> Which leads to another question: Do surface hilights really exist? Or are 
> they just reflections of the shape of the light source?

They do exist, and they are not just reflections of the shape of the light 
source.  They are due to the geometric structure of the surface on a very 
small scale.  Think about each surface IRL actually having something like a 
normal{bumps scale 0.000001} block and then rendered with perfect 
anti-aliasing... even a point light source will give you a non-zero sized 
highlight.

Diffuse, specular and reflection are all the same thing IRL, it just depends 
on the geometry of the surface.  IIRC some light reflection model used in 3D 
cards (or maybe POV too) is based on theoretical analysis of random surface 
variations.  Something like assuming the surface to be made up of a 
triangle-wave with each peak a random height.  You can calculate the % of 
light reflected in certain directions mathematically and with a few 
simplifications come up with a model for diffuse and specular components.

Post a reply to this message