TITLE: Solitaire Confinement
NAME: Gordon Griesel
COUNTRY: USA
EMAIL: gordon@mpgenius.com
WEBPAGE: www.mpgenius.com

TOPIC: Loneliness
COPYRIGHT: I SUBMIT TO THE STANDARD RAYTRACING COMPETITION COPYRIGHT.
JPGFILE: soliconf.jpg
RENDERER USED: 

        My own ray tracer written in C, which I developed after reading
        the book Photorealism and Ray Tracing in C, Watkins, Coy, Finlay, 1992.


TOOLS USED: 

        TextPad editor
        Paintshop Pro
        pencil and paper
        an old deck of cards


RENDER TIME: 
    2 39 00


HARDWARE USED: 
    Pentium-III 850 laptop with 544MB


IMAGE DESCRIPTION: 

        The king is being held in solitary (or solitaire ha ha) confinement.


DESCRIPTION OF HOW THIS IMAGE WAS CREATED: 


        I developed this scene to showcase my new radiosity calculator.

        There are only two light sources in the scene, both outside the cell
window.

        All other lit areas are accomplished through radiosity.

        I used a system of radiosity patches which are adaptively sub-divided
into smaller
        sub-patches where needed to make the lit surface appear smoother.

        One can easily see the patches in my scene.  A smoother surface will
require more
        rendering time which I did not have for this contest entry.  I can
produce a
        really nice version of the Cornell Box rather quickly.

        The king and queen texture maps are also covered with radiosity
patches.

        Please notice how in areas of higher light gradient, the patches are
smaller because
        they have been adaptively sub-divided as the scene was rendered.

        Three passes were made, each pass distributing energy from the 6400
brightest patches.

        Hopefully, my next contest entry will be more detailed and have smoother
lighting,
        as I will have some time to refine further my lighting techniques.

        Here are some of the optimization methods used to speed up the radiosity
calculations.
         . vertex meshes used across each polygon and shared by patches and
sub-patches.
         . backface culling of patches.
         . hidden surface culling of patches.
         . frustrum checks for patch division.
         . layered approach to energy gathering for patch verticies.

        I got most of my ideas for the radiosity C code from the following
link.

        http://www.cfxweb.net/~aggrav8d/tutorials/radiosity.html

        Thanks for viewing my image.
        :)

