Final Report to the National Endowment of the Arts

Excerpt of the Final Project Report concerning control structures for digital computers used in art-making.

General Statement of the Problem

Analog computers have been much more successful as tools for the creative artist than digital computers. Analog computers are quite successfully used in music and more recently, in video art. Most of the current work in digital computer art suffers from a real lack of fluidity in control, which is why the work appears so sterile and mechanical. What this project has attempted is the creation of dynamic, flexible, user-programmable input devices and requisite control structures. This report will attempt to communicate the actual progress we have made in humanizing digital computing for aesthetic ends.

The problem in general arises from the reliance on linguistic/numeric control of digital computers. Computers were invented primarily to do accounting and highly numerical scientific tasks. If they had been invented with the purpose of doing art, they would probably be quite different in structure. Analog machines, on the other hand, rely mostly on sensing and processing continuous events and data collected via well-tuned input devices. This orientation makes analog computers much like musical instruments, video cameras and other artistic tools that transform continuous phenomena into continuous phenomena.

Language in general and computer languages in particular do not easily perform this type of function. The solution, then, is to provide rich non-linguistic inputs to the digital system and simulate the workings of an analog system but retain the advantages of working with computers.

We had done quite a bit of preliminary investigation before applying for the NEA grant. We have been using slide potentiometers and dials as well as joysticks and data tablets for roughly eight years. Nevertheless, dials, slide potentiometers and the like are very inefficient in their connections to humans. One can only control two of these devices at once, getting a maximum of four dimensions of coherent control. The situation is worse with computer keyboards and function button arrays. Yet the amount of independent degrees of freedom the human body can exert is quite fantastic, perhaps as much as one hundred, but the problem of wiring oneself up to a computer has yet to receive enough attention, to say the least.

So the goals of this project were to provide convenient multi-dimensional controls laden with constructive feedback. Other self-imposed constraints were adopted as well: the controls had to be easy to construct by individuals without access to an elaborate machine shop, the documentation had to be usable by someone with only moderate electronic skills and the construction materials had to be affordable by people without the resources of a major grant.

What we developed falls into four categories:

• Pressure sensitive keyboards
• Sensing clothing
• Force feedback dials
• Time-based computer variables