Steven Pigeon wrote:
> 
> > Remco de Korte wrote:
> > >


> > > I would like to build it from small parts like triangles and I
> know

> thinking
> > > about a number of points randomly scattered over the surface and
> then
> > > make triangles with those.
> 
> There are numerous ways to do "spheres" with lots of polygons.The best
> I can
> see is to use a "Buckminster-Fuller"-type polyedron.
> You start with an icosaedron (20 faces... you know, d&d 20 sided
> dice) then progressively divide the faces into 4 other triangles. You
> keep dividing triangles until you've got a satisfying result.
> 
> IMHO, there's no real use to that. A sphere is much better repre-
> sented when it's a z^2+y^2+X^2=r^2 form; especially if your
> only goal is to render a simple sphere.
> 
> On the other hand, the buckminsterized sphere provide almost
> uniform sampling points on the sphere; you dont have the point
> concentration effect you have with polar/spheric coordinates. It
> is also ideal for spheric fractals (or "rocks", either John's
> generator,
> http://www.erols.com/vansickl/rock.htm, or my own generator,
> http://www.iro.umontreal.ca/~pigeon/pub/PovPage/RockGen/RockGen.html
> (with source included)).
> 
> I first buckminsterized spheres to do uniform sampling, in the
> context of environment mapping (to reduce rendering time).
> Then for to fractal planets.
> 
> Best,
> 
>     S.
>  ----------------------------------------------------------
> Steven Pigeon                     Ph. D. Student.
> University of Montreal.
> pig### [at] iroumontrealca           Topics: data compression,
> pig### [at] jspumontrealca          signal processing,
> ste### [at] researchattcom           non stationnary signals
>                                   and wavelets.
> ----------------------------------------------------------
>          http://www.iro.umontreal.ca/~pigeon

I downloaded the program Ken mentioned (Geodome1). There's a really nice
image by the same name at the FTP-site.
I had a look at the program and though it does do what it's supposed to
do it wasn't exactly what I needed (my question was a bit vague). I had
a look at the source though and that brought me to the idea to do
something like you describe here: a recursive function, creating a
sphere with triangles, subdivided by triangles, etc. With this I could
do what I actually intended to: to blow up a sphere into fragments.

You're right that it wouldn't be very smart to use this method if you'd
just need a sphere. Even with smooth-triangles I found that there were
seams all over the object. But if you'd need any 'special effects' (
abump in a sphere) you could consider this (or use a blob).

Regards,

Remco de Korte
http://www.xs4all.nl/~remcodek

Post a reply to this message