Page 1 of 1

Luxcore really needs an anisotropic rotation option

Posted: Fri Apr 16, 2021 10:23 am
by khaldrogo
REDACTED: This was a spam post: Post something reasonable sounding first (probably copied from elsewhere), then edit it into spam later...

Re: Luxcore really needs an anisotropic rotation option

Posted: Fri Apr 16, 2021 10:35 am
by Dade
khaldrogo wrote: Fri Apr 16, 2021 10:23 am I have not seen an X/Y workflow
Actually, LuxCore uses an U/V workflow: the anisotropic follows U/V coordinates so it can be rotated by rotating the UV coordinates. However, I'm not sure to understand your topic, can you add some link about this feature ? Like how it is supported in Vray, etc. ?

Re: Luxcore really needs an anisotropic rotation option

Posted: Sun May 11, 2025 6:42 pm
by nostalphia
Dade wrote: Fri Apr 16, 2021 10:35 am
khaldrogo wrote: Fri Apr 16, 2021 10:23 am I have not seen an X/Y workflow
Actually, LuxCore uses an U/V workflow: the anisotropic follows U/V coordinates so it can be rotated by rotating the UV coordinates. However, I'm not sure to understand your topic, can you add some link about this feature ? Like how it is supported in Vray, etc. ?
Greetings!
Apologies for my poor English,
I can understand that {u-rough=1&v-rough=0} can achieve a tight brushed texture sorted by left - right; and a circular or radial texture by rotating the uv,
ImageImage
(Adding subD (level 3) to the circular texture brings terrible saw aliasings)

but it becomes difficult when cannot changing the uv for some reason:
Image
in the U/V workflow, the u-rough = v-rough of Apoint and Bpoint(two points at ±45°) , who should be opposite value :o .

Re: Luxcore really needs an anisotropic rotation option

Posted: Mon May 12, 2025 8:35 am
by CodeHD
nostalphia wrote: Sun May 11, 2025 6:42 pm ...
(Adding subD (level 3) to the circular texture brings terrible saw aliasings)

but it becomes difficult when cannot changing the uv for some reason
in the U/V workflow, the u-rough = v-rough of Apoint and Bpoint(two points at ±45°) , who should be opposite value :o .
Hello,

Could you explain this in a little more detail, please?
For example, about the second issue, can you show a comparison of how it is supposed to look?
Ideally, always include a blend file (simple and with textures included) so that your setup can be better understood.

Re: Luxcore really needs an anisotropic rotation option

Posted: Tue May 13, 2025 10:03 am
by nostalphia
CodeHD wrote: Mon May 12, 2025 8:35 am
nostalphia wrote: Sun May 11, 2025 6:42 pm ...
(Adding subD (level 3) to the circular texture brings terrible saw aliasings)

but it becomes difficult when cannot changing the uv for some reason
in the U/V workflow, the u-rough = v-rough of Apoint and Bpoint(two points at ±45°) , who should be opposite value :o .
Hello,

Could you explain this in a little more detail, please?
For example, about the second issue, can you show a comparison of how it is supposed to look?
Ideally, always include a blend file (simple and with textures included) so that your setup can be better understood.
Hello Hello,
When I'm dealing with some oddly shaped objects, I can't get the desired effect by changing their UVs, here's an example:
Image
Its scratches are roughly like this:
Image
This is a mechanical watch movement deck, decorated with a regular (sort of) texture called Geneva stripes,which made by countless circular scratches arranged in parallel.
When using anisotropic rotation to implement a circular texture, this can be achieved by setting anisotropy ≠ 0 and linking a conical gradient to the anisotropic rotation:
Image
(Looking from the z-axis, sun shine's highlights are generated in the area perpendicular to the scratches)
Geneva stripes,which made by countless circular scratches arranged in parallel,can be achieved in a similar way:
Image

But if in u/v roughness workflow, I find it difficult to find some intuitive way to get a physically consistent result.

p.s, I haven't found a way to upload .blend files yet, maybe I could get your email address?
-------------------------------
to

Re: Luxcore really needs an anisotropic rotation option

Posted: Tue May 13, 2025 10:11 am
by CodeHD
nostalphia wrote: Tue May 13, 2025 10:03 am p.s, I haven't found a way to upload .blend files yet, maybe I could get your email address?
You can attach files directly to messages in this board, including .blend files.

The same with images, you don't need to use postimages.cc!

Re: Luxcore really needs an anisotropic rotation option

Posted: Wed May 14, 2025 2:16 am
by nostalphia
CodeHD wrote: Tue May 13, 2025 10:11 am
nostalphia wrote: Tue May 13, 2025 10:03 am p.s, I haven't found a way to upload .blend files yet, maybe I could get your email address?
You can attach files directly to messages in this board, including .blend files.

The same with images, you don't need to use postimages.cc!
:lol: thx
the file too large 2.2Mb (if include texture), I have to upload .blend to this board,texture to that .cc website :D
geneva waves.blend
(945.49 KiB) Downloaded 383 times
with the textureImage

Re: Luxcore really needs an anisotropic rotation option

Posted: Wed May 14, 2025 2:53 am
by nostalphia
CodeHD wrote: Mon May 12, 2025 8:35 am
nostalphia wrote: Sun May 11, 2025 6:42 pm ...
(Adding subD (level 3) to the circular texture brings terrible saw aliasings)

but it becomes difficult when cannot changing the uv for some reason
in the U/V workflow, the u-rough = v-rough of Apoint and Bpoint(two points at ±45°) , who should be opposite value :o .
Hello,

Could you explain this in a little more detail, please?
For example, about the second issue, can you show a comparison of how it is supposed to look?
Ideally, always include a blend file (simple and with textures included) so that your setup can be better understood.
there is another example to explain the convenience of{anisotropic rotation} in dealing with certain situations:
1.jpg
A metal ball with many very slight scratches. When the ball reflects the highlight of the light source, the scratches around the highlight (although their directions are randomly distributed) are also partially illuminated, but always appear to be rotating around the highlight.
scratches in blender viewport:
2.jpg
Anisotropic rotation calculated by node in blender viewport:
3.jpg
texture file:
Image
the .blend file:
Messy scratches.blend
(1.83 MiB) Downloaded 397 times
======================================================
using [Vector Brush] or [Flow Map Painting] by Mari or Substance Painter, to get the RG two-color map with scratch direction information, and then get the anisotropic rotation map through node calculation:
4.jpg
(in some renderer may add 0.25 at the end)