For a graphics class at the University of North Carolina I developed an algorithm to approximate global illumination. These were the first two tests:

Rob Wheeler and I designed an architecture for a low-end graphics chip. A simulator was used to generate this teapot:

Here’s some advice on premultiplication of colors with alpha.

I spent a lot of time at PDI working on issues of gamma correction in graphics.

I had some insights into how homogeneous coordinates work in graphics.

In 1989, I wrote my first computer graphics program with a high-school friend, Fredrik Fatemi, on his 286 PC with an EGA card. We filmed the frames one at a time using my Super-8 camera. The first two clips are on his computer and the second two are movies we made on an SGI Iris in college a year later.

Six years later, he and I worked on the dynamics of a bowling strike. It was rendered on the simulator for the PixelFlow graphics machine I was working on at UNC.