The name Raymond Kurzweil is likely to be familiar to most readers. One of his many inventions, the Kurzweil 250, was the first 88-key polyphonic digital synthesizer on which chords could be played and that was capable of realistic reproduction of the sound of a grand piano and other acoustic instruments based on digital sampling and recording of real sounds. Stevie Wonder knew of Kurzweil because of the latter's earlier invention of a reading aid for the blind and interest in building a synthesizer, Wonder asked Kurzweil if he could create an electronic musical instrument specifically adapted for him. Because Wonder represented the kind of professional musician Kurzweil wanted to cultivate as a user and advocate for his instrument, he actually created a Braille prototype of the synthesizer.
With the expected functional complexity and a target selling price of $1,000, the design of the new machine presented a significant and very interesting challenge. For BBN Technologies, the opportunity to confront that challenge came in the summer of 1982. It all started when I received a phone call from Aaron Kleiner, a principal in the New Yorkbased startup company Kurzweil Music Systems, who asked if I would undertake the human factors design and packaging of "a revolutionary keyboard-based electronic music synthesizer." The only thing I was not to be responsible for was the piano-like keyboard because Kurzweil had a "friend" who had invented a way to make a keyboard that simulated the feel and percussive dynamics of a real piano.
There was no formal request for proposal. My colleague Carl Feehrer and I were shown a glossy prospectus that pictured (opposite page) and touted the features of this unusual machine for the benefit of potential investors. There appeared to be no one else invited to compete for this HCI opportunity. They wanted a proposal now and a completed, detailed specification within six months. At BBN we called such an inquiry a "bluebird"an unsolicited opportunity that just "flew in the window." Since neither Carl nor I nor anyone at BBN had real experience with the packaging aspects of industrial design, we immediately located an industrial design collaborator, Paul Brefka of Latham, Brefka Associates in Boston. Both Carl and I could claim human factors expertise, but Carl was trained in music and I had no real music experience at all, except that I survived listening to my son "percuss" daily on his drum set in the basement. We acquired a Casio VL-1, almost a toy, for under $30 just to find out what a keyboard synthesizer was all about.
Together with Brefka, we submitted brief proposals, and after a financial negotiation during which BBN refused to accept stock in the fledgling company with an uncertain future in lieu of cash payment for our work, we embarked on a brief but fascinating project.
We immediately broke up into two teams. Brefka worked with the mechanical and electrical engineers on packaging constraints and specifications. The human factors team included Pew, Feehrer, a rock musician, and two or three Kurzweil employees"software gurus," one of whom was accomplished both as an electronic and acoustical engineer and as a musician. We were given a very detailed 300-page technical specification for the instrument's functional characteristics and circuit design that we were told had been written single-handedly by Ray Kurzweil. We visited local music studios and synthesizer repair shops to acquire the requisite contextual data about the use, programming, and repair of such instruments. We had two kinds of meetings at Kurzweil headquartersweekly design meetings of the two teams independently and, less frequently, coordination meetings of the combined staff.
Ray Kurzweil himself attended most of these coordination meetings. He was a friendly, soft-spoken young man in his mid-30s. He exuded confidence and frequently redirected some of our activities.
The design problem for this intricate bundle of electronics hardware and software was twofold. First, musicians are rarely, and usually don't want to be, electronics engineers. They want to play. The control panel had to look slick and be perceived as simple to operate at performance time. We needed to bury the inherent complexity. Second, there were actually three levels of complexity from the point of view of the musician-user. Besides the performance time instrument, there were the pre-performance setup requirements. Contemporary professional musicians use sequencesrepetitive series of notes in selected "voices" that are pre-programmed and can be selected quickly and introduced into the performance as repetitive background accompaniment. Like organists, the keyboard players want to be able to call up different sound effects while playing. Some of these to-be-called features were made directly available on the panel, but others had to be assigned to buttons ahead of time. Besides pre-performance setup, the more sophisticated user was also given the capability to record new sounds or to modify or adapt a factory-preset library of sounds.
Oh, I forgot to mention an additional design problem. Because of cost considerationsremember, this was 1982we were told, in spite of protests, that we were limited to a one-line, 16-character LED display with which to communicate all of these interactive control activities.
At performance time, the user was given direct access to controls needed in the course of a "gig" through sliders, wheels, buttons, and foot pedals. A digital number pad was used to call up prearranged sequences and keyboard assignments rapidly. For example, the keyboard could be "split" and have different instrument sounds assigned to different blocks of keys. Some numbers were assigned to factory presets and many more to user-defined presets.
For setting up keyboards, the time and effort constraints associated with live performance could be relaxed. We utilized the display to provide specific prompting of what to do next at each step. Preparing pre-recorded sequences was handled similarly.
The more difficult problem was providing access for purposes of creating new or modified versions of the stored keyboard layouts, i.e., the detailed features of stored sequences of notes and the parameters that make up each instrument sound specification. This was accomplished by thinking of the keyboard layout in terms of a spatial, two-dimensional matrix with different keyboard layouts arrayed vertically while horizontally, the detailed parameter specifications of key assignments were provided. The 16-character display could present only one cell of this matrix at a time, plus enough information to identify the context. The user employed the left-right and up-down arrow keys to move freely among different layouts and among specifications. While in principle these arrays were very complex, the use of a spatial layout as a mental model and maintenance of consistency in the order of parameter specifications allowed a simple, predicable way of providing access.
We completed our design specification on time and almost within budget and delivered it to Kurzweil in March 1983, fully expecting to hear back from them so that we could answer questions and/or iterate our design. However, as is so often the case with consulting contracts, especially those with cash-hungry startup companies, we never heard from them again.
The commercial version of the Kurzweil 250 was officially announced at the 1984 summer show of NAMM, the International Music Products Association, with a price of $10,715 plus options just a bit above Kurzweil's original price point. It was manufactured commercially until 1990 with several follow-on versions. Of course, comparable synthesizers today are available even below his original thousand-dollar target price. The version introduced appeared, at least superficially, to conform largely to our original specifications, but it had capabilities well beyond those we provided for, including midi- and Apple Macintosh connections. At some point, the small display was replaced with a two-line, 48-character display.
In 1990 Kurzweil Music Systems was sold to Young Chang, a large Korean musical instrument company. It remains active as a division of Young Chang today as one of the market leaders in computer-based musical instruments, marketed in more than 40 countries.
Dick Pew spent 11 years at the University of Michigan and the past 33 years at BBN Technologies in Cambridge, MA, where his current employment status is "part-time irregular." His interests have spanned a range of human factors activities, from HCI to human performance modeling. At this stage of his career, history becomes an attractive topic.
©2008 ACM 1072-5220/08/1100 $5.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2008 ACM, Inc.