Margus Ramst <mar### [at] peakeduee> writes:

> Thomas Willhalm wrote:
> /.../
> > 
> > You can however add some turbulence to the function. This can result
> > in hairs that point in different direction at different regions of the
> > workpiece. Of course, you could also use a non-random function to
> > change the direction.
> 
> Yes, that's basically what I meant - hairs not pointing straight up from
> the surface normal. Also, how controllable is the shape of these hairs -
> can they be bent etc? I did a quick search, but I couldn't find the paper
> you mentioned. Could you provide a link to it?

As far as I know, the paper isn't available for free on the net. (The ACM
wants to earn some money.) You have to look for it in a library.

They model the fur like this:
You must provide a function R^3 -> R that is 1 at the surface of the object
and 0 at the top of the hairs. Then, the density function is roughly given
like this:
1. project the point to the surface
2. decide whether it is part of a hair by a noise function (The noise 
   function that is built in POV-Ray)
The values are adjusted by some functions to give a good ratio for empty
space and hairs. To incorporate reflections, they use the gradient to
simulate a normal.
You can modify the location of the point by a 3D noise function before
projecting it. That gives you some curly hairs.

I guess that it is also possible to include other modifications. It is
probably possible to simulate gravity and the like.

Thomas


-- 
http://www.fmi.uni-konstanz.de/~willhalm

Post a reply to this message