Video camera

While a graduate student at UC Berkeley, I discovered that one of the research groups had a high-speed video camera. I knew immediately that it had to be used to record water-rocket flights. The camera is a Kodak Ektapro model 1012E, built around 1994, and hence primitive in comparison to current models. Nevertheless, the camera was extremely capable and fun to use.

The camera can capture up to 239x192 pixels at 256 gray levels per pixel. In order to accomodate the high bandwidth of rapid video capture, all images are stored in digital memory. Only after the video capture is completed can images be transferred to a computer as consecutive TIFF files. The camera has about 53MB of memory (not much by today’s standards, but pretty expensive in 1994) so at the highest capture rates you get less than two seconds of capture at best. The maximum speed is 3000 frames per second (fps). At these high frame rates, very little light is available for each frame, so the camera contains a photomultiplier tube before the camera sensor to dramatically increase the sensitivity to light. Unfortunately, this also increases the noise. For these reasons, this camera cannot create the kind of quality still images that one can get with an ordinary camera and a high-speed strobe.

lights schematic


I wanted a way to light up the rockets during their upward flight while at the same time leaving the background dark. My solution was to use two overhead transparency projectors as light sources in the darkened room. I put strips of paper on the projectors to block the light so that each produced only a vertical slit of light which was directed at the flight path. The projectors were angled so that they would cast no light on the background of the camera’s field of vision.

Also, I wanted the water to be more visible to the camera. I simply added a small amount of milk to the water used in most of my rocket launches—the fluid’s ability to disperse light increases dramatically.


While I used a number of bottle sizes, most of the launches were of 2L bottles. Because of the tendency of the rockets to fracture, I used naked unmodified bottles until my supply ran out and I was forced to use a bottle with fins attached. I added a symmetrically formed chunk of modeling clay to the top of each bottle to provide more flight stability, to adjust the mass to my liking, and to absorb some of the impact energy. The clay wad deformed and often detached when the rocket impacted the foam pad mounted on the 3-meter-high ceiling. Clay spattering is not fun to clean up.