A Monthly Column
Computers And Society

David D. Thorn burg Associate Editor

ET. & T.

I just finished seeing Spielberg's film E.T. for the second time. I can't imagine anyone who didn't feel a deep sense of sympathy for the poor creature left on earth by accident. His valiant attempts to contact his ship made this one of the most touching and heartwarming films ever made.

It is interesting, though, to consider just how unsophisticated our own computer-based communications services really are. Those of us who use various information networks and bulletin boards have to carry a collection of phone numbers and log-on procedures. Furthermore, once we are connected, we need to remember the correct command syntax to access services on each of the systems.

While I find that I derive a great deal of value from these services, I can't pretend to say that I am happy with the primitive state of their development. For example, suppose you belong to information utility A and I belong to utility B. Each service has its own electronic mail system, but neither allows us to communicate across the utility boundaries. This makes as much sense as saying that people with phone service provided by G.T. & E. can't contact people on the Bell system.

There are many solutions to this problem. The creation of one massive utility is one solution I don't care for. There is no technical reason that messages destined for one network can't be automatically forwarded to the other networks as needed. Since both sender and receiver pay for this service, the various utilities don't even have to keep track of the balance of message traffic. If such a system becomes commonplace (and it must if electronic mail is to become a commercial reality), then we would be free to subscribe to one utility on the basis of the services it provides and know that our decision would not preclude our communication with someone who made a different selection.

I am a strong proponent of computer-based message systems, but until the various utilities let people communicate across their boundaries, it will be hard to think of these utilities as more than hobbyist ventures.

More Movie Magic

There are a few reasons I haven't reviewed movies in this column. First, most films are long gone by the time this column gets into print and, second, I hate standing in lines.

Nonetheless, I find that I have been going to movies more often – if for no other reason than to see the commercial use of computer animation. Conventional animation techniques (such as celanimation using hand-painted celluloid sheets) have long been used to create the illusion of computer graphics, but until recently, the actual use of computer generated imagery has been limited to very short segments. Note that I am restricting my comments to big box-office films. Numerous films using computer-generated images have been around for years. Anyone who has not seen John Whitney's Arabesque, for example, should watch for it. It is sometimes shown on PBS and is very much worth seeing.

Prior to this year, one of the most successful films to use computer-generated imagery was Star Wars, specifically the "Death Star sequence," in which the attackers are briefed on their strategy for the destruction of Death Star. This monochrome sequence was made for Lucasfilm by artist/computerist Larry Cuba. From this beginning Lucasfilm has built up a computer graphics group that includes talented people like Loren Carpenter Loren has made extraordinary practical use of fractal geometry in the generation of landscapes and planet land masses. His work is so realistic that, given the pace of an action film, it is virtually impossible to tell that the image has been synthesized.

While the new wave of young film makers is perhaps more inclined to make use of new technology, the old established studios clearly do not want to be left behind. In Walt Disney's TRON, for example, a significant portion of the film (about one half hour) was made from computer-generated images. I saw this film along with a group of teenagers. After the film one of the kids said "Boy! You sure enjoyed yourself. How come? After all, it's a pretty dumb story."

Well, the "how come" part is pretty easy. I wasn't paying much attention to the plot (that being all too easy to follow). I was paying a lot of attention to some of the most spectacular computer graphics I have ever seen. Two graphics groups were involved in making the bulk of the images: Information International, Inc., and MAGI. Each seemed to be trying to show the other how good they were. The result is that the viewer is the real winner. Whether the effect was the partial transparency of a craft's wing, the subtle shading and reflection from a spherical surface, or any of the other myriad effects in the film, the use of computer animation in TRON created a film that would have been totally different had it been made any other way.

This brings me back to an old topic. The computer animation in TRON is a new art form distinct from any other. The goal was not to have the computers replicate scenes that would be the natural result of hand sketches or photographs, but rather to let the computer generate images of a highly fantasized model of its own inner workings. This aspect of the computer graphics was so important that those scenes using conventional sets were painted to resemble the computer-generated imagery as closely as possible.

This doesn't mean that computers can't generate realistic images. They can. It remains to be seen in which direction the continued creativity of film makers will push this new medium.

For a glimpse into the world of high resolution computer graphics, you might want to read The Computer Image (by Donald Greenberg, Aaron Marcus, Allan Schmidt, and Vernon Gorter, Addison-Wesley Publishing Co., 1982), a new book that not only presents a beautiful sampler of computer-generated images, but also provides a good description of the techniques used in computer graphics. Although outside the capabilities of the computer systems found in most homes, the illustrations point the way toward the day when resolutions on the order of 2048 by 2048 pixels will be affordable to the average computer artist. Until that time, books like The Computer Image and films like TRON will provide clearly marked goals towards which we home computerists can reach.