My graphical stuff

I have uploaded some of the pictures generated by my path tracer which I wrote for Image Synthesis (CS6620 Spring 1999, U of U). It is similar to a ray tracer, but it statistically samples the entire hemisphere. Thus, the program is MUCH slower, but can have more realistic looking features without the need for cool 'hacks'.

Just for personal interest, here is a copy of a web page that I started writing about 7 years ago, that I intended to be An introduction to raytracing. It is not complete (probably never will be), but it may be useful to someone somewhere.

If you have any interest in my work from CS6620, you can always checkout my CS6620 pages, which were originally hosted on Shadow (a machine on which I spent most of my time for 3+ years).

Here's an image to demonstrate image maps, spheres, lights, reflective surfaces, etc.

This one shows mirrors (near perfect mirrors), and wide-angle lenses in a physically impossible situation: the camera (invisible) inside of a mirrored box with the top mirror being a light source. The distortion of the spheres near the edges is caused by having a wide-angle lens.

Finally, a couple of images to demonstrate solid textures.



Here are two images of the same scene. One was rendered with my pathtracer, and the other was rendered with the GNU project Panorama (including my modifications).
The important part to notice is the bottom of the spheres. The path tracer includes indirect light, whereas the raytracer only does direct light calculations.
Both images were rendered to a reflection depth of 5, and with approx 256 samples per pixel.
To make the raytraced image more visible, I doubled the intensity of the area light over that of the path traced image.
Pathtracer generated image
Pathtracer generated
Panorama (raytracer) generated image
Panorama (raytracer) generated

Around June of 2001, I started working on my old project again, the main goals being ease of modification (not met) and incredible flexibility (it has a lot). I rewrote almost everything (except the pathtracer portion), and wrote a new raytracer for it. The new version has the ability to have better textures, faster rendering (due to bounding heirarchies), and more shapes (disc, polygon, triangular mesh [obj file import], transformed objects [correctly working], etc.).
Here's an image showing quite a few of those shapes:
Currently possible shapes

Here's another image showing some more textures:
Texture/shape demo
I'll have to check to see if that glass is correct... It now looks a little weird to me.

Again, as my interests shifted, I stopped (temporarily) working on my own project to continue on Panorama. I have written a few thousand lines of code which are not in the main distribution (yet). They improve stability (for what works -- not everything does, yet), increase flexibility, and ease the burden of making changes to some major items (rt scene loader in particular -- although it lost it's backwards compatibility). When Angel adds me as an administrator (I'm the only person working on it right now), I'll commit some more of those changes. Well... It's something to look forward to.

When I get the time, I'll put more images up here, complete the program that I used to generate them (my own program, that is), and then release it to the public (source code and all).


Go back to Kevin's Page.

Revision History:
07Dec1999 Moved the graphical related stuff here from my main page.
29Jan2001 Added the scenes showing pathtracer vs. raytracer.
03Jan2004 Added the link to the CS6620 pages from shadow.

Here is a count of people who have seen this page since 25Jan2001:
Counter

1