[Radiance-general] Anisotropic Roughness Vector Alignment

[Gregory J. Ward]

Hi Marcus,

A radial anisotropy is one of the easiest types to generate.  The  
values for the vector [ux uy uz] are variables for a reason.  These  
can be set to [(Px-xcent) (Py-ycent) 0] for a clock face in the x-y  
plane, which may be rotated and moved wherever you like afterwards.   
To make things even simpler, we can assume [xcent ycent] == [0 0] and  
use:

void metal2 brushed_around
4 Px Py 0 .
0
6 .5 .5 .5 .9 .03 .12

brushed_around ring shiny_face
0
0
8
	0	0	0
	0	0	1
	0	2

Using xform, you can then move the surface into position.

-Greg

> From: "Marcus Jacobs" <marcdevon at hotmail.com>
> Date: November 3, 2006 4:47:19 PM PST
>
> Dear Group,
>
> I have a question concerning anisotropic roughness vector  
> alignment. Specifically I would like to know is there anyway to  
> correlate the u and v mapping values to a desired roughness vector.  
> Here is an example what I am attempting to simulate (see photo below)
>
> http://i35.photobucket.com/albums/d178/marcdevon/100_0050.jpg
>
> You can see how the brush strokes follow the base of the clock  
> circumferentially. Ideally, if one were to model the same object,  
> one roughness vector value would align itself around the clock  
> base. The other value would be in the radial direction. For an  
> object such as this, I can model the object such that the image map  
> will align around itself in a similar manner (see image below):
>
> http://i35.photobucket.com/albums/d178/marcdevon/ring01_render.jpg
>
> Since I can align a texture in the direction that I want, I was  
> wondering if there is I can translate the uv coordinates in a  
> manner that could be used for specifying the anisotropic roughness  
> values.
>
> Please let me know if I do not make sense. Sometimes I have been  
> known to confuse myself :>)
>
> Thanks,
>
> Marcus