A Technical History of Computer Music
at the TIMARA Department
(Technology in Music and The Related Arts)
of Oberlin Conservatory
Electronic music study at Oberlin began when the field was still largely undeveloped. In 1968 John Clough and Olly Wilson, then members of the theory and composition faculty, received a $100,000 grant from the National Science Foundation (NSF) that allowed the Conservatory to purchase its first Moog synthesizer and Scully tape recorders to be housed in a studio in the basement of Bibbins.
Another chunk of the NSF grant went toward an IBM 360/44 computer, installed in the Wright Physics Building, for computer music processing. The IBM also served as Oberlin's principal computing resource until 1975. This system included digital-to-analog conversion (DAC) equipment and professional quality audio equipment for recording and monitoring output from the IBM. Oberlin was one of only a few such computer music production facilities in the country at the time. Most other schools had to send out their digital tapes to have them converted to audio at other facilities such as Princeton.
John Clough sets to work developing combined synthesis and composition algorithms for the system that results in TEMPO, a composer's programming language, and IRMA, an Interactive Real-TIME Music Assembler (1970).
The TIMARA (Technology in Music and Related Arts) program is established.
Sergio Franco is hired as a full-time engineer/lecturer for the TIMARA program. Sergio Franco had just come from the University of Illinois working with composer Salvatore Martirano and other engineers to design the Sal-Mar Construction, a massive analog/digital synthesizer with 250 suspended speakers built for live performance (Martirano 1971). Although the control interface for the Sal-Mar was digital, there was no overall computer control. While at Oberlin Sergio Franco took the next step and introduced computer control of analog synthesizer modules of his own design. By 1975 the project was completed with a PDP-8M computer controlling the new 'Hybrid Synthesizer'. (Hybrid.pdf - 2MB)top
In 1975 the College outgrew its computer resources and replaced the IBM 360 with a new Xerox SIGMA9 mainframe computer to be housed in new facilities at Mudd Learning Center. The TIMARA program also moved its computer music production facilities to Mudd Learning Center, on the fourth floor. The DAC system and its related control circuitry were replaced as part of the changeover. That equipment included:
Xerox 7910 analog output controller
Xerox 7969-7972 frequency control subsystem
4 Xerox DA50, 15-bit digital-to-analog converters
This was a quadraphonic system using a sampling rate of 10,240 samples per second per channel. Computational time for the conversions was about 30 times the actual music time, requiring overnight or weekend runs. The samples had to be stored on several large digital tape reels between the computation and conversion stages.
Communication with the SIGMA9 was accomplished via two serial lines, one for a DataMedia computer terminal and the other for a Diablo wide printer, plus a high speed data line for transmission of sound samples to the DAC system.
Throughout this time Gary Nelson is developing a computer language to be used for music composition, MPL (Music Programming Language), an extension of the general-purpose programming language, APL. It is introduced to students in the fall of 1976 resulting in a major jump in computer music productivity in the facility.top
Work on Sergio Franco's computer controlled Hybrid Synthesizer has moved to newly hired engineer Bob Hoover. The PDP8 computer is replaced with a more manageable Ohio Scientific OSI C2-8S Challenger II microcomputer based on the 6502 microcomputer chip. Bob Hoover and student Robert Moore develop the 'Hydrox' (Hybrid Doodle Rocks) 6502 assembly language program to control the Hybrid.
John Talbert (your author) is hired in the fall of 1978 as the TIMARA program engineer.
The OSI microcomputer takes on Analog Synthesizer Control in addition to its Hybrid Synthesizer duties. The following interface devices were designed and built into the OSI computer to allow computer control of the TIMARA program's Moog, Buchla, Arp 2600, and Putney VCS3 synthesizers:
8 DACs (8-bit digital to analog converters)
8 ADCs (8-bit analog to digital converters)
8 Pulse Detectors
8 Pulse Generators
An AY-3-8910 sound synthesizer chip
Users programmed the devices mainly through a tiny Basic language and saved their programs on cassette tapes. (OSI.pdf - 1.3MB)top
Bell Laboratories generously donates its prototype digital synthesizer, named the 'Alles Machine' after its primary designer, Hal Alles. The synthesizer is prominently displayed on the cover of Computer Music Journal (Vol.1, No.4). This was a completely digital sound synthesizer controlled by an LSI-11 Microcomputer.
The Alles Machine was given to the TIMARA program without maintenance or software support and thus required several years of work before it could be put to use in the studio. Gary Nelson took on the job of software development and I took on the work of hardware development. We immediately added a Heath disk drive and a Pascal software package. My job was to glean the operation of each board and document my findings. In addition, I redesigned part of the controller circuitry for the working console (the keyboard, sliders, lights, and switches), added a video monitor and video graphics circuit, designed and built Timer circuits and hardware FIFO stacks used in feeding the synthesizer its time coded parameters.
The Alles Machine gave Oberlin a unique opportunity to be at the center of research around the latest generation of microcomputers and digital synthesizers. It provided a vehicle for research into the particular problems associated with real time music synthesis. In 1983 Gary Nelson and I presented a report on our work with the Alles Machine at the Fall Computer Music Convention in Rochester, New York. (Alles.pdf - 0.3Mb)top
MIDI (Musical Instrument Digital Interface) is introduced to the world. By January of '84 we have purchased our first MIDI keyboard in the Yamaha DX9. By the fall we also have Yamaha DX7 keyboard followed by the 8-voice TX816 (1985). These synths come close to duplicating all the capabilities we have been working on with the Alles Machine.
For the next several years I work making MIDI interfaces for everything in sight: our OSI micros, IEEE interfaces on the Osborne microcomputers (1985), Radio Shack Color computer, the Alles Machine, and later, the Macintosh Plus (1986). (Midi.pdf 1.2Mb)top
The prototype for the MIDI Horn is designed and built. This is a wind controller with a pressure sensor, 8 switches, and several sliders/pedals. The controller data is fed to a single board Z8 microcomputer that interprets the data and converts it to MIDI control signals. Forth language was used for programming the device as Basic language proved to be too slow.
The college retires the Xerox computer and along with it, our venerable Xerox Digital to Analog Converter connection. The TIMARA program buys into a Ridge 32C Minicomputer complete with maintenance contract. This was a Unix based system with the following peripherals: 445Mb Winchester drive, 60Mb Priam drive, streaming tape drive, 8" floppy disk drive, 4 RS232 serial ports, Hi Resolution Display and optical mouse. A parallel interface DR11 board was added for connecting to our Xerox DACs. I worked on building the interface between the two, programming drivers in C. Also added were two Analogic MP2735 ADCs and a MIDI interface.
The Ridge Computer allowed us to continue our computer music capabilities for several more years until the Macintosh computers became powerful enough to take over extensive MIDI sequencing and CSOUND synthesis. (Ridge.pdf - 2.8MB)top
Since '85 Gary Nelson has been experimenting with Apple's Lisa computer. Late '86 the TIMARA program buys its first Macintosh Plus computer. The Macintosh computers are to become the program's main "computer of choice".
The final version of the Midi Horn is built. This is a MIDI controller instrument based on a single board microcomputer and programmed in Forth programming language. Gary Nelson takes the MIDI Horn on the road with over 200 performances around the world. He uses the Midi Horn as the performance interface part of a 'hyperinstrument' consisting of a Macintosh computer, a set of digital synthesizers, and the software (Max/MSP) linking them all together. In a 'hyperinstrument' the controller does not necessarily play 'notes', it sends performance signals acted upon by a computer program composed to control how the music is played out.
After working with Forth Programming Language on the MIDI Horn I am impressed with its speed, compactness, and ease of use. The Forth language consists of a dictionary of words (subroutines) and several stacks for storing the subroutine data. Programming in Forth is a matter of building new 'words' by combining previously built words that are already in the dictionary, thus creating a hierarchy of words. The higher-level words can easily be tested by running their lower level components. Lower level words that deal directly with the processor hardware are easily built, even using assembly code if that is deemed necessary for speed.
The Ohio Scientific Microcomputer is upgraded in 1987 with a Forth based system (RSC Forth). The language is significantly extended with words that deal with the Hybrid Synthesizer, MIDI input and output, a Timer device, a new SID synthesizer chip, and all the devices used to control the analog synthesizers such as control voltage DACs and ADCs, pulse detectors and generators. The Hybrid Synthesizer interface is rebuilt with new waveform generators and timer control. Floppy drives are installed for user storage of programs. (RSC.pdf - 1.2Mb, RSC Code.pdf - 1.2Mb, Hybrid.pdf - 2Mb)top
Analog synthesizer circuitry reaches a certain maturity with the availability of chips such as Solid State Music chips and the Curtis music chips. I use these to design and build an octal Voltage Controlled Amplifier (VCA), a quad Voltage Controlled Filter (VCF), an Aural Exciter with all its components available, and an Analog Delay Line box. All of these are controllable with the Ohio Scientific Micro using the extended Forth utility words.
In the fall of 1989 a new TIMARA facility is opened with dedication ceremonies. The acoustically designed facility includes two quad studios, a recording studio with control room and recording space, two faculty offices, a computer workstation lab, tech room and lobby. Work is begun moving the computer music studio from the fourth floor of Mudd Library to its new home, filling the studios with equipment, and pulling audio and computer lines between them.
The Macintosh computer starts to take a more prominent role in our studios as we replace the Mac Plus and SE models with the Macintosh IIcx loaded with a Digidesign Audiomedia board.
TIMARA receives the gift of its first PC computer, a Gateway 2000 PC with Windows 3.1 from alumni Greg Hendershott, founder-owner of Twelve Tone Systems, creator of Cakewalk.
Work is started on designing and building control voltage to MIDI devices. One project uses an 8088 microprocessor with a ROM based Forth system controlling 8 bit ADCs. Another project is to reprogram a Roland PG1000 slider box to put out any type of MIDI signal.
TIMARA plugs into the College's internet after AppleNet cables are pulled between the studios.top
TIMARA joins the World Wide Web with one of the first web sites on campus.
A new Computer Music Workstation Lab is designed and built. It will eventually include nine Macintosh computer workstations, each with a Yamaha Keyboard Synthesizer, a MIDI interface, a Sampling Synthesizer, a Mixer, and a full complement of Music Software: Notation (Finale), Sound Editing (ProTools), Midi Sequencing (Vision), Music Programming (Max/MSP). All the stations are connected to racks containing several different recording equipment formats: Turntable, Cassette, DAT, Reel to Reel.
I take on the job of Computer Lab Manager, using the management program RevrDist from Purdue University to keep the lab and studio computers updated and regularly cleaned.top
TIMARA introduces video editing into the curriculum. A new studio is outfitted with video decks and monitors and a MEDIA 100 Video Editing system installed on a Macintosh 8600 computer. Video cameras are made available for check out.
TIMARA keeps updated with the latest in computer technology. We have more than 20 Apple G4 and G5 computers and several PC systems. This includes a Web Server (using 4D Webstar), office and faculty computers, a Video Studio with surround sound, a Digital Recording Studio (with Digidesign HD system), 3 Macintosh computers on Carts, 2 PC computers on carts with Windows XP Pro and Mandrake Linux, and a public workstation lab with 9 G4 Macintosh computers.
The computers have the latest system OSX installed and are managed with Radmind from the University of Michigan and Carbon Copy Cloner by Michael Bombich. (Radmind Lab Setup Instructions.pdf 124KB)top