Same results, I attach the render and the scene, if someone want to have some test.


Anisotropy test build 2015-10-06.zip (97.9 KB)

Does your material’s anisotropy rotation is set to UVW space by any chance?

Did you by any change create this material using older version of Corona?
edit: ok, I just checked the scene, it looks like something is wrong indeed

Romullus, as you can check in the scene, no maps and no anisotropy rotation.
Maru, as you’ve seen, is a new created very simple scene. There’s actually something wrong in anisotropy shader.

Hi Alessandro,

this is a defect of the ‘Local Axis’ anisotropy orientation. On the top of the cube local Z axis gets aligned with the normal direction closer you get to the center which causes abrupt anistropy axis change. When you ‘disable smoothing groups’ there is no normal interpolation close to the edges and neither the abrupt aniso axis change.

Have you tried ‘UVW’ anisotropy orientation mode in your material?

Cheers,

Robin

Hi Robin, sorry for the late, I’ve got a little car crash (…) and I read your message just now.
Yes, with UVW orientation it work correctly, on uvw maps direction.
Thanks.