In article <4000b0b1$1@news.povray.org>,
 "Dan P" <dan### [at] yahoocom> wrote:

> What an interesting idea! It might take an eon to render, but it would
> definitely make amazing hair.

Well, taking a single media sample ought to be faster than testing for 
intersection with a complex hair object, and you should be able to get 
good results with fewer media samples than you would have hair 
intersection tests. Not fast, but almost certainly faster...hopefully.


> Using the exported guide-hairs, you could use
> an isosurface and fill-in between them using some custom function (perhaps
> build it into MegaPov initially?). It would be pretty hard-work unless
> you're a math wiz. If you are, man, you'll be a hero!

The media-hair would give the hair more body, the true-hair would be a 
visual cue to tell the viewer's brain that it really is hair. (A mass of 
something that looks like it might be hair should look much more like 
hair if you can see a few actual hairs.)

And what I actually had in mind was using the function to create some 
true-geometry hairs to keep the overall appearance from looking fuzzy or 
blurry, but that would require modeling the hair direction with a 
density field, which most people probably aren't going to want to do. 
Hmm...one way would be to find the closest point on the nearest guide 
hair, and use the direction of the hair at that point, maybe with some 
noise mixed in. Sounds slow, though. Maybe do something like a blob 
pattern with the guard hairs, compute a weighted average of hair 
directions...should be much faster, and give smooth transitions between 
guide hairs.

Simple angle-dependence isn't enough for it though...it would need to 
react to light sources as well, and hard coding them into the density 
would be impractical. Basically, the media density needs to be a 
complete texture.

-- 
Christopher James Huff <cja### [at] earthlinknet>
http://home.earthlink.net/~cjameshuff/
POV-Ray TAG: <chr### [at] tagpovrayorg>
http://tag.povray.org/

Post a reply to this message