Hardware Development for Computer-Based Instruction
In discussing the development of computers with the layman, terminology must be defined as it applies to the relatively new field of computer-based music instruction (CBMI). Hardware, software, courseware, and program have been used freely by those involved in the field of computer instruction, without clear definition for the outsider, the novice, or the interested pedagogue.
In dividing computer terminology into two domains for discussion, hardware includes the physical computer and connected components, while software includes operational intangibles which reside inside the computer, such as programs, operating systems, courseware, computer lessons, and computer data. Computer software can be seen in printed form on computer cards or printouts, but it also consists of electronic impulses which are invisible and intangible while being used.
This article is a survey of the tangible domain of computer developments as it applies to the use of computers as an educational tool in music. The development of hardware is an encompassing term used in computer jargon to define the machines (computers), physical attachments to computers (peripheral devices), and the interconnecting equipment (communications devices). Hardware is common to all computers by definition, but the type, cost and amount of equipment used with a particular computer can vary greatly.
Hardware can be classified by size, starting with the size of the computer. Minicomputers were once "small" computers, but only until the development of microcomputers. Since microcomputers have as their "brain" a small processing unit called a micro-processor, the two terms are used interchangeably. The home computers advertised in the media today are such small computers, i.e. microcomputers. The trend is toward miniaturization of computing hardware without loss of computing power. An example of an earlier technological miniaturization is the hand calculator. A similar reduction in size has taken place with computers since the mid-1970s.
In discussing computer size computing power can no longer be measured in size of physical presence. The large computing center with rows of impressive cabinets we saw in the 1960s could be replaced today with much smaller equipment possessing an increase in computational power. Computers continue to be discussed in terms of size, however, with three main classifications: the maxicomputer, the minicomputer and the microcomputer. The large computer is capable of multiple-user interaction. Typically this class of computer, such as the IBM 360 or 370, the Control Data Corporation CYBER and the CREY, is capable of computing information for as many as 1000 users simultaneously. The minicomputer also has the capability of being used by multiple users, but the computing power will generally allow access by only 20 to 30 users at one time.
The new era—that of the microcomputer—is one of the personal computer. Since the size is small, the computing power is limited to one computer programing task at a time. However, this has cut the cost of computers to an affordable level for home and school use. With this perspective, one of an industry that has worked from a large multiple-user environment to one of a small, personal user environment, the reader can assess the developments in computer hardware as they apply to music instruction. Several unique problems are presented to the developers of computer hardware when the concept of audio instruction is considered. Since music is an aural medium, computers had to be developed to generate sound, play back music, record music performance, and analyze music performances.
The development of sound generation capabilities did not reside in the development of computers per se, but in the development of the peripheral devices which attach to the computer. The methods of appending equipment to a computer have expanded rapidly within recent years, with interface equipment becoming more and more standard. In 1970 any peripheral device developed for a computer tended to be of unique design for a single computer. A sound-generating peripheral device developed for one computer would not necessarily work on any other computer. This problem has been resolved with industry conventions for standardization.
This tying together of equipment is probably the hardest area for the computer novice to comprehend. The fact that a tone generator can operate on one computer which uses the BASIC language and will not function on a similar computer which also uses BASIC is difficult to explain and maddening to the potential computer user. The variation between computer types, models, languages, and construction perpetuate these problems. Normally an interface device is required to connect any peripheral device to a computer or to a computer terminal. Such devices can be quite simple in construction, but are essential to direct information to and from the computer. One computer may store information differently than other computers, necessitating a particular connection. As procedures become more standardized the problem of developing music peripherals becomes more manageable. Indeed, current trends in computer design give evidence that standardization will be common in the future.
COMPUTER MUSIC HARDWARE
The hardware which is needed for CBMI is peripheral equipment to computers. The peripheral equipment and the required interface equipment become the main focus of what we have termed "computer music hardware." The hardware required for operational instruction in music has been envisioned for some time. Peters outlined the requisite hardware components for CBMI in 1977,1 as did Hofstetter the same way.2 These components fall into three categories:
- Random-access audio playback/record
- Computer sound synthesis and generation
- Computer analysis of music performance
The first area of sound reproduction is that of random-access audio. This method of recording information is that of using prerecorded examples on tape. The idea that computers could search for prerecorded information and play it back to students was researched as early as 1969 by Deihl.3 But the problem with using conventional tape recorders (reel-to-reel) was the length of time it took to retrieve the information. The process could take several minutes if the audio tape had to be rewound or searched.
The demand for random access of prerecorded information was pressed by those who wished to be able to move a student through listening material more freely. If the student wanted to repeat an exercise, the random access feature would allow for a quick repeat of the listening example. This type of device, first used for music instruction by Placek4 in 1970, allowed such quick access of information. Each listening example was available to students within half a second, a capability which is quite acceptable for music instruction.
The device used by Placek was also capable of recording student performances. This added feature was demonstrated on the PLATO computer with a lesson by Peters and Sanders in 1975. Students could hear a model of a sight singing exercise, hear a reference pitch and tempo, then record their performance. The major drawback in this demonstration lesson was that the computer did not judge the music performance. The students were required to compare the prerecorded example with their own recorded performance.
The random access audio recorder continues to be refined commercially as a means of presenting verbal and musical examples for students in quick, random order under computer control. Limitations have been the fidelity of the sound, the length of examples which could be recorded, and the cost of the equipment. Other problems with this technology include the problems of selecting audio materials for the computer, recording the material, and then duplicating copies for multiple usage. The recording and duplicating process requires sophisticated audio equipment, which adds to the cost. At present the cost for duplicating one floppy disk can range from 10 to 25 dollars. Random access equipment now includes cassette tape equipment and floppy disk equipment which can be controlled by various computers, including microcomputers.
The application of this computer hardware to music instruction is very attractive. The use of music examples, played when needed or desired by the student, can enhance computer lessons in music appreciation, ear training, and music methods. For example a computer can present a student with recorded examples of stretto, Baroque music, poor clarinet tone, or Dorian mode. Eddins5 has outlined the use of random access audio in music appreciation on the PLATO computer system, based on his work at Southern Illinois University in recent years. His work has demonstrated the feasibility of the random access of audio information when used in a learning environment at the college level.
COMPUTER SOUND SYNTHESIS
The most common thought that comes to mind when discussing computers and music is that of digital to analog (D to A) sound generation. This technique has been developed by composers in several forms to use computing power to create musical compositions. Some of the same techniques have been used to generate sound for CBMI. The computer music techniques developed in the late 1950s and the 1960s have been used in the construction of peripheral devices for music examples in computer instruction in theory and ear training.
The digital to analog (numbers to sound) concept allows for the storage of sound in digital form or the generation of sound from digital form. These are two slightly different approaches to sound synthesis. In the first, music or verbal messages are performed, with the results stored in digital form. This process is beginning to surface as digital information on phonograph recordings. The approach requires that each audio example be performed and essentially recorded; however, the recording process is one of digital storage, not an electrical signal. This process is also the basis for the operation of the Texas Instrument "Speak and Spell" educational toy.
The digital storage of musical examples has not been pursued in the development of CBMI to date. The process of prerecording examples and the amount of digital storage required has made several other alternatives more attractive to the developers of computer hardware. One alternative is computer generation of the digital information required to generate music. In this computation process, the computer can be programed to generate a series of numbers which can be converted into an audio signal. No storage of digital information is required, only computer programs to compute the required digital information.
The advantage that computer-generated sound has in writing computer lessons is that the computer can randomly generate a series of exercises for students, based upon prescribed limits. The computer can shape instruction for each student individually, even as the student works through the computer-generated materials. Such a device was developed by Gooch6 in 1973 at the University of Illinois. Later application of the technology was used by Hofstetter7 at the University of Delaware. A complete evaluation of this approach to computer instruction in ear training has been detailed by Hofstetter.8
Various peripheral devices have been developed as D-to-A sound synthesizers for microcomputers. At least two of the "music boards" as they are called, have been used effectively in ear training in recent years. The microcomputers can be used by one student using one lesson at a time, but the micro technology has been made available on low cost computers so that a large number of schools can afford computer-generated music programs such as those developed by Hofstetter.
COMPUTER ANALYSIS OF MUSIC PERFORMANCE
The third area of peripheral hardware in development for CBMI involves the ability to judge music performance. This task is the most formidable to undertake when one considers the many variations in music performance that can be considered correct. Even when restricting the computer judgments to pitch and rhythmic performance, acceptable levels of variations exist.
Two devices have been developed to investigate the feasibility of judging music performance. The first is the piano keyboard, which has been interfaced to a computer to judge keyboard performance. Kent9 in an early study judged this approach to individualization of keyboard instruction feasible. The use of various keyboards with computers continues to be developed. The computer can be programed to judge whether the student has pressed the correct key and to accommodate variations in rhythmic performance.
The harder task is that of judging the accuracy of a singer or an instrumentalist by analyzing the sound produced. This process is the reverse of the D-to-A process. The analog to digital (A to D) process requires that a computer analyze sound and convert it to digital information. The amount of information analyzed and the number of times that the computer checks the electronic signal directly affects the complexity of the task. As with the recording devices, the requirement for effective teaching is immediate feedback. A device that takes five minutes to inform a singer that he or she is out of tune is unacceptable for CBMI.
Devices have been developed which can be used to judge music performance and these devices have been interfaced into computers for CBMI. An early study by Peters10 started in 1969 found it feasible to judge trumpet performance (pitch/rhythm) by computer. The cost of developing this type of music performance judging device has been high. The control needed for ambient noise persists as a problem since the computer judges any sound, musical or unmusical, as part of the performance. Extraneous sounds such as foot-tapping can result in erroneous judgment of music.
As with the D-to-A devices, recent technology has made the analysis of sound available on low cost microprocessors. At this writing, the sophistication of these devices is not at an acceptable level for music performance. While the devices are capable of pitch analysis, the data generated quickly overpowers the small computers so that only a few notes could be analyzed at a time. The use of larger computers continues to allow for longer or even continuous evaluation of music performance.
The source for the most current information on computer hardware is computer journals and magazines. The technology is developing at such a rapid rate that announcements of new peripheral devices are made monthly. Yet the three basic categories of peripheral devices remain constant and will not change in the near future. The newest available technology that will be used with computers and instruction in the next few years is the video disk for storage and retrieval of video and audio information, and the bubble memory for vastly enlarged computer storage. These developments continue to make the computer hardware more adaptable to the music classroom at the college level and at the public school level.
1C. David Peters, "The Complete Computer-Based Music System: A Teaching System—A Musician's Tool," Proceedings of the 1977 Winter Conference of the Association for the Development of Computer-Based Instructional Systems (1977), pp. 93-100.
2Fred T. Hofstetter, "Music Dream Machines: New Realities for Computer-Based Musical Instruction," Creative Computing III, No. 2 (1977), 50-54.
3Ned C. Deihl, Development and Evaluation of Computer-Assisted Instruction in Instrumental Music (Final Report) (University Park: Pennsylvania State University, 1969).
4Robert W. Placek, Design and Trial of a Computer-Assisted Lesson in Rhythm (dissertation, University of Illinois , University Microfilms No. 73-17.362).
5John M. Eddins, "Random Access Audio in Computer Assisted Music Instruction," Journal of Computer-Based Instruction V, No. 1 and 2 (1978), 22-29.
6Sherwin Gooch, "PLATO Music System," Proceedings of the Conference of the Association for the Development of Computer-Based Instructional Systems (1978), pp. 314-324.
7Fred T. Hofstetter, "Results of the 1975 Delaware PLATO Project," Proceedings of the Association for the Development of Computer-Based Instructional Systems (1976).
8Fred T. Hofstetter, "GUIDO: An Interactive Computer-Based System for Improvement of Instruction and Research in Ear-Training," Journal of Computer-Based Instruction I, No. 4 (1975), 40-42.
9William P. Kent, Feasibility of Computer-Assisted Elementary Keyboard Music Instruction (Final Report) (Falls Church: System Development Corporation and Wichita Public Schools, 1970).
10G. David Peters, Feasibility of Computer-Assisted Instruction for Instrumental Music Education (dissertation, University of Illinois , University Microfilms No. 74-14.598).