THE VIDEODISC REVOLUTION

I. Sebestyen International Institute for Applied Systems Analysis, Laxenburg, Austria

RR-82-27 July 1982

Reprinted from Electronic Publishing Review, volume 2(1) (1982)

INTERNATIONAL INSTITUTE FOR APPLIED SYSTEMS ANALYSIS Laxenburg, Austria Research Reports, which record research conducted at IIASA, are independently reviewed before publication. However, the views and opinions they express are not necessarily those of the Institute or the National Member Organizations that support it.

Reprinted with permission from Electronic Publishing Review 2(1):41-89, 1982. Copyright© 1982 Learned Information Ltd.

All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopy, recording, or any information storage or retrieval system, without permission in writing from the copyright holder. iii

PREFACE

In 1981 the International Institute for Applied Systems Analysis began a program of research on the impacts of information technology. This work was planned as a cluster of related tasks, rather than a unitary whole ; and, indeed, the various activities were in­ tended to explore various possibilities, and therefore were not necessarily predicated on the same set of technological and societal assumptions. One of these clusters has been concerned with the applications and social impacts of Viewdata (Videotex) systems - and the author of this report is a member of the re­ search team that has been carrying out the work. This is only one of a number of papers that team members have written either jointly or individually. Too, its content intersects that of papers from another task con­ cerned with computer-based messaging (or conferencing) systems. An appendix listing related publications appears at the end of this report. The potential social impacts of both Viewdata and computer-based messaging sys­ tems are immense - the basis for the inquiry whose results are reported here.

ALECM. LEE Chainnan Management and Technology Area

ELECTRONIC PUBLISHING REVIEW

The Videodisc Revolution

Istvan Sebestyen

Abstract: This paper attempts to make a comprehensive analysis of present and future videodisc technologies and a thorough examination of the impacts of this technology on different information application classes and on other media. First the basic principles of this new technology are described. This is followed by a summary of some major hardware and software functions of such systems. In the subsequent chapter, the extremely broad range of videodisc applications is dealt with . In the final summarizing chapter some conclu­ sions are drawn pointing to the vast potential of this new technology, which according to the author, could lead to a new revolution in the information and entertainment industry.

1. Introduction robots perform their duties in manufacturing; We are at the beginning of a new revolution in industrial processes are increasingly the use of information technology, triggered being controlled by real-time computers; the off by the rapid developent of new technolo­ microelectronics revolution has started gies in the field of electronics. While the winning one battle after the other in daily life. performance of processors, primary and For example, today we can find cheap secondary storage equipment, input-output pocket calculators in practically every house­ devices and telecommunication devices has hold, and low-cost digital watches controlled improved at an accelerated pace in recent by internal microprocessors have practically years, the unit prices of these items have overrun and replaced traditional watch­ dropped dramatically. New technologies making methods. Furthermore, advanced have found their way into many different areas microelectronic technology has moved into of application. so-called "consumer electronics". In the A major step forward was the emergence of newer types of TV sets, hi-fi equipment, computers in telecommunications, which videotape players etc., microelectronic devi­ marked the beginning of a new era: "tele­ ces such as processors ¥e being applied en matics", a word coined by S.Nora and masse. The "video arena" in particular is A. Mine [7] . being invaded by these modem technologies. The convergence of information and tele­ New applications such as broadcast and inter­ communications technologies was only a first active videotex services (e.g. Oracle and step and was followed by others, as new Prestel in the United Kingdom) are marking information technologies gradually started to the beginning of a new era, which might be influence other fields: processor-controlled called "videomatics" , or a convergence of video and information technologies. The author is with the International Institute for With the advent of laser technology and Applied Systems Analysis, 2361 Laxenberg , optical fibers, revolutionary changes are now Austria. expected in the field of information storage

1982, Vol. 2, No. J 41 ELECTRONIC PUBLISHING REVIEW

About the author

Istvan Sebestyen Dr Sebestyen studied Electrical Engi­ neering at the Technical University in Budapest; he received his M.S. in 1970 and his Ph.D. in 1974, both from the Technical University. He then joined, as computer hardware engineer, the computing laboratory of the Institute for Coordination of Computer Tech­ niques. His special field of interest during this period was software engin­ eering; from 1974 to 1977 he participated in a number of joint research projects on this subject with Siemens Ag, Munich. From 1977 to 1978 he worked as an expert on Management Informa­ became the lnstitute's Computer tion Systems at the UNIDO Headquar­ Communications Services Depart­ ters in Vienna, and joined llASA in ment. Here he is carrying out impact 1978. While at llASA he has been studi.es on new information techno­ involved with the activities of the Infor­ logies such as videotex, teletext and matics Task, which early in 1981 videodiscs. and telecommunications. In this paper, an attempt will be made to collect all the major attempt is made to understand, analyze and classes of videodisc application. Some of assess the impacts of laser and optical storage their requirements can be met by other techni­ technology on the converging fields of cal means and media-perhaps in a less information technology and consumer elec­ convenient way and at a higher cost. Other tronics. The article focuses on a new device, applications,however,are brand new - and which we will call 'videodisc'. probably will have considerable impact on The structure of the article is as follows . future life styles. Section 6 briefly summar­ First a short overview of videodisc technol­ izes the study and the major conclusions to be ogies will be provided, showing why they drawn from it. It is hoped that by that point will achieve a major breakthrough. In Section readers will agree with the author: the Video­ 3, the "dualistic" nature of videodisc techno­ disc revolution is coming and will greatly logy will be explained: on the one hand such affect our lives, as did the book, the radio, the devices will be portrayed as "interactive TVs" tape recorder, the telephone, and the TV be­ - as looked at from the consumer electronics fore it. point of view - and on the other hand, they will be seen as cheap mass storage devices for random access of coded information - as 2. A Description of Videodisc looked at from the information technology Technology area. In Section 4, the hardware and system Harry Collier (22] calls the videodisc scene at software aspects of videodisc systems will be the beginning of the eighties a "jungle". dealt with briefly and major operational Many forms of ·videodisc technology, functions of videodisc players will be designed for slightly varying purposes, are discussed at some length. In Section 5, an now in existence or are being developed. It is

42 1982, Vol. 2, No. 1 ELECTRONIC PUBLISHING REVIEW

not the aim of this paper to make a full analy­ coding; in the case of digital coding only two sis of the technology itself; this is done at analog values exist: zero and some discrete some length by other authors. See, for example, value. This means that on the surface of the Barrett [23] and Sigel et al [29]. Neverthekss videodisc record either there is a pit under the before discussing the application classes and information sensor (information" l ")at reading possible impacts of this new medium, we time or there is none (information "O''). should briefly discuss the piece of hardware From the storage and reading point of that will be the focal point of the study. view, analog and digital videodiscs are thus The first videodiscs, developed in the early very similar. This gives the hope that one day seventies, were intended for use in the enter­ it will be possible to produce videodisc devices tainment industry. Since then , however, which will be able to handle both type of strong interest has arisen for their use in infor­ codes - analog and digital. As we will show mation and document storage, retrieval, later, this would bring major advantages to computer-aided instruction , etc. An essential the system. At present no such system on the feature of this technology is that it lends itself market or in development is known to the readily to the integration of text, image, and author. audio information and, at the same time, it is Why is it that already at the beginning of programmable and can be randomly the videodisc era, two different types of accessed. However, each of the above types coding techniques are being applied? of application places different requirements The simultaneous emergence of analog and on the equipment's development, and thus digital coding techniques has both historical after only a few years of development, a and technical explanations. Analog storage number of different forms of videodisc tech­ for videodiscs was linked to the effort to store nology have come into being. These tech­ analog television frames according to NTSC, nologies can be classified according to the PAL, or SECAM standards on videodisc. It following criteria: has been found that the best type of storage is analog versus digital information coding achieved when pits are engraved at the speed - mechanical (contact) versus non­ of the FM carrier frequency of about 8 MHz mechanical (non-contact) pickup of inform­ onto the surface of the videodisc, where the ation from videodisc records length of the individual pits is proportional to - mechanical replication versus laser beam the modulation on the carrier frequency recording of information on to records. (Figure 1). With this technique any TV signal

Analog versus Digital Information Coding At present there are several systems using each of the two types of coding. What do analog and digital coding mean in the case of videodisc technology? As will be described at some length in the following chapter, the information units on a videodisc record are tiny pits engraved in the I I I surface of th e disc. In the case of analog a~~ coding the information stored is coded in the length of the pits; the longer or shorter the pits are, the higher or lower the analog value (e.g., voltage) produced by the sensor, which reads the information serially. Di gital coding Fig. 1.Micropits e 11 gra1•ed 011 the disc of an can be regarded as a special case of analog analog optical recording system.

1982 , Vol. 2, No. I 43 ELECTRONIC PUBLISHING REVIEW

can be recorded and read, including their such as for analog videodisc players, the usual properties: thus about 200,000 pixels "color TV" type moving frames. Or is it per videoframe, sound and color coding, and possible that one day a videodisc player will in the vertical blanking intervals even teletext enter the market which will combine both type digital information, could be stored. coding techniques and utilize their advan­ Digital storage applied to videodisc tages when handling different information emerged when Philips in the Netherlands application classes. realized that through this technology much Depending on the type of information more coded alphanumerical information stored, analog and digital techniques have (text, data, computer programs) can be stored their advantages and disadvantages in highly economically than by any other comparison with each other. information storage techniques. (a) Coded alphanumerical information It was soon realized, however, that in such as digitally coded text, data, and principle other types of information as well computer programs can be stored ideally should and can be stored on videodisc: first of through digital videodisc technology. Digital all, on the analog videodisc 'color TV' type coding and storage allows the most sophisti­ freeze-frame pictures, which is no more than cated data compression and error recovery. In repeated access to the same analog TV frame, addition this type of data is immediately ready and second, on the digital disc the storage of for further data handling and processing 'Facsimile' type black and white static purposes without any special preparation or pictures with high resolution. conversion. It may sound strange, but in Since then, it has been recognized that if principle, it is possible (with the addition of the same videodisc handled all the above­ some hardware features) to store digitally mentioned classes of information, i.e., coded alphanumerical information through - coded alphanumerical information, analog videodisc technology as well, in a - 'Facsimile'" type static pictures, "pseudo" digital fo1m . In this case, obviously, - 'Color TV' type freeze-frame pictures, the degree of data compression is less favorable - 'Color TV' type motion pictures, and than that used in the real digital technology. - audio (sound) information, At present, there are two possible ways in one would get a revolutionary new piece of which digitally coded information can be equipment, which would allow not only vast mixed with the standard analog television amounts of storage but also flexible mixing of signal to be stored on analog videodiscs coded alphanumerical , audio, and video records: one technique is described by Barrett information. This would lead to new , revolu­ [23] and Sigel [29]. They suggest to replace tionary fields of applications. the active picture material in the standard Along this line we see the same trend of color video signal with data, while retaining development for both analog and digital the various synchronizing signals that are videodiscs. Analog videodiscs will be used by playback electronics to identify an developed to be able to handle coded alpha­ analog video signal. Encoded in this fashion , numerical information in an efficient way; according to Barrett, at a data rate of digital videodisc will be able to handle voice, 7.16Mbits per second in an NTSC signal , a 'color TV' type freeze-frames, and maybe total of 375 stored bits per line can be stored. later at some point - although this is far The total bits per TV picure could be 185 ,625 down the road - produce 'color TV' type bits. Utilizing a standard 108,000-track motion pictures. analog videodisc a total storage capacity of Which coding techniques will actually 2 x 10 10 bits per disc could be reached . "win" is almost impossible to predict at this Assuming 30-40% for error recovery time. Both technologies will predominate in purposes this would lead to the impressive certain information classes for a long time, figure of 1.2-1.4 x 10 10 user bit per disc.

44 1982, Vol. 2, No. 1 ELECTRONIC PUBLISHING REVIEW

The other technique which would allow the disc systems. However, at present, no such storage of a limited amount of coded alpha­ application is known to the author. numerical information on videodisc utilizes (d) 'Color TV' type motion pictures are typi­ the opportunities provided by the well-known cally suitable for analog videodisc systems as teletext technology mentioned before. described above. It seems at present that the According to Sigel [29], as it is implemented coding of motion pictures on digital video­ in Europe, teletext can transmit as many as discs is beyond the presently known disc 160 alphanumerical characters hidden within storage and data processing speed capabili­ four scanning lines of television frames each, ties . According to J. W. Klimbie [30] about an which are restored to "pages" of alphanumeric 80-100 MbitJsec processing speed would be and crude pictorial data by a special decoder. required and on the presently known digital On a full videodisc record (P AL/SECAM) videodisc only about a 10 second motion 1.44 x 107 characters - the full Encyclopedia picture program could be stored. Brittanica having roughly 2.2 x 108 charac­ (e) Audio (sound) information at usual TV ters could be stored. broadcast quality can be easily stored on One possible application of such character analog videodiscs. The storage of sound on might be to provide subtitles to videodisc digital videodisc is in principle also possible. motion picture programs for handicapped The basic principle behind recording voice people or for subtitling a program in a foreign on digital videodisc players will be PCM. language. PCM is an acronym for Pulse Code Modulati­ Another might be the transmission of data on, a typical digital technique. Conventional into a computer while pictorial information is audio and. video systems, relying on analog being watched; the videodisc controller could techniques, handle sound and video signals be fed with programming instructions to without modification. The key principle modify the videodisc control program, as will behind PCM is that both the frequency and the be explained later. Since teletext is also avail­ levels of the signals are handled in an inter­ able with standing frames, it might be used to mittent form regarding both the time and their provide descriptive text for freeze-frame levels. Sound level and frequency are encoded pictures. with a binary coding system using only" 1" or (b) 'Facsimile' type static pictures are more "O" and are then recorded as equal-amplitude suitable for digital videodisc systems. Their digital pulses. And here lies the secret of high quality resolution according to CCITT's PCM's capability for yielding specifications Group III standard can only be assured by superior to those of conventional systems. digital coding technology. Such pictures, however, can only be printed out by appro­ What are the advantages of digital audio priate matrix printers. This type of informa­ recording in consumer electronics? While tion is less important for the residential video­ analog recording techniques have been disc user, but will find major application in remarkably improved, they are still limited by office automation and document filing a number of weaknesses that result in distortion systems. Storage of 'facsimile' type static and dynamic range limitations. These limita­ type pictures on analog disc in digital fashion tions are inherent in the tapes, heads, and is principally possible as described in the pre­ other mechanical parts, and it is virtually vious paragraph. However, at present, no impossible to eliminate them completely. such solution is known. With digital systems, the sound or image (c) 'Color TV' type freeze-frame pictures are signal is recorded and transmitted in the form primarily suitable for analog videodisc of digital codes, and this has a whole host of systems. Appropriate digital coding, perhaps advantages, among them: using some PCM decoding methods, would - Wide dynamic range in principle make it possible for digital video-

1982, Vol. 2, No. 1 45 ELECTRONIC PUBLISHING REVIEW

- Flat frequency response regardless of It should be noted that the technology for input level digital recording on optical discs is presently - Extremely low distortion developing in two different directions: - Superior transient characteristics 'videodiscs' and 'digital audio discs.' The - No deterioration even with repeated storage principle is basically the same on duplication both. However, usually the quantity of data - Error correction capability. stored is considerably higher on videodiscs because of its physical size. Needless to say, The basic principle of the PCM recording in principle, it is also possible to store voice process is shown in Table I [ 16]. information on videodisc devices.

Table 1. Principle of the PCM recording process

What is the basic principle of operation behind PCM?

There are three stages in the di11;italization; sampling, quantizing, and encod­ ing. The quantizing and encoding functions are handled by the AID (analog-to­ digital) converter, and the reverse function is performed by the DIA (digital-to­ analog) converter. (1) Sampling The sampling circuit serves to convert the analog signals into PAM (pulse amplitude modulation) signals at fixed time Intervals. These intervals are com­ monly known as the sampling intervals and their reciprocal as the sampling fre­ quency. Analog signals whose frequency is less .than half that of the sampling fre­ quency can be completely restored to their original form even after digitalization. The sampling circuit is usually associated with a holding circuit since there are limits to the operating speed of the quantizer. The sampling frequency for digital audio disc is usually 32,000 KHz, for studio applications, higher. (2) Quantizing Quantizin& refers to the process where the levels of the individual sampled signals are divided into fixed steps. The tlner these steps, the more accurately the signal level is expressed, and the better the dynamic range and signal-to-noise ratio. If the number of quantizing bits i.s taken as "n", then the number of steps which can be expressed becomes 2". In equipment which aims to produce top­ grade hi-ft. "n" is usually equal to 14-16 bits due to the sound quality, economic feasibility and other considerations. In PCM video systems 3-4 bits can be used to express brightness of a spot. (3) Encoding The signals which have been converted into discrete values with respect to both time and their amplitude by sampling and quantizing are further converted into pulse codes and recorded. A binary code (0 or 1) is normally used for this purpose.

1010100101100 ~ .. ~ .. B .. rlJlJV1JL An•log Signal Sampling Ouantizalion PCM signal

PCM Recording Proceu

46 1982, Vol. 2, No. 1 -- Table 2. Characteristics of present and future videodisc systems ~ Videodisc lnformulion lnformollon pickup Pickup Information recordtnc Media Random Type or information stored Price (USI) ~ ayslem slornge prln- Mechcinical Laser beam mhcln1 ace es."' video wldeo coded videodisc mechanical non- cl pie repllcatlon recordlna capablllly capability audio motion data/ control analoa dteitail •lnale equipment record contact contact (alampin1) ("write only") frames led pro1rHm "'<: 0 Selecla Vision x x capnci- x x x x 600 16 (RCA) live and above end above

Video HiKh capacl- Density x x live x x x x x x ? ? (JVC)

Philips! x optical Magnc1voK x x x x x 700 16-24

MCA/ op lien II Pioneer x x x x x x 760 16-24 Discovison x A.ssoclalion x x oplical x x x x x x 2,600 6-26 fllJMIM CAl

Thomson-CSf x op lie al dtsc x x x x x x x 2,000 10-20

Uigilul 100,000 250-300 Optical (In 1982) (In 1982) Recu,-dina x x optical x x x x Of:M price (0011) 10,000 100 Philios ..J!!!...t085l lln 1967)

Document nline 4~ . 000- 140 •ystem "Df'·2000" x x oplico.1 x x x 60.000 l'o:d1iba lln 19821 fin 19821 ...... m Optical disc di1llnl dtttn x optical recorder/player x x x x x ? ? !CRCA) z~ DREXON di::tc c=; (Drexler Tech· x x optical x x x x 3,500 50 "'Cl 110lo uv C2!£...) c: m "ldeol videodisc ? optional 500- ...... systems" (not ? x oplJcol x feature x x x x x x x 5000 IHiO c;; ovoilnble vell :::z::z C) ::a ~ .,, ;;; -.J :e ELECTRONIC PUBLISHING REVIEW

In principle, about IO hours of super hifi characteristics of the videodisc record under a quality sound could be stored on an average laser beam spot. This technique will be dis­ digital videodisc record. It is no doubt that cussed in more depth, since it is regarded as there will be special applications when this more suitable for major videodisc applica­ capability of the digital videodisc recorder tions. will be utilized. As we have seen, at present no videodisc Mechanical replication (stamping) versus system, be it analog or digital , exists which laser-beam recording of information on to would fulfill the basic requirements - vast records information storage and media mixing capa­ Both methods of record production are of basic bility - of an 'ideal videodisc system.' importance for the videodisc applications that Whether it will be an analog, digital , or a follow. The "mechanical" manufacture of mixed system is hard to predict at this stage. videodisc records will allow mass production of programs for videodisc players, including Mechanical versus non-mechanical movies, encyclopedias, electronic diction­ information pickup from videodisc records aries, telephone directories, railway time­ Although there are a few systems with the tables and similar products that can be sold mechanical pickup on the market (one, the and distributed in large quantities. At present, Selectavision of RCA, for example, uses a information for "mechanical" replication special needle on a grooved capacitance usually arrives at the videodisc factory on disc), they seem to be less important, as this videotape. technology has several disadvantages that In the case of Philips videodiscs [26]. the exclude it from being used for the applica­ beam of a 100 MW laser is modulated with tions to be discussed throughout this paper. the tape 's signals and is focused on· to a One drawback is that it does not allow random rotating glass disc coated with a photo­ access to the recorded information , which sensitive emulsion. When the exposed disc excludes its use for information retrieval has been developed and etched, it contains purposes. Another deficiency is that so-called some 25,000 million pits. This master disc is "freeze frame" pictures cannot be shown then coated with silver and replayed to check using this method (these deficiencies are not its quality. If the disc be acceptable, the next true for the VHD system of JVC). And last stage is to electroplate it with nickel, and coat but not least, the lifetime of both needle and it with aluminium to produce a "father" disc: record is limited due to the high rotation speed this negative replica of the master is stripped of the disc and the mechanical contact from the glass, and electroplated with nickel between needle and record. Therefore we to produce a "mother", or positive replica of have limited the following discussion to tech­ the original master. From this another electro­ nologies with non-mechanical pickup capab­ nickel plating process produces a negative ilities. "son" or stamper, which is used in the produc­ Pickup from videodisc records can be either tion process. The stampers are used to make capacitive or optical. The basic principle of "half' discs, pairs of which are glued together the capacitive pickup techniques li es in sens­ to make the final double-sided product. ing the change of the electrical capacitance However, at the time of this writing, there between a conductive surface on the disc and remain some "snags" in this production a thin metallic electrode carried by a stylus process. In a UK factory, according to press tracking either in the groove (mechanical reports [26]. failure rates vary from 10 to 90 pickup) or smoothly above it on an "air per cent and are quite unpredictable. It can be cushion" (non-mechanical pickup). Optical expected, however, that these "infantile dis­ pickup techniques are always non-mechanical orders" in videodisc record production will be and are based upon the change in the optical overcome in the future.

48 1982 , Vol . 2, No. I ELECTRONIC PUBLISHING REVIEW

The other type, "non-mechanical" infor­ A description of optical videodisc technology mation recording is carried out withh a laser Optical videodisc technology has changed beam, which bums the information on to an little since 1978 when Philips launched its empty videodisc record in a sequential first laser-optical system (Philips/Magnavox) recording process. This technology does not [I]. A 30cm record, superficially resembling allow real mass production of videodisc an audio disc with a grooveless surface, is records; it is however, of utmost importance coated with a reflective material (Figure I). The for applications such as archiving, filing and disc's surface is covered with a spiral of tiny office automation. A similar technology used pits. For playback, the disc is rotated at high in the Digital Optical Recording System speed on a gramophone turntable while a finely (DOR) of Philips is described in the following focused laser beam tracks from the inside out­ section in more detail. In Table 3, the applic­ wards. As each pit passes through the laser ation classes are listed according to the beam it makes a minute alteration in the recording technology to be used. The switch­ pattern of the reflection off the disc's surface. ing point between the two recording technolo­ These changes are detected by a photosensor gies lies at a production level of around 100 arranged along .the same axis as the laser copies. beam. The photosensor produces an electrical signal that can be decoded to produce color television pictures, stereo sounds, or (in dig­ Table 3. Applications for mechanical replication and laser beam recording of information. ital systems), special codes for digital infor­ mation. A similar technique for optical stor­ /14echanicalrecording Laser beam recording age such as the one used by Philips' other (replicating ofrecords) system, the experimental Digital Optical Recording (DOR) system - a one-time writing Entertainment Mass storage and read-only system - is shown in Figure 3. Encyclopedia Archives of data The DOR system uses a 30 cm disc formed Automated dictionary Program archives from two glass substrates placed back-to­ Electronic directory Document filing back in a sealed airtight construction. The so­ Information bank Electronic publishing formed experimental optical disc has a spiral Timetables groove equivalent to 45 ,000 usable tracks. Education The disc is divided into 128 sectors. Each Mass storage, archives track/sector combination is given an indiv­ (many copies) idual segment address. In this way segments can be found in random access mode. DOR is in principle an empty disc, as are magnetic In dealing with the above classifications, tape and discs . DOR discs can be written this article will focus primarily on digital (once only) by the user. The user may write videodiscs with multimedia mixing capability about 1 ,000 bits of information in each seg­ that use non-mechanical pickup (primarily ment. Thus on the experimental DOR system optical) reading techniquess allowing random it is possible to store 5 x I 09-bits per side; since 10 access to frames and with mechanical (stamping the disc is double-sided, 10 - bits can be of records) and laser-beam recording tech­ stored per disc . niques. As shown in Table 2, such systems Philips sources say [30] that in the final which combine all the above requirements do version of DOR, double sided 2 Gbyte user not exist yet, although their characteristics are information (1 .6 x IOwbits can be stored . individually part of one or other presently The disc is pregrooved, and the recording known systems. In Table 2, no attempt is surface is completely protected; it is engraved made to compile all videodisc systems and read through the glass substrate. Engrav­ presently known. ing is carried out by so modulating the diode

1982, Vol. 2, No . I 49 ELECTRONIC PUBLISHING REVIEW

Laser beam Figure 2. Simplified diagram ofa videodisc record based on the laser beam reflection principle (Philips/ Magnavox) [I]. Write laser Rud laser .-...... ·.·.···n' .a ...... ··=·~~?:

Figure 3. The optical disc memory for data storage and retrieval. Some - such as the Philips DOR system - employ laser light to write data by burning holes in the medium on a spinning disc. The laser for writing the data is shown at the far left. A laser for reading the data is shown at the center top. Its light reflects from the disc at the places where no hole has been burned, then makes a second passge through the optical train of the device, arriving back at a beam splitter with a vertical polarization. This reflects the beam to a detector array. The reflected light also yields feedback signals for control of tracking and foc using [2].

50 1982, Vol. 2, No. I ELECTRONIC PUBLISHING REVIEW

laser beam that a hole less than one micron in fields of information technology such as mass diameter is burnt in the sensitive layer; once storage of information and random retrieval. engraved, the disc cannot be rewritten and In order to assess videodisc technology and its retains its properties for at least 10 years, potential impacts, we shall follow a two-sided making it suitable for archival purposes. approach: first, we will look at it from the Another type of system , used, for example, consumer electronics - mainly audio-video by Thomson-CSF, (another analog video­ recording - point of view, and then we shall disc) is based on the so-called laser transmis­ examine the phenomenon from the informa­ sion principle. The beam shines through the tion technology perspective. Finally, the "du­ disc that has been previously burned by a laser, alistic" nature of this technology will be the photosensor being situated on the other shown, classifying videodiscs as typical "vi­ side of the record. Thomson's system utilizes deomatics" devices. The term "videomatics" a flexible rather than a rigid disc (30 cm) has been defined as the convergence of video carrying approximately 30 minutes' (or and information technologies on the analogy 54,000 frames) on each side. Since it is a of the term "telematics" - or French "tele­ transmissive system, though, any of the matique" - coined hy Simon Nora and Alain 108 ,000 frames can be randomly accessed Mine in their report to the President of without flipping over the disc , simply by France {7]. When Nora and Mine use tele­ changing the focal point of th e optics from matics to describe the growing link between one side to the other. computers and telecommunications, they While conventionai audio discs usually point out a new concept: the computer is not have only one long track of information along only a computation machine but also a tele­ a spiral line going from the outside inwards on communications device. The term "video­ each side, video discs have numerous tracks matics" is used to describe the growing inter­ (e.g. up to 54 ,000 per side on Discovision's connection between computers and audio­ PR-7820 model, manufactured by Disco­ video technologies such as television and video vision Associates (subsidiary of IBM and recording. The term videomatics defines the MCA, Inc), each separated by 65 millionths computer not only as a computational and tel­ of an inch along concentric circles contain­ ecommunications device, but also as a ma­ ing independent frames of information. When chine for controlling, mixing and processing the PR-7820 is working in video mode, each textual, audio and video information. track contains information for one TV picture, with approximately 200,000 picture elements per frame. Sound and computer 3 .1 Videodisc technology from the consumer instructions are recorded on the disc as well. electronics point of view On the model PR-7820, a helium-neon laser Comparing videodisc technology as described is used to "read" the information from the in this paper with the presently known standard disc, which is spun on a turntable at 1800 videotape recording technology, we see cer­ rpm , 54 times the speed of a normal 33 rpm tain basic differences in recording and play­ audio turntable. back: a) most future videodisc players will, we believe, use digital signals; videotape recorders 3. Videodisc as a typical "videomatics" usually work with analog signals device b) video frames (individual pictures) can be There seems to be much confusion about th e addressed separately on videodiscs but not on possible role of videodisc technology and its videotapes impacts on different fields of consumer elec­ c) on videodisc, access to picture frames may tronics, such as video recording, hi-fi, personal be random; in videotape recorders, it must be computing, and the like - and also on various sequential

1982, Vol. 2, No. I 51 ELECTRONIC PUBLISHING REVIEW

Optical recording on video disc ~ 1118 .5"' -...... ~- ) ~ Vertical recording---_-;-_ "'= ,,,...... ____ _ ...~ ,,,..------... 107 ... Metallic mediums e:: ·~ ...= 106 ...... :ii= ..."' ..... 106 Q"'

1114 1960 1970 1980 1990 Vear

Figure 4. Data package density of storage technologies. Advances in disc technology are reflected by advances in the packing density of data, expressed in this chart as bits per square inch on the surface of a disc. In rigid discs the magnetic medium is coated onto an aluminium substrate; in "floppy," or flexible discs it is coated onto Mylarplactis. Two improvements are foreseen for the magnetic technology: the use of metal film instead of iron oxide as the medium on rigid discs and the recording of data in regions of magnetization oriented vertically, or perpendicular to the plane of the disc, instead of horizontally, the current practice. Videodisc technology might attain the greatest storage density of all [2]. 3 .2 Videodisc technology from the informa­ d) at present, videodiscs are "read only" tion technology point of view devices (and "write once" in case of digital optical recording systems). Videotape Looking from the point of view of computer recorders may be used to record and/or erase technology and comparing videodisc tech­ selected programs. nology with other (mainly magnetic disc) From the technological point of view, the technologies, the following differences can videodisc player and videotape recorder are be observed: completely different. The only similarity is ( 1) Videodisc technology (more precisely in that TV-like video information (movies, our case optical disc technology) attains the sports events, etc.) can be replayed on both greatest storage density of any of the informa­ types of device. (Incidentally, this was the tion storage technologies presently known original purpose of the videodisc design). If (see Figures 4 and 5). the only criterion for comparison were this In magnetic storage the cell size is limited by one, there would probably be no market for the magnetic particle size limit, which sets the videodiscs, since their unit prices are, and ultimate magnetic limit somewhere below 8 2 still would be, somewhat higher than those of 10 - bits/in- , corresponding to a memory cell videotape recorders. A one-hour videodisc size of approximately 2 microns. program will cost approximately five times In optical storage the cell size is limited by the ($30) that of a one-hour continuous program optical diffraction limit which sets the ulti­ on videotape ($6) [ l]. mate limit for optical recording above 10-9

52 1982, Vol. 2, No. I ELECTRONIC PUBLISHING REVIEW

.1 .2 .4 ----Optical -;;;­ .6 diffraction =Q .8 ... limit 1 .. ·e Magnetic =1==-=::-==t~:z~-~l~:::::-~;--=--=:r:=-:~ --Q> ·;;;N particle size limit

10 20 40 60 80 100 200 400 1960 1970 1980 1990

Figure S. Trends in information storage [22]

2 bits/in- , corresponding to a memory cell size (3) The storage capacity of videodiscs is on the order of 0.4 micron. These limits are remarkably high. affected by ambient temperatures and by the In Table 4, the storage capacities of various wavelength of the light used; for instance, the videodisc systems are compil~d according to limit for silver halide film is lower than this different types of storage "measurement owing to limitations on the optics and the units." No attempt is made to collect informa­ requirement to work with visible light. Figure tion on all presently available videodisc 5 gives a reasonable projection of the limiting systems; the table is intended to provide only capabilities of various storage media. a general picture of the storage capabilities of (2) The access time to videodisc is similar to present and future videodisc systems. those of the magnetic moving head discs, i.e. As an example of a typical digital storage between 10-2 and 10-1 seconds. The recorder/ device the Digital Optical Recording System playback system of the Philips DOR system, (DOR) was taken . The experimental system for example, enables a record to be accessed as mentioned previously may store double­ with a mean random access time of 135 sided 10 10 bits of user information; however, milliseconds; the very compact optical sys­ according to Philips sources [30] this will be tem weighing only 40 grams is mounted on an upgraded to 1.6 x 10 10 bits or 2 GBytes of arm driven by a lin~ar motor. The arm has an storage capacity . optical grating, which very quickly brings the If, on the experimental system, coded optics to within 50 tracks of the one selected. characters are stored, each page will require Direct reading from the disc then enables about 4 x I 04 btts. If, however, one stores rapid selection of the exact track. Less sophis­ optically scanned data according to the Group ticated videodisc players need longer access Ill facsimile standard of eight lines per times (up to five seconds).

1982, Vol.2.No. / 53 Ul Table 4. Comparison of videodisc storage capacities m .i::. ~ Storaee capacity per disc ~- ---·- ·------· --·-- Videodisc Mode of "atom le" usable "VIJJEOTEX Standard High resolu- Length of Length of ~ system Information Information digital characters Uke" resolution lion (Facsimile color TV movie audio z atorage units (pita) bits rrame1 (TV) pictures Ill) pictures program Ill program n "'Cl 200,000 pixels 4 million pixels normal speed c:: mc::: 2.l:Ho" en Experimental .. 1.3·1o' (without data :::r:: Digital Optical 1.25·101 digital 2.1·1010 l.05· 1010 (coded 15-1114 compreoslon) z Recording (digital I VTXlrame = (with data 2.5·10' - - C) Phillps (DOR) coding) 880 characters) c ompreaalon) (with dale ::D compreaslon) ~ Ho" S1 New series 2·101 2· 1o' 8·10• (without data of Phillps DOR (digital (digit.al (with data compression digital 2.3·101• l .8·1010 (available coding) coding) compression) 4·1D' - - 1983) (with data compression) Dlscovlslon .. Association 1 ! .2·1o" 1.08· lo" 1.08·10" I hour analog a -10 • I hour (IBM/MCA) - (TV mode- (TV mode- (TV mode- - (TV mode) (TV mode) PR-7820 NTSC) NTSC) NTSC)

Thomson- .. 1.08· 10" l.OB·lo" CSF disc 1.2·!o" (TV mode-NTSC) (TV mode-NTSC) I hour 1 hour analog a -10 1• - (TV mode- 8·10' 8· 1114 - (TV mode) (TV mode) TTV 3620 HTSC/PALI (TV mode- (TV mode- SECAM) PAL/SECAll) PAL/SECAM)

l.2·1o" 1.08·108 1.0B·lo" 3-3.2·103 (TV mode) (TV mode) (TV mode) (digital coding I hour "Future !.2-l.4· 1010 or or or without data (TV mode) analog (digital coding) 1>5-1.7·!01 1 analog 3·1010 1.5-t.B· lo" 1.5-1.8· 10 compression) I hour videodisc and (digit.al coding) (digit.al coding) (digital coding) 3-3.2· IO' (TV mode) 7.5 hours .._ system" 1.15·!o" In and and and (digit.al coding (digital (Teletext mode) 1.44·10' In l.OB·lo" In I.OB· Io" with data audio) ~ (Teletext mode) (Teletext mode) (Teletext mode) compreoolon)

2·!0' ~ "Fu lure 10·101 10• 4·1D" (without data 50 hours tv digital (digit.al (digital (with data compreulon) digital I 1.2· 1010 8·!010 (dig Ital videodisc cod In&) coding) compreoslon) 2·!o" - audio) ~ system" (with data .._ compreaaton) ELECTRONIC PUBLISHING REVIEW

12" EXPERIMENTAL OPTICAL DISC A4-DOCUMENT (2-SIDED)

210mm

1728 PIXELS

2376 LINES 294mm

1'1010 BITS/DISC USER SPACE ALPHA NUMERICAL: 45,000 TRACKS/SIDE 50 LINES x 80 CHARACTERS 4 kBYTE 128 SECTORS/TRACK = 1 PJCTOR!AL A4 "' 32 TRACKS PICTORIAL: HIGH RES. /WITHOUT COMPR. 2,500 M's/DISC 2376 z 1728PIXELS = 4, I 05,728PIXELS HIGH RES. /WITH COMPR. 25,000 A4's/DISC STANDARD RES./WITH COMPR. 50,000 A4's/DISC "'500kBYTE ALPHA NUMERICAL 500,000 A4's/DJSC

Figure 6 Philips experimental DOR disc. millimeter, 4 x 106 picture elements will be PR-7820 stores almost 30 billions bits of analog produced per page, or 4 x 106 bits in black information. This allows the storage of and white (See Figure 6). Using data compres­ approximately 1.2 x 108 characters in "TV sion techniques one can reduce this by a factor mode", considering that on one TV screen no of 10, so that on the average 4 x 105 bits are more than lOOG-2000 characters can be required to record a page; i.e., a disc would shown because of the relatively low resolu­ hold a maximum of 25 ,000 facsimile pages. tion of the screen. If "videotex-like" frames* As a comparison, it should be mentioned are shown on videodisc - which in the case here that an average TV picture contains of analog storage does not have a particular "only" about 2 x 105 picture of elements, meaning- this only allows one to compare it although with grades in colors. which is less with the "videotex" storage capacity of digital by a factor of about 20 than the Group III videodisc systems in which I .08 x 105 facsimile standard picture of the DOR frames can be stored. In freeze-frame mode system. The resolution quality of the TV obviously also 1.08 x 105 pictures can be picture is accordingly also lower, but for stored, a one-hour program can be played on a moving pictures it is fully acceptable. two sided videodisc record. With the new type of DOR system obviously *"Yideotex-like" frames are used throughout this the above storage figures go up accordingly. study to define the unit for the maximum of alpha­ The storage capacity of analog videodisc numerical information that could sensibly be put models are also impressive - DY A s Model on an ordinary TV screen.

1982, Vol. 2, No . I 55 ELECTRONIC PUBLISHING REVIEW

1013....------....,.--.,.-..,----.,------. Digital optical recording systems**** 1012 and Video disc Mass 1011 storage systems 1010 EBAM* -... 109 ..+-----t Single Mai:netic e- drive moving 8 head -~ 10 :ii disc ...Cl. 107 ...> Mos ram Magnetic tape Cl 6 E 10 .. The Gap" ... .. Magnetic E 105 '! fixed .....Cl head 104 CCDs** disc 103 Bipolar ram 102 10 10-s 10-7 10-S 1o-3 10-1 10 Access time (sec)

Figure 7. Total memory capacity of storage technologies in terms of access time; (data exclusive of videodiscs were taken from [4]). The capacity of the moving disc head represents maximum system capacity and not that of a single drive, which would be around 4.8. 109 bit. * EBAM - Electronic beam accessed memories (under development). ** CCDs - charge-coupled devices (new technology). *** 100 G bit videodisc already under development would be available before 1985 [22 , 28]. **** The Philips juke-box system Megadoc will be able to store around 128 GBytes of information. Philips is also working on linear selectors with a mass storage capacity of 128 GBytes (?5.12 · 10 11 bit) and an access time of around 10 seconds.

Future videodisc records will contain in between the so-called random secondary addition, as pointed out earlier, other types of storage devices (e.g. disc)and tertiary storage information; thus on a future analog system devices (e.g. magnetic tape, mass storage 1.5-1.7 GByte of digital information might systems)(Figure 6) . The new technology in be stored in a "pseudo" digital fashion and this category offers herewith new categories 1.44 x 107 characters in teletext type mode. of applications, previously unknown in the Future digital videodiscs will be able to store era of "classical computer technology", such digitally up to 50 hours of super-quality audio as on line storage of mass information such as hifi program. for encyclopedias, and voice and picture The vast storage capacity or a single record images for storage, retrieval , and processing. closes a wide gap in terms of access time In closing this gap (see Figure 7), we also get

56 1982, Vol. 2, No . 1 ELECTRONIC PUBLISHING REVIEW

in hand a technology allowing computer­ (4) The price of videodisc allows cheap, controlled manipulation of classical TV randomly-accessible mass storage of digital broadcast-like information. As mentioned information (see Figure 8). A very simple but earlier, we will see a convergence of informa­ illustrative way to look at the impact of a tech­ tion and video technologies, to be called nology is through the so-called "wedge" illus­ "videomatics", similar to the convergence of tration [5] . Its principle is shown in Figure 9. information and telecommunications tech­ The tip of the wedge represents the intro­ nologies known as "telematics". In Figure 7, duction of a product. The wedge itself traces 6 9 in the range of 10- -;-. 10- second access the evolution of any one of the product's key time, another rather broad "gap" can be features (a disc drive's capacity in megabytes, observed (called "the gap"), which cannot yet for example). be fully covered by operational storage tech­ The product can evolve in either of two nologies. basic directions. As technical advances lower

(2314) $100 000 29 MB • (1311) (2311) (3330) (3350) (3370) Digital optical 300 MB 600 MB recording systems $30,000 2 MB 7 MB 100 MB Multiplatter discs Digital_ opti~al $10,000 2 GB 10 GB recording discs 1.5 MB 12 MB ...... • (OEM) .., . ;::==::::::::::o.12...§.! t5-10GB"lntelligent ·;:::"' Cl. $3000 Cartridge discs ------~ video disc :::!: w $1000 0.4 MB 0.8 MB 1.6 MB players 0 oe::::::::::::: FI op py dis cs $300 o.~ Minifloppy $100 1965 1970 1975 1980 1985

Figure 8. "Performance" and "learning curves" for families of moving head discs. (Data for magnetic discs are taken from [3].)

100 .. Performance curve" ~ 1 = Time Figure 9. The principle of "performance" and "learning" curves. The "wedge" illustrates the two directions in which a product can evolve; a manufacturer can either keep performance constant and reduce the cost or keep the cost constant and increase performance. [5 /.

1982 , Vol. 2, No. I 57 ELECTRONIC PUBLISHING REVIEW

10 Magnetic Fixed head disc

t: ...0 ~ ..0e ::::E

Video disc 10-1 1 10 100 Access time (sec)

Figure 10. Unit costs versus access time of storage technologies (data exclusive videodisc technology are taken from [4]). manufacturing costs, the price of the product The videodisc is represented by a sixth will follow a downward curve. This path has wedge, likewise emerging between the been labeled the "learning curve" (Figure 9) cartridge disc and the floppy disc. We have and is analogous to the term "experience tried to predict its path (Figure 9). For cheap curve" used in the semiconductor industry. videodisc players we have taken as a starting The second path involves increasing the point the present analog type of system (such functionality of the product while keeping its as the one produced by Discovision Asso­ price constant. In other words, as the manu­ ciates) and have predicted that cheap digital facturing cost for the product's initial capa­ or analog optical videodisc players will be bility decreases, more capability can be added available in the future at the same price level. without raising the manufacturing cost and The price of a DY A Model PR7820/3 is about price beyond its original point. This path is US $2,500, of a universal external interface, the "performance curve" seen in Figure 9. US $225, and of a videodisc record, US $5 Normally, a product grows into a family of (without royalty for the program stored). A products by following both paths. This second category of the new device is repre­ concept offers a perspective on all computer sented by the digital optical recording devi­ hardware development, and provides a ces, which is also represented separately as an convenient way of looking at hardware tech­ OEM device. nology trends. In Figure I 0 storage unit costs are shown in Figure 8 shows the evolution of disc tech­ terms of access time . For a videodisc system nology over the last 20 years. The chart shows using mechanical (cheap replicated) video­ four wedges-one for multiplatter discs, one disc records, the average storage cost per for single-platter cartridge discs, and two for information bit is about 0.6 · I o-6 cents (US). flexible discs. A fifth wedge is emerging This is not only the cheapest storage presently between the cartridge disc and the floppy disc known on random access devices but also the (i.e. nonremovable Winchester), although it cheapest storage in the entire field of informa­ is still too early to plot its path. tion technology. In comparing this technology

58 1982, Vol. 2. No . I ELECTRONIC PUBLISHING REVIEW

Table S. Characteristics of mass storage devices caken parc/y from Kenney [24}

User Access Data Hardware Media Media Archival Device capacity time rate cost cost cost/bit life (Mbytes) (ms) (Mbits/s) (dollars) (dollars) (cents) (years)

Magnetic disc 70 35 7.0 20 ,000 2,200 3.l x lQ- s 2-3 IBM3340 disc pack

6250 bits/in 91 45 ,000 3.3 28 ,440 16.50 1.8X lQ - 7 1- 2 tape 2,400 IBM3420-8 ft reel (2000byte records)

Mass storage 462,500 16,000 7 2,400,000 188 ,000 4 x lQ - 7 1- 2 system (9 ,400 IMB3850 cartridges $20each)

Philips 2,000 1~500 5- 10 10 ,000 250 5 x lQ - 6 >10 optical disc (DOR-OEM)

Philips 128,000 3,000 20-50 200,000 16,000 1.5x10- 6 >10 MEGADOC (64 juke-box discs at $250each)

with that of magnetic moving head discs, a year if a sufficient volume is produced and single videodisc storage unit is cheaper by a sold. It is expected that in the second half of factor of 10·4(!) . Such an achievement in cost the decade the DOR disc drives could be sold savings obviously represents a revolutionary at an OEM price of around $10,000 per unit. breakthrough in the field of random access Thus the projected unit storage price in a storage devices . According to [23] it can be Digital Optical Recording system such as that expected that a storage price of between 10·7 of Philips will be around I 0-3 cent per bit in and 10-8 cent per bit stored should be achiev­ 1982 and around 10·4 cent per bit in 1985 . able by 1985 . However, it is also expected This is considerably higher than the cost per that the other storage technologies listed on information unit predicted for stamped (repli­ Figure 10 will undergo similar price improve­ cated) videodisc record systems. Therefore, ments and will cost less by a factor of at least in applications such as office automation 10. A shift in this direction can already be where digital optical recording systems seen in Kenney's comparison Table 5 [24], would be applied , the prices are still attrac­ which is based on more recent data. tive, but they cannot be compared with For its Digital Optical Recording systems storage unit cost of videodisc systems using Philips expects that the user will have to pay "mass produced" videodisc records. between US $250 and $300 for a disc record; (5) Semiconductor technology: The orders­ it is hoped that in the second half of the 1980s of-magnitude advances that were character­ this price wi II fall below $100. The company istic of semiconductor technology have pro­ plans to have a starting price in 1982 of duced similar startling changes in the price around US $100,000 for the disc drive itself. and performance of processors. The price could be half of that in the following Figure 11 illustrates the learning curve for a

1982 , Vol . 2, No. 1 59 ELECTRONIC PUBLISHING REVIEW

Mainframe computer

$100K ~ital optical "-recording systems $10K ·1 ntelligent" """----:;;::::..___ video disc "' player $1K Programmable calculator "'Standard" ' video disc $100 ...... ______...______player 1955 '60 1965 '70 1975 '80 1985

Figure 11. Constant performance plotted against mainframe price (Figures except videodisc are taken from[J]) mainframe computer of 1955 vintage. By the stamped by the stampers - cannot be mid- l 960s, the emerging minicomputer had changed. Reading the data with a laser light of achieved the same or better performance at a lower intensity does not (and should not) alter price nearly ten times lower. By the early 70s the information stored on the record. But an intelligent terminal had reached the same videodisc technology is still in its infancy; it is level, but now the price was almost 100 times hoped that in the future it can be developed lower. In 1977 a hand-held programmable into a form in which the data can be erased calculator with a performance equal to that of and rewritten many times. Intense research in the 1955 mainframe computer was available this direction is being undertaken at different at a price almost 1000 times lower than its research centers around the world: the Japanese much larger predecessor. firm Matsushita is planning to bring such a Standard analog videodisc players, first system to the market in about two years' time. introduced in the consumer electronics In the opinion of the author, the "read only" market around the end of the seventies, are nature of videodisc is not a hindering factor, expected to slide down along the learning since large, low cost random access data curve to about US $400 by 1985 . At the same archives - even with read-only character­ time, more sophisticated intelligent digital istics - can be utilized in many essential videodisc players controlled by advanced applications. This will be shown in Section 5. microcomputers with capacities of 256-512 These applications include their use as backup KB RAM will be sold as dedicated data bank archive systems in computer centers and as computers. Such devices, as will be shown videodisc reference books (encyclopedia). later, will be able to perform as third party Multilanguage dictionaries are another computers ("data hosts") in interactive video­ excellent, long awaited "read only" tex systems, thus videodisc development may application. influence videotex technology as well . We can summarize by saying that in the light (6) The "read only" nature of the videodisc: of the broad field of applications for video­ Videodisc records as we presently know them disc, and taking into account the other aspects are non-erasable: the data, once written - or mentioned above, such as the preparation of more accurately, burned by a laser beam or mass information and production of videodisc

60 1982, Vol. 2, No. I ELECTRONIC PUBLISHING REVIEW

records, the "read only" nature of the video­ disc track cannot be used by the system any disc technology does not seem to be a hinder­ more, preallocated "reserve" tracks will be ing factor in their applicability; in fact, for introduced to replace the old track. However, archiving purposes, it has advantages. such methods are not possible with replicated (7) Data accuracy problems: Errors due to videodiscs. With the digital optical recording faults in videodisc production (e .g. in master­ system it is easier, since the information can ing and replication) cause major problems of be checked immediately after writing and if data accuracy. Failure rates are high. (As the original information cannot be restored by mentioned earlier, for videodisc records with using error control coding methods, a new TV-like programs, failure rates vary between track would immediately be allocated for the 10 to 90 percent at a Philips factory in the UK) same stream of information which now could and are quite unpredictable. Unwanted parti­ again be burned on to the disc. The mass pro­ cles (dust, dirt, etc) are unavoidable in any duced (replicated) videodisc records would system and in videodisc recording a particle allow no such possibility; the original just a few microns in diameter can prevent the information has to be restored - supported recording of many signal elements or obscure by sophisticated error control coding methods them after recording. The effects of this type - on the basis of what is on the videodisc of "dropout" are less significant for television record. recordings since redundancy in individual Nevertheless, videodisc experts are confi­ frames coupled with the persistence of vision dent that in the long run, this can be done: as reduces their effects to negligible propor­ J. W. Klimbie [30] from Philips Research tions. Nevertheless, there are serious prob­ Laboratory says, this is already applied, and lems with recording even television-like pro­ works satisfactorily, in their DOR system grams on videodisc records [26]. The situa­ where the empty disc is preformatted in a tion is even worse if a stream of coded, similar manner as their mass-produced video compressed text or data is to be recorded, long-play disc and where each DOR section since dropouts can obliterate a whole block of contains its individual address. digits and cause significant numbers of As an example of error recovery , let us errors. Methods for overcoming these prob­ consider the Philips ORA W (Direct read after lems through protection of the storage media write) [27] method used in the experimental surface and introduction of error control version of the DOR system. coding are being extensively researched and There are 45,000 tracks on each side, every introduced. Whether or not this problem is track containing 128 sectors of 1,900 bits. solved will be decisive to the success of this Each sector has a unique address and 1,024 technology. bits of protected use data, followed by a Data accuracy is not a new problem in secondary error-checking code. ·The error information storage and transmission. How­ correction bits are split up into 22 so-called ever, the "read only" or "frozen" nature of code words of 16 characters of 4 bits, which videodiscs makes a solution to the problem of are interleaved throughout the sector. It is data accuracy most difficult. In other storage claimed that a primary and backup error­ or transmission technologies, the computer encoding system corrects the error rate to less 10 system itself attempts to correct false data by than 10· • means of parity checks and other error recovery The primary encoder and interleaver intro­ methods; in cases of data transmission errors, duce structured redundancy to enable detec­ where correction is not possible, the system is tion and correction of all errors except very requested to repeat the transmission. In the long bursts. The backup error-correction case of data read from magnetic disc, repeti­ system monitors the playback quality during tion of data access is usual practice, too. If it recording. Any discrepancy between the turns out that, after a few more trials , a certain incoming data and the recovered playback

1982 , Vol. 2, No. I 61 ELECTRONIC PUBLISHING REVIEW

"'--~M Double Coding+ D1t1 source one sector interleaving buffer next \ sector I Enable I next Sector sector dltl Sector acetpt/reject eI I I

Rewrite Deinterleaving decision +decoding Demodulation ---<

Figure 12. Error detection/correction system . signal that approaches the correcting power the error-correcting capability of the system, limit o(the code causes the system to rewrite as determined from the playback signal, the the entire sector in the next block. This burst­ secondary error-checking ACCEPT code is trapping method virtually eliminates uncorrec­ written. The "enable-next-sector" signal is ted errors, even in the presence of large then given, and the data source is advanced. defects. When no errors (or correctable errors) There are other ways of detecting and are detected, an ACCEPT code is written. correcting errors, of course, without direct The design provides for subsequent indirect playback and secondary error checking, but addressing by leaving in the sector a gap that they would require a more robust coding­ may later be recorded with a posting sync interleaving scheme that would be less efficient sequence and also with the post (or indirect and would decrease the usable disc capacity . address). It is this technique which has to be applied for The error-detection correction system [23] reading "mass produced" (pressed) videodisc transfers input data from a source to a double , records. one-sector buffer (figure 12) . The complexity of the data accuracy prob­ The first sector is coded, interleaved, mod­ lem in this technology will cause many head­ ulated, and then used to drive the recording aches in the years to come. First the problem laser. Modulation matches the data spectrum has to be solved satisfactorily, and this applies to the spectral requirements of the record/read also to mass produced (pressed) digital video­ process. As writing is recorded on the disc, it disc records; then it has to be standardized in is read - through the same objective - to order to ensure that the technology will check the accuracy of the recording. Should achieve full market penetration. If too sophis­ the sector recording exceed (or equal) the ticated hardware or software methods have to error-correcting capability of the system, a be applied to the mass-production of digital secondary error-checking REJECT code is videodisc records and their play back, this written, and the "enable next-sector" signal is might have a negative effect on the very inhibited. In this way "bad" sectors are re­ favorable cost level of this storage technol­ written. Errors may occur only through disc ogy. But answers to this question can only be degradation or improper reading. expected in the coming years. If, however, the sector recording satisfies

62 1982, Vol. 2, No. I ELECTRONIC PUBLISHING REVIEW

3 .3 The "dualistic nature" of videodisc techniques: 25 frames per second, each frame As we have seen, videodisc has a "Janus­ having 625 lines (210,000 picture elements faced dualistic" nature. On the one hand, it per frame). Since in principle access to video­ fulfils technically all the requirements for a disc frames is random, the microcomputer­ consumer electronics product: it can play based videodisc controller executes an inter­ back pre-recorded video information, it can nal program for generating sequential frame present slow or fast motion, it has two channel access, moving step by step from the first stereo sound, and it can perform several other frame near the center of the disc to the last one consumer electronics functions. One can at the outside. A possible flow of control is argue about whether or not it satisfies these shown in Figure 13. needs as efficiently and cheaply as standard The flow chart in Figure 13 also includes videotape devices, but this is not the point. other classical functions of the videodisc The basic issue is that in addition to these controller, such as slow motion, fast motion, functions, videodisc can perform a number of freeze-frame, and reverse motion. qualitatively new functions . The other side of The most important videodisc functions the "Janus face" which looks in the direction (both classical ones and those related to infor­ of information technology, sees a "gap-filling" mation technology) are compiled in Table 6, technology in the field of fast random mass which describes the major operational features storage devices , with the added advantage of of a videodisc controller. extremely low costs. In particular, the tech­ nological revolution of the videodisc, backed 4 .2 The information technology-related by optical storage and microcomputer tech­ functions of videodiscs nology, provides a new basis for the storage Before discussing the information techno­ and manipulation of large quantities of textual , logy-related functions of videodisc, we audio, and visual information. The videodisc should again summarize the existing new drive' s ability to perform the above-mentioned technologies that form the basis for them. consumer electronics functions represents only a sm:all fraction of the capabilities for its a) Ultra high density storage of any kind of built-in microcomputer. A new horizon of analog or digital information videodisc applications is described in Section b) Fast random access to addressable units 5. (frames) of digital information by means of a highly focused laser beam c) Application of microcomputers to control 4. Operational Functions of Videodisc the functions of the videodisc handling Players program, which adds greatly enhanced local 4.1 "Classical" functions intelligence to the system d) A broad variety of possible hardware inter­ As mentioned earlier, the typical videodisc faces to other types of applications, tech­ player presently available on the market can niques, and services, such ·as TV, Hi-fi, randomly access over 54,000 videodisc personal computer, remote and local host tracks per disc side. The most classical function computer, telecommunication and computer of this type of videodisc player is motion networks, videotex systems, and combina­ playback of video image frames. tions thereof. Clearly, the present format of a videodisc image frame follows the standard TY-video The information technology-related functions and audio format. Hence for the USA, accord­ provided by an "ideal" videodisc controller ing to NTSC standards, 525 lines (about can be divided into two major categories: 150 ,000 picture elements, but not bits) must - fast random access and control functions be produced at a speed of 30 frames per of preferably digitalized mass information; second. In Europe, this must be done in and accordance with PAL, SEC AM , or other TV - data output channel functions.

1982 , Vol. 2, No . I 63 ELECTRONIC PUBLISHING REVIEW

Table 5. Operational features of a videodisc controller.

Operational In.ilia! Stop Frame Frame Access to I features frame frame cowiter counter I frame with Function increment decrement time [/.J [f,] [i] [i] delay 1. Normal ,, /, /1 >/, - one by one - Read con- video play tent of backwards track and display on TV screen 5. Video play /<>/. /1 >/, - more than - Read con- reverse, one tent of fut motion track and display on TV screen e. Video play I<>/, /1 >/, - one by one yea Read con- reverse, tent of slow mo- track and lion display on TV screen 7 . I Freese /1 =/, ,, =J. none none - Read con- frame tent of track and display on TV screen 8. Frame by /1

Note: f B = first frame of the videodisc f z = last frame of the videodisc where I B .a/,"" /z and Is""/, .a/c

64 1982, Vol. 2, No. 1 ELECTRONIC PUBLISHING REVIEW

Operational I Initial Stop Frame Frame Access to features frame frame counter counter frame with Function increment decrement time [/.J [! .] [\] [\] delay 9. Frame by ,, >f. ,, >t. - one by one yes Read con- frame, re- (manually) tent of verse track and di.splay an TV screen 10. Individual ,, .. ,. ,, .. ,. one by one - - Search for frame ac- individual ce:ss frame 11. Frame --- - - Displays number frame display number 12. Selectable -- - - - Two chan- audio nels of hi&h tideli- ty audio can be stared in addition ta videoframes, enablina either streophon- ic sound, duel lane uag e. or question- answer format 13. Proeram------In the min& model "PR-7820," 1024 bytes of nan- volatil memory pro- eremmed either mannually (cambina- tion of previous instruc- tians) or autamati- caly by pre- programmed videodiscs

1982, Vol. 2, No. 1 65 ELECTRONIC PUBLISHING REVIEW

START

-s.ttlllt pnlCldun• for 'fideo disc. Call lit dolll 111 diffwlftt W1YC

I)b) --by pr.-nal'lll In 1111 llll1llOrY of vidlo disc COll1rOll.- llld d--loldtd, 1.g. fnlm a 1iftn video disc framl or fnlm an in1mlal Comrollint sptld Pf1llfllli lilmry, sucll 11 •d pl1yi111 dinctio11 program for tilt "nol'lllll by Sltti119 of motion pll'fblck" funaian ,ncrtment Vllue (I)" cl by muns of a 11111 bml lllllllllftllll syftallll illlllit- if (1)>0 play 11111111d on 1llt video disc forMnl comroller. 1.g. by 11111' nqllllt (I)< 0 play '"Show II scons hudld in slow lll01ion by the Hunpriln t11111 (I)> 1 at the World 5oCC9I' ~ •ill 111 1916 in Entllncl" fnlm 0<(1)<1 video dill: m:on1s s11ow1no 111 .,_icons of ..... lllljar illtlnmional (1)•0

Seekhlt STARTtm..

Rlld COlltlllt of fr11111 (nl

II• r Put COlltlnt Sannlnt a. proclucint of frlllll (11) fmnl SICllllllCI till r•r+i into lllllllOrY STOPfrmM forTVim191

l"fl'lllll :-:.. ya counttr(n)"1~\o R11111rlcs: (n, s) an '1""91(' vllu11 '----<-!'"STOP fmRI STOP (r, ii an "1'111" VllUll counttr(sl"I ;>(i)

Figure 13. A simplified flow chart of the "videodisc frame display" controlling program .

66 1982, Vol. 2, No. 1 ELECTRONIC PUBLISHING REVIEW

It has to be mentioned that such an "ideal" an encyclopedia. The language used would, videodisc system is not on the market yet , but in a simplified fashion , be similar to the data­ present trends show that this will be the very base languages now used in computer infor­ likely way . mation retrieval systems. c) Downloading of access control program 4.2 .1 Fast random-access control of digital­ from videodisc: Each frame of the videodisc ized mass information. As mentioned earlier, can contain different types of information. It all information on the tracks of the videodisc can contain output data (textual, video, device can be addressed separately. If coded audio), or data required for specific videodisc textual information be stored, an additional control functions , such as index information subdivision of the entities of information (thesaurus) for the videodisc database man­ might be carried out by defining a number of agement program or executable programs for information blocks and recording each on a the videodisc controller itself. These would separate information track as done by Philips be fed into the core memory of the controller in their DOR. The random access process on for immediate execution. videodisc is very similar to the well-known d) Downloading of access control program random access method on moving head from external data sources: This category is magnetic discs. The mode of access control identical in nature to the category mentioned can be one of the following: in paragraph c). The only difference is that the - by direct manual ("physical") addressing access-control program is fed into the video­ by the operator of the videodisc disc controller from outside data sources - by indirect addressing by the operator through an external interface. In this fashion, through an appropriate videodisc database access control information may come from management program stored by the videodisc local or remote external computers or from controller videotex and other computer network nodes, - by downloading an access-control program etc. from predefined tracks of the videodisc into The modes of access listed above are the videodisc controller summarized in Figure 14. It should be noted - by downloading an access-control program that the access control functions can be linked from external data sources (e.g. interactive in many ways, making the access system videotex channels) through hardware inter­ extremely flexible. For example, by direct faces . manual access, an access program can be downloaded to perform a complicated chain a) Manual direct access:An operator know­ of access functions to other videodisc frames. ing the physical address of the required frame initiates access to the videodisc by typing in 4 .2 .2 Data-output channels and their func­ on the operator panel/keyboard all the infor­ tions. Any of the following types of.data can mation necessary for access. He may access be stored on videodisc: individual frames, which might be informa­ tion frames or videodisc control programs to a) frames with video information be downloaded, or he may initiate the output b) frames with audio information (sound) of frame sequences to be played back, etc. c) coded textual information (e.g., videotex b) Indirect addressing through a videodisc frames) database management program: The user of d) coded data information for data processing the videodisc player may interact with the programs system through an internal database manage­ e) data processing source and object code ment system tailored to the functions of the (excluding the videodisc control program videodisc controller and drive. Such a system codes mentioned in the previous chapter). could be used to retrieve information recorded Since all information is stored in the same on videodisc records from a databank, such as physical (preferably digital) form, a separate

1982 , Vol. 2, No . I 67 ELECTRONIC PUBLISHING REVIEW

"Downloading" of access program through external interface (d)

"Downloading of access program from video disc (c)

---J....-...... , Frame ROM for storage Video disc of "classical" controller video disc access programs I Direct access Video disc by DBMS drive Direct (b) Index access Data base manually information management --~(b) (a) program (DBMS) Information retrieval in DBMS language .. Classical .. (b) video disc operation manually (e)

Figure·l4. Access-control paths for videodisc drives: (e) represents the "classical" videodisc player functions, such as "play forward" , as discussed in Chapter 4. I. identifier must be placed at the beginning of b) channel to a hi-fi amplifier for reproduc­ each videodisc frame stating the type of tion of audio information information it contains. c) channel to a textual data (e .g., videotex) The output channels for information can decoder for decoding stored textual frame also differ; their selection depends mainly information on the type of information to be channeled d) channel to an external interface for local from the videodisc and on the category of (e .g., printer, local computer) and remote application. (e.g., to the videotex network, remote termi­ The data-output channels for videodiscs nals) connections for transmission of coded are the following: textual information, data for computer a) channel to an attached television set for programs, and computer program source and displaying video frames object codes .

68 1982, Vol . 2, No. I ELECTRONIC PUBLISHING REVIEW

Textual data (like videotex) decoder (c)

Video decoder TV (a) set

Video disc drive Audio HIFI decoder equipment (b) Video disc controller

External Local DP interface equipment (d) Internal core memory of video disc Remote player DP equipment (e) and networks (e.g. videotex)

Figure IS. Dara-output channel options of videodisc devices. e) channel to the built-in core memory of the that the list is by no means complete. It i~ videodisc player (e.g., information of a difficult at present to predict all of the suitable videodisc record thesaurus) for internal types of videodisc usage, just as it would have usage. been hardly possible in Gutenberg's time These data-output channels are summarized some five hundred years ago to classify all of in Figure 15 . the applications for printed information. The major classes of videodisc application are: 5. Videodisc Applications - Videodisc as an entertainment medium In this section an attempt is made to describe (5 . 1) some possible videodisc applications as they - Videodisc as an information and reference are presently seen. It should be pointed out medium (5 .2)

1982, Vol. 2, No. I 69 ELECTRONIC PUBLISHING REVIEW

- Videodisc as a tool for education (5.3) tape recorders cost less than US $750 and four - Videodisc for mass storage and for data hours of continuous program can be taped on and program archives (5.4) a single cassette at, or for less than US $5. 7 - Videodisc in office automation and docu­ per hour. In the entertainment category, ment filing (5.5) . videodisc players now cost about US $8~ 900 and the discs cost around US $25-30 for a 5.1 Videodisc as an entertainment medium full length film . Thus, at present, videotape 5 .1 .1 The "Classical" type of entertainment. technology seems not only to be technically The primary purpose for developing video­ more advanced than videodisc, but it is also discs was to create a machine for reproducing cheaper. motion pictures, on the analogy: "videodisc The next problem is the availability of should be in the video field what the gramo­ programs. For this purpose new recording phone is in the audio field". From the indus­ lines would have to be installed at high costs , try's point of view, these disc techniques have whereas recording of videotapes, be it over a great advantage over tape recording techno­ TV or through a separate camera, is flexible, logies (both video and audio) in that the copy­ cheap, and simple. In addition, centralized right law cannot be evaded. Videodisc and production of videotape programs has already gramophone are "read only" devices: a repro­ been developed. duction of the stored program on to other Thus this classical application for the records can only be carried out through a videodisc technology does not seem to justify complicated manufacturing process. Thus its viability. those interested in protecting copyright hope that a wide distribution of videodisc players will promote the sale of stored video programs, 5.1 .2 New types ofentertainment . The variety such as films, in a fully controlled way. of new types of videodisc entertainment is At present, however, videodiscs have endless. Videogames stored on videodisc certain disadvantages in comparison with frames will add new flavor to the current "vi­ videotapes, which may hinder the acceptance deogame bonanza". For example, a pro­ of videodisc technology by the public, should grammed labyrinth game stored on videodisc it be used solely for entertainment purposes. record frames may become a competitor to First of all there are technical problems. the labyrinth game installed on interactive Because videodisc technology is still in its videotex services [ 14] . In this game each dis­ infancy, there are still technical difficulties played picture frame could offer alternative due to the limited lifetime of the disc's read paths leading to other tracks which would be laser head (about 1000 hours), which, more­ selected manually by the player. On some of over, can presently be produced only at the tracks a short frame-access program might relatively high costs. Also, the microscopic be stored. Upon selection of such frame, a sized pits, which are the actual carriers of program would be downloaded into the digital information, place fearsome demands videodisc controller and a movie scene would on the servo system, which must keep the be suddenly triggered off (e.g. the "Wumpus" laser beam in focus and in alignment with or a monster who "lives" in the labyrinth spiral turns, despite inevitable eccentricity would appear to frighten the player and the and warpage of the disc [I]. As Barry Fox [8] game would be lost). points out, with so little margin for error, Another kind of videodisc entertainment making the discs becomes a nightmare. In one might utilize the programming capability of system now on the market the laser beam the videodisc controller. For example, must accurately follow a spiral of pits I/30th specially prepared videodisc frames might be the width of a human hair apart that are programmed to combine a given color, travelling at over 60 kilometers an hour! graphics, and one of the constant tones of the The next hindering factor is cost. Video musical scale to produce appealing melodies .

70 1982, Vol . 2, No . 1 ELECTRONIC PUBLISHING REVIEW

A similar application in the field of music player could directly compare his ''playing would be the establishment of a database on art" with that of the style of the artist whose videodisc containing musical notations of performance is stored on the first sound music written by famous composers. The channel. A similar application might be par­ Common Musical Notation (CMN), or staff ticularly useful for singers. notation, as it is also called, is one of the few Interesting opportunities are offered for graphical notations that is truly accepted analog videodisc with "color TV-like" short internationally. It is well-known that in recent motion picture programs and " hidden" alpha­ years, the cost of printing music has risen numerical lines in teletext mode. In such man­ dramatically; one is now likely to pay well ner, for example, "hidden" alphanumerical over $I per sheet for printed sheet music. lines in teletext mode. In such manner, for According to Cheek and Simpson (15], the example, "quiz" programs can be stored on storage of music on interactive videotex the videodisc and the solution to the questions systems and its display on home TV screens would be given on the appropriate teletext might become an interesting application in lines belonging to the motion picture session the near future. containing the given question. In such a way, The storage of music notation on video­ an analog videodisc equipped with standard discs - in its more advanced applications it teletext decoder would be able to show the might be accompanied by a sound recording replay on the TV screen on pressing the tele­ of the music on hi-fi channels - might itself text button on the remote controller keypad. become an exciting entertainment and educa­ This type of application could also be most tional application. interesting in educational applications. In its simplest form, individual music nota­ The potential for gaming with videodisc is tion frames would be organized by a music endless: if the reader regards the above forms notation data bank and stored on separate of entertainment as rather simple minded, he videodisc tracks. Using the retrieval program should accept the author's apologies, but the of the videodisc controller, compositions may aim of this passage is simply to point out the then be retrieved and displayed on the screen, vast gaming potential of this technology. or hard copies may be taken, if an appropriate interface and output device be connected to the videodisc player through its external inter­ 5 .2 Videodisc applications in the field of face . Another form of musical notation appli­ information storage and retrieval cation would be to play the music stored on 5 .2. I Videodisc as an information medium. hi -fi equipment while showing the corre­ The use of videodisc for information storage sponding musical notation frames simultane­ is an important type of application in view of ously on the screen. In this fashion, up to one the impressively large amounts of informa­ hour of music and notation can be stored on tion that can be stored and randomly accessed, each record. An excellent educational tool and this at extremely low cost. and means of performing home exercises The total storage capacity of videodisc would be to store the recorded music on the records varies considerably - as referred to two separate audio channels (which as we in the previous Section, on one Discovision know, is indeed possible) and to have the user videodisc record nearly 30 billion, on a Phil­ switch off that channel playing the music ips DOR disc, 10 billion bits of information from the instrument he wants to play himself. are stored. According to Predicast Inc . [28], a In this case, the other audio channel would high-density/laser-encoded disc storage accompany the player while the correspond­ system patented by RCA stores 100 billion ing frames with the musical notation would be bits of information on a 30 cm disc. displayed on his home TV. By recording his Before going into details special attention own "performance" with a tape recorder, the should be given here again to Table 4 , where

1982, Vol. 2, No. I 71 ELECTRONIC PUBLISHING REVIEW

the storage capac1t1es of different types of Examples of storage capacities for typical - present and future videodisc systems are not mixed - applications are shown in Table listed according to different "media" categories 7. such as audio, video, freeze-frame, coded Looking at Table 7, we find that an enor­ alphanumerical, "atomic" information unit, mous amount of information can be stored on etc. The information for most of the applica­ a single videodisc record. If the capacity to tions that will follow could be put on different store a 60-minute movie (first entry) does not types of videodiscs - first of all, in the form seem particularly outstanding, one should of different types of "media". This is primarily reflect that all of the major paintings of the up to the author and the electronic publisher. world, the contents of a large personal library, For example, in the design of a videodisc or all of the references in one of the biggest encyclopedia one can go several ways. One bibliographical databases such as NTIS, alternative is to follow the pattern of the clas­ Inspec, or Predicast can easily be stored on a sical printed encyclopedia. In this case main­ single videodisc record. ly alphanumerical and standing picture types of information should be stored-hence a dig­ 5 .2 .2 Use of videodisc for old and new types ital videodisc seems to be optimal. The se­ of 'reference books'. As already pointed out cond alternative could be to provide as many in previous chapters, the use of videodisc as motion picture scenes complemented by au­ 'reference books' is one of the optimal types dio information as possible; in this case, an of usage for this medium. Similar to the analog videodisc such as that of Discovision extremely numerous categories of possible Association should be applied. A third way categories of possible classical references might be the so-called "talking" encyclopedia books, their videodisc equivalents represent a as suggested by Barrett [23); in this case, the large family as well. A few uses for which the majority of the information should be audio, videodisc would be ideal are mentioned here. complemented by freeze-frame and motion pictures. Here the "future analog or digital (a) Videodisc as 'encyclopedia'. The Encyc­ videodisc system" should be used. lopedia Britanica has 23 volumes of text The exact videodisc storage needs and containing approximately 20,000 figures and technical requirements thus very much a separate volume containing indexes and a depend on the outlay of the encyclopedia and geographical atlas. A single page of full text will be determined when the detailed design might contain a maximum of 9,600 text of such system is performed. Therefore, in characters. what follows all the figures given in this paper If we use a digital videodisc with storage with regard to applications should be looked characteristics according to the new DOR at from that angle. In actual design, figures disc (2 GBytes/disc) , assuming that there are can differ considerably; here, with a few 23 volumes of full text and 20,000 black and examples, only some rough major character­ white pictures of standard resolution, with istics should be outlined to provide the right data compression a videodisc capacity of 7 .2 feeling to the readers about the orders of x 108 Bytes would be needed, which repre­ magnitudes involved in different applica­ sents approximately 36% of a single DOR tions. capacity. If an analog videodisc such as DVA's In our first example we have taken some 7820 model 3 was used, the 23 volumes of typical figures for different types of storage full text in TV mode and the 20,000 pictures "media" which can be stored on one of the of color TV quality (according l track per pic­ present or future videodisc players. Since the ture) could be put on 2- 3 videodisc records . types of stored information can be mixed on Greenagel [25) describes the first trial any videodisc record, the total amount of along this line , a Channel 2000 videotex actual information varies according to the experiment currently taking place in Colum­ type and amount of stored information entities. bus , Ohio in the USA, in which a new general

72 1982, Vol . 2, No. l ELECTRONIC PUBLISHING REVIEW

Table 7. Examples of data stored on videodisc records

Type of information Examples for total Example storage amount of information entities

Moving picture program Scenes in total of a 60- Disc containing all scores of the national (video and audio minute program at football championship of last year storage) normal speed

Standing pictures 108,000picture All pages of a larger encyclopedia of twenty -TV quality odd volumes on two records -Facsimile III quality 25 ,000 pictures All famous etchings of the world

Coded information on Over 3 million frames 300,000 encyclopedia pages of information videotex-type frames information (The Encyclopedia Brittanica has "only" (960 characters per frame about 24,000) using the CCITT/ISO standard 7-bit code 15 bilingual dictionaries with each approxi­ mately 270 ,000 entries (words, expressions)

Over 3 million references inclusive abstracts of database information (the largest data­ base "Chemical Abstracts" has over 4 million references)

Coded data (detailed Over 3 billion bytes E.g., over 750 million figures (Fortran Real discussion in 5 .4) 4 bytes) for archive of statistical data

Computer programs Over 3 billion bytes E.g., 100 million lines of source code (a -source code larger program system has 10,000 lines of --object code source program) (detailed discussion in 5.4) E.g., storage of 100,000 object program of 30 Kbyte size

encyclopedia for secondary school and film clip of the hydrofoil rising up and university students, the Academic American skimming across the water, check the Encyclopedia, is being recorded on videodisc bibliography for further-references and records. Just the text of the encyclopedia finally, cut quickly to the article on "air would fit on two sides of a videodisc, but it cushioned vehicles" to compare the will occupy three or four discs because of the advantages and physics of these two sound and motion sequences illustrating the new water-crossing technologies." material. It is planned to make the entire encyclo­ "With a videodisc [the DY A 7820 model 3 - pedia available via two-way interactive cable N. B., an analog videodisc model - TV in six American cities. It is estimated that was used because of the more sophisti­ the cost of this videodisc encyclopedia to the cated progamming facilities] one is customers could be less than half of the cost of able to call up the article on "hydro­ the printed version. foil," read the text, see a full-color cut­ Barrett's "talking Encyclopedia Britanica" away illustration, switch to a 29-second [23] is estimated to have a volume of I 0

1982, Vol. 2, No. 1 73 ELECTRONIC PUBLISHING REVIEW

videodiscs which would additionally have a eventually they will find their ways into most capacity of 270,000 color television frames households. and it would take 750 hours to listen. As explained previously, there are two (b) Videodisc as an "automated dictionary". "extreme directions" in which a product can An automated dictionary is a computer-based evolve: either its cost will be reduced while device - in our case realized on a videodisc performance is kept constant, or the cost kept record player - that holds all the information constant at increased performance (see Figure of a dictionary and allows dictionary entries 9) . Of course, a mixed strategy between the to be accessed and displayed. The dictionary two extremes is also possible. Falling prices might contain entries of a single language, for the same performance would mean that such as a standard dictionary in English, the present price of the Encyclopedia German, or Hungarian, or it could be bilingual Britanica could drop by a factor of around (e.g., English-German, German-English). In 5~100, provided that the drop in price would special cases, for example for a "dictionary of result in the mass purchase of encyclopedias, computer techniques", it might be multi­ which would allow coverage of the extremely lingual. There are many advantages in the high intellectual costs of collecting and use of an automated dictionary instead of a updating its contents. dictionary in book form: The other "extreme direction" would be to - the user-interface will provide a simpler keep the cost the same (e.g., AS 25,000 in means of accessing dictionary entries Austria for a complete encyclopedia) and - the device will provide a set of search increase its performance. As suggested capabilities enhancing the usability of the earlier, this could be done by including dictionary "motion picture" encyclopedia entries - it will allow sophisticated new applica­ whenever the information to be provided to tions, such as spelling correction, synonym the user is less of a descriptive and more of a and antonym search, voiced word pronuncia­ visual or audio nature. If this new type of tion, games such as picture naming, and in encyclopedia were made up of about twenty more advanced systems, perhaps grammar records - similar to the number of volumes correction and sentence building capabilities. in printed encyclopedias - it might contain A detailed description of a similar system is about 2500 half-minute or one-minute provided by Fox, Bebe! and Parker [8]. While moving picture or audio screens. Such a one can fully agree with their definition of the "mixed media" encyclopedia could still be functions of the automated directory, their sold at about half the price of the present assessments concerning physical realization printed version. An encyclopedia with voice of such systems do not take into account the and motion in addition to text would be a potential storage and retrieval capacities qualitatively improved new medium, in the offered by videodisc technology, which is author's view. perhaps the only cheap mass storage techno­ To keep the price level constant, especially logy presently known that could provide a during the introductory period, it can be sound basis for this type of application. The expected that dedicated videodisc players will limit of the system they describe would be be offered with the records. Later, however, around 30,000 entries stored on approximately when the videodisc technology (hardware, . 120,000 frames . However, using videodisc software, stored data) is standardized, the storage technology, up to 3,000,000 frames encyclopedia records could be played on any can be stored per videodisc record: enough videodisc player. The first videodisc based space for the largest of the presently known encyclopedias will appear in libraries and on printed dictionaries. So even the complete computer networks - this includes videotex Oxford English Dictionary, with its 12 volumes networks as third party data centers - but and three supplementary volumes (price

74 1982 . Vol. 2. No. I ELECTRONIC PUBLISHING REVIEW

Table 8. A classical language dictionary entry and how it is used in an automated dictionary entry according to [8]

The dictionary entry. The primary purpose of the automated dictionary is to store and allow access to a dictionary database. Consider the definition of the word "flag" taken from Webster's New Twentieth Century Dictionary of the English Language:

flag':(flag) n. [LME.flagge

3flag2: (flag) n. [MEflagge< ON. jlaga , slab of stone< IE. base plak-, to spread out, flat , whence L. plasidus, flat] same as FLAGSTONE -vi. flagged, flag'ging to pave with flagstones

We list two separate entries for the word "flag," although the dictionary actually contains four separate entries. The rationale for separate entries is that each entry represents a conceptually different definition of the word. Notice that each entry contains multiple definitions which are not necessarily related.

Each definition also contains a variety of information, including correct spelling, syntactic category, pronunciation (written), etymology, synonyms, and syntactic variations. Information such as antonyms, examples of usage, hyponyms, and pictures can also be found.

approximately US $570) and approximately tions", such as voiced word pronunciation 300 million characters, would fit on to a (another revolutionary step forward!), spell­ single videodisc record. A typical dictionary ing corrections, games, etc. entry is shown in Table 8; a list of the most It is expected that automated dictionaries important automated dictionary functions are will be offered on the market relatively soon. presented in Table 9. Figure 16 shows the The first version will probably be accompanied same dictionary entry given in Table 8 imple­ by dedicated videodisc record player devices. mented on the experimental - not based on The first devices will be placed either in videodisc - Zog System of the Carnegie libraries and educational institutions or Mellon University team [8]. on computer networks including videotex Figure 9 shows the two "extreme direc­ networks. Thus , in the first phase, the prin­ tions" in which the automated dictionary can ciple of resource sharing would apply. Later, evolve. Following a "learning curve", the when videodisc prices have come down, the simplest versions of automatic directories standardization problem has been settled, and stored on videodisc records will become videodiscs have found their ways into homes, cheaper and cheaper;.it can be expected that videodisc records containing automated they will be sold at average videodisc record dictionaries could be found in the storage prices of approximately US $25-50 per cabinet of every videodisc device owner. record, which is considerably cheaper than an equivalent printed version (for example, by a factor between 10 and 25 in the case of the full ( c) Videodisc as an online information bank. Oxford English Dictionary). If one follows Videodisc technology is likely to have a the ''performance curve" however, more major impact on online information retrieval. sophisfr::ated automated dictionaries will be At present, online information services such available offering in addition to the standard as Dialog, SDC, or ESA/IRS offer huge dictionary functions "value added applica- amounts of information over private and

1982, Vol. 2, No. I 75 ELECTRONIC PUBLISHING REVIEW

Table 9. The most important functions of automated dictionaries (Source: Fox et al. {8])

Word retrieval: The user types in a word, and the corresponding dictionary information is displayed.

Spelling test: The model presents to the user, by voice output, a word at a level of difficulty chosen by the user. The word is then typed in by the user. If spelled correctly, a positive response is given to the user. Functions that gather statistics on user performance could also be provided.

Word recognition: A word is randomly chosen and permuted according to standard spelling errors. The user must confirm or deny its legality as a word.

Anagrams: A list of letters is provided to the user, who then tries to compose as many legal words as possible from them.

Word search: A definition is displayed, and the user must provide the corresponding word.

Picture naming: A picture is displayed, and the user must name it or parts of it.

Synonym search: By following synonym paths, the user must determine whether two words are related.

Voiced word A displayed word will be accompanied by voiced word pronunciation. pronunciation:

Grammaroorrection, This capability would be dependent upon more sophisticated processor and sentence building: software becoming available.

public computer networks to some ten thousand data center configuration is considerably users. There are three major characteristics of lower - about one sixth the cost of an equi­ the present technology: valent traditional database center [9] (see Table 10). Comparing total operating costs, ( l) frequent centralized update (daily, weekly, the videodisc-based large online information monthly, quarterly) of the databases; bank center is cheaper than an equivalent (2) the principle of online ordering and offline center based on magnetic disc storage tech­ delivery of primary documents and nology by a factor of approximately three. A (3) the aspect of cost and resource sharing by possible configuration of a videodisc­ a large user population. The videodisc tech­ technology based center of the size of an nology will have major impacts on the pres­ ESA/IRS (about 30 online databases, 18 ently known, online databank technology million references) is shown in Figure 17. (based on large central mainframes and The fifty videodisc units are grouped accord­ special database management programs - ing to the usage of the individual databases. suitable for fast retrieval from large stores of More frequently used databases, such as In­ information on large 'magnetic disc devices). spec, Biosis, or Chemic.al Abstracts, would First, the amount of stored information can be be operated separately on two to three parallel multiplied. For example, the 17 million on­ disc installations separately, allowing parallel line references of a present large database searches by different users. The configuration service center can easily be stored on twelve shown in Figure 17 could serve at least 50 parallel working videodisc devices controlled users simultaneously. About 50 percent of the by one central processor ("head computer"). total online disc capacity of 75 billion bytes The installation cost of the videodisc-based (an approximately 75 million references)

76 1982, Vol . 2. No . I ELECTRONIC PUBLISHING REVIEW

- 51 -

f119 lln) Didi llagl\n) Ck!l PRO JrCUNCl.6.T !ON nag OEFINtTIQN· A IMC• al cJoftl II' tll.lnhtlQ . °"'" lttkl'ltG to. ETYMOLOGY · LME _ F~ o1.tosiQ11a1 . 11t.

DEFINIT IO NS 1 AplfCIOICIO!ltortlUllli119 , 0fltnlftKll«llOIStl1fl. witncllSl1narv9t:QiorS. p.anerns . OI symCOlc c.wictS. SYNONYMS · ANTONYMS · HYPONYMS used U I nl!IOl\ll 0t ll•1 symbol . lo &igNI. £ . 1. Batl,.,. z. [,.,.Rlr•llonQIUIMBorQuilllUOftaf\IWk, z. Ensign J . Thet.iolldttf. l 5'"'41rd. 4. Thi DuSlly tail oe CllUin dogt, u Mim w somt USAGE EXAMJl\.ES· llounes . • o. Wor10ll11111t1011t . ,...... , ______?.WorllOirtillibotll . < .PfrfiovsWotd. l .lnde. NciilWord t .lnda____. > .Mtx!Word _. > < .PrrmutWord. ------... ----.---

bln.,..(n.I Dlct9 11.11\MFWll(n.) O!cl17 PflONUNCIATION: ~ - - PRONUNCIATION : ban"• it . 1 r•· ETYMOlOGY : ME. ban«e _ H111 ....,_Word t ino ... > Not Word .

tel

Figure 16 (Part I). Example of automated dictionary entry .

- 52 -

B. lndu """' AVAILABLE FUNCTIONS ""'' A. Dlnner l. blnquert1 o. 1W1111 Magrarnpne. 9. !Wlnerrt J_ Banouo A. Bint11a111 C llanntfl>I JC IW'ltl'lff S. !wlyan SoeWno ttst. D Dainn1Sle r l . llanllm T. bll'll... E. tiannock M.IW'lwnwttQht u. NoOltl J. 'M:Wd.-.cogmtton. F Blnnockbvrn N!Wittt V . ~ . G. !Wins o. Ban 11no W. b.IOll W 4. Word 58'dl. H. llanouet P. b.lntling I . baptism

< Jlrw¥1011S8ll\du f ll\Ot.J. . > .NulBWonl. t.Main lnOO: ----·-- ·---.-..-- ...... ------.. -....-- ... -- ,.. (Q

!.ch diet..., Miry .. ~ .., • pMa8ry tr.me 48) tNI ,,....._ pronunc.l.9t'°" Md atymcMogy u t•al Md cteHntttoM .. GIP"°" .. Se'9ctl"I • Mao~ PT"· Clct21 dllllnltion option cau... tnnefef ao • tec:ondary "8me(b) THE ANAGRAM GAME COll'l!Mnfng synonym .. .,.,.°"""' hyponym.. .wt uuge An ~ram is a s.quenc• of leners tflll an tll com110S1d to eumpte1 th.II go wtth tNt dlillnttlon. Tl'lt. rr.me • rormo111orl'l'IClt1WOfGs. Foteumf111 tl'l1 l11tll'Sin ··1111 ·· i;an fNChed by ..i.cuog the optkN'l IHt Oft tM ~ ot by tlllHITlllQld 1o lonn11'11WOfdS ·· ~ ,. -- 11\d "'rile. " Youi;annot hhUng the .. , .. on IM keyboeni. n. f,.me '°' tM fsnaYIOl'idOifftll"Sta thoSlgNllllO'fOU. SJftOftY"' ""bannV' (~Is ,.llCMd by ae&ecting Hem-.- m lrame (b). Two opUons-NDT WORD and ll'REVIOUS Durit1Qlfl1pqyitw;laf fh1pnayoucanacusstMdid10rWYn WORD-1llow Hquentl•I mow....m throuih •Ofd• in l,llU. To git back to ttns QM. USI 11'11 rlhlm p.ad . the dlettonary. FDf eaampa., tM fraMe few '"banneNI.. (d) .. ruch9d by 1elacllng IM NEXT WORO opt&on .:rhU• In U. "'benn.,.. !tame (4 Tl'le INDEX option Inn.ten to• G. Guas Wonl. tnme wMte Ute ~word 19 lndeaM (e). The lmat HIMI F. AMII Gaml. Oft tM Plilil• contains global COfNIWMls-1pedal options ··-"""°· tor l'Mnlpulatlng tM ...twon cM! framn.. To plaJ oam--. the FUNC option ts Ni.cted. Md the FUNC tnm. (f) dl1p6ep th• l"ftlilabtll tvnr:Uoft• In,,..,. ,..,., SeMettng the ..eNgram pme"' ..,.... trMI tftle ,,.,. cel'9 • the •• In.pan! fnl'M CCI- Figure 16 (Part II). Example of automated dictionarv entry.

1982, Vol . 2, No. I 77 ELECTRONIC PUBLISHING REVIEW

Table 10. Comparison of service costs of "classical" and videodisc-based data centers

Yearly cost of on-line service centers Based on a "classical" Based on a distributed data base configuration (US$) videodisc-based configuration* (US$)

Hardware 430,000 90,000 Software 145,000 10,000 Office expenses 25,000 20,000 Operating staff salaries 100,000 100,000 Total 700,000 220,000

*calculated on 50(!) parallel working videodisc (2-3 years lifetime) with a total on-line capacity of75 billion(!) bytes attached to a "head" minicomputer would be used to store all online references at information bank centers. Royalties and tele­ least twice for simultaneous access during communication charges would remain the multiple usage, and another 50% would be same. This will probably lead to the establish­ used to store the latest online primary docu­ ment of local databank centers based on ments. In this way, about five million one­ videodisc technology on the premises of page online documents A4 size can be stored, major users, e.g. libraries, which might nega­ the equivalent of a library shelf 300m in tively influence the revenue growth of the length or 500,000 papers with an average of centralized databank centers. In the long run 10 pages each. This amount of online primary it can be expected that database suppliers will information is approximately equal to the start to supply their subscribers with the most total of referenced primary documents over up-dated videodisc versions of their databases the last 12 months. on a regular basis. Thus there will be separate Online document delivery of primary videodisc records for Biosis, NTIS or Chemabs information might become an attractive databases. In most cases, each of the present service for centralized information data bank databases could be put on to a single video­ centers; however, to this end, the data trans­ disc record. Only the largest databases such mission capabilities of the telecommunication as Chemabs or Pascal would need more networks has to be improved considerably space. Eventually - once standardization is and the cost of telecommunication has to achieved - all these records could be put on come down. Perhaps the introduction of any more sophisticated videodisc player. advanced computer networks and new hard­ In the field of information retrieval, the ware technologies (such as fiber optics net­ impact of videodisc on videotex systems will works and telecommunication satellites) will be twofold. First, it will be beneficial from remedy the present situation. the viewpoint of information suppliers and The cost of telecommunication is still very videotex data center operators since it will high: as a rule of thumb for online retrieval allow the storage of a large number of infor­ outside North America, one can say that one­ mation frames at considerably lower costs. A third of user costs is spent on telecommuni­ configuration similar to the one shown in cations, one third on databank center usage Figure 17 can be imagined for videotex data charges, and one third goes through the data centers as well. However, if the videodisc spinner (operators of databank centers) in the technology finds its way into every house­ form of database royalties to the database hold, there will be limited need for special suppliers. Only the operating costs of the types of centralized information to be stored center could be reduced through the intro­ at videotex data centers; for example, a data­ duction of videodisc technology to online base for a railway timetable or even an elec-

78 1982, Vol. 2, No. I ELECTRONIC PUBLISHING REVIEW

Dial up lines, connections to computer networks 7incl. (videotex network) Telecommu· nication interfaces

Head

Program Interfaces disc Printers to the video disc recorder (controller)

Video disc controller 2

Figure 17. Example of an on line data service center based on videodisc storage devices (*such as the Philips DOR system). tronic directory could be stored on home 1971 to date) as mounted on Bibliographic videodiscs. Only information requiring fre­ Retrieval Services (BRS) in the US with the quent up-date (e.g. daily) and the "umbrella" facility for two-second retrieval of patent type of information, when many information drawings using a Discovision videodisc providers put their information under a player. The 700,000 drawings are stored on common "umbrella" (as commonly used in 14 discs (each disc side providing 26,000 the interactive videotex type systems such as frames for drawings). The microcomputer Prestel) are unsuitable for videodisc applica­ terminal/videoplayer link permits automatic tion. These categories are better suited for retrieval of selected drawings; discs will be videotex type of services. updated quarterly by Pergamon. There are Last, but not least, it should be mentioned plans to use two parallel screens to allow text in this chapter that a new database for patent and drawing to be viewed simultaneously. drawings based on videodisc has been The price of the system will be US $6,000 per demonstrated by Pergamon International year (including terminal, videodisc player, Information Corporation. The new service and videodisc subscription). called Video Patsearch [ 13] offers micro­ The novelty in the Video Patsearch system computer-microcomputer access to the lies in its capacity to retrieve pictures. Similar Pergamon Patsearch file (U.S. patents from applications in the past were mainly based on

1982, Vol. 2, No. J 79 ELECTRONIC PUBLISHING REVIEW

Table 11. Number of videodisc records needed to store "electronic directories" for selected countries (source [11/)

Telephone population Number of videodisc records Country (January 1980) containing "electronic directory" (million) through videotex stand alone

Austria 2.812 2 1 Bulgaria* 0.946 1 1 Czechoslovakia 3.073 3 2 Finland 2.244 2 1 FRG 26.632 18 9 GDR 3.071 3 2 Hungary 1.186 1 1 Italy 10.085 12 7 Netherlands 6.853 5 3 Sweden 6.407 5 3 UK 26.651 18 9 USA 175.505 117 59 USSR 22.464 15 8 Japan 55.422 37 19 Canada 15.560 11 6

*data in 1977. micro-fiche; videodisc will probably bring a million frames of information. If data on all breakthrough in this field . telephone subscribers were stored on one page each, then an online videodisc record (d) Videodisc as an "electronic directory". could contain information on about 1.5 The first Electronic Directory System is million subscribers, a large telephone user presently being developed for 250,000 tele­ population. If the videodisc units were stand­ phone subscribers in the Ille and Vilaine alone devices and each subscriber had his region of France. It will be implemented on a own videodisc player at home, then twice this videotex-type computer network using amount of information storage would be at his simple CRT terminals [IO] . It is hoped that by disposal, since the other side of the record replacing the paper telephone directories could be used as well. Table 11 gives a rough issued yearly, that not only will more up-to­ overview of the number of videodisc records date information be provided, but significant needed in selected countries according to savings in resources (paper, distribution, etc) 1980 AT&T statistics [21] on their telephone can be achieved. populations. The videodisc technology might have a The number of videodisc players installed two-fold impact on the "electronic directory". on videotex-type networks to be in the online First, if one keeps to videotex-system based version of "electronic directory" would on the central "electronic directory" philo­ depend on the telephone subscriber popula­ sophy, then the storage and retrieval function tion of the country. With the appearance of of the central "electronic directory" center - the home videodisc version of the "electronic which is basically an online information bank directory", every owner of a videodisc player center - can be fulfilled by means of video­ who is also a telephone subscriber would get disc players working in parallel (similar to yearly "electronic directory" records from the Figure 17) enabling simultaneous usage of the PTTs. Since videodisc players are multi­ system by many users. The storage capacity purpose devices, this would not require any of one online videodisc record is up to 1.5 major hardware installation at the user's site

80 1982, Vol . 2, No. 1 ELECTRONIC PUBLISHING REVIEW

or on the telephone network - again pro­ point will be the degree to which it can pene­ vided that videodisc standardization be trate the mass market, which in tum will achieved. In countries with under-developed heavily depend on its usefulness and on user telephone infrastructures, the concept of an acceptance. The cost factor will be decisive in online telephone directory might be arguable this process; both videodisc players and since its usage would add unnecessarily to an videodisc records have to be mass produced already overloaded telephone network. This and distributed in order for costs to come would also have a negative impact on the down to a reasonable level. The production quality of the online "electronic directory" and broad selection of videodisc programs - service itself. be it video,audio, or text - is of basic However, whether the concept of an elec­ importance. If the choice of information on tronic directory were to come about through a videodisc records is not attractive to users and videotex network or through videodisc, there is not available in all the major application would remain one question: would the user categories - entertainment, movies, voice accept the concept, or would he prefer to recordings, encyclopedias, automatic direc­ continue to use the paper telephone direc­ tories, educational programs, etc. - it cannot tories and ask the central telephone informa­ attract the necessary broad consumer popula­ tion service for any unknown number? tion . Therefore, while building up videodisc player production sites, special attention has ( e) Videodisc as a tool for "electronic pub­ to be paid to the establishment of a videodisc lishing". Videodisc can be used to store image record industry. A necessary precondition for pages of any printed publication. The huge the healthy growth of the record production storage capacity of videodiscs would allow, industry will be the adoption of videodisc say, all the novels of Thomas Mann to be put standards; otherwise the flexibility of video­ on a single disc record. The contents of the disc technology will not be exploited and we record could then be shown page by page on will only see a limited diffusion of dedicated any videodisc recorder. The advantage of videodisc systems (for example, a dedicated such usage would be its low cost, and the system for a Chemical Abstracts, separate savings in paper and storage space. The dis­ automatic directory systems etc.). If the de­ advantage would be that reading from a TV velopment moves in that direction, then the screen would be far more tiring than reading a mass market cannot be reached and both hard­ book. ware, software, and videodisc record prices Other sorts of electronic publications might must remain high. It can be expected that the better utilize the "value added services" videodisc record manufacturing industry will offered by videodisc technology. Electronic appear in a concentrated form since the instal­ publications such as a "Guinness Book of lation of videodisc record production lines are Records" with voice and motion scenes or an expensive (on an order of magnitude of some "electronic cookbook" would utilize the millions of dollars) . In order to keep produc­ videodisc technology's sound and motion tion costs down, records must be produced in dimensions. Similarly, history books might large quantities. include motion and voice scenes. An encyclo­ In the audio area, hard competition can be pedia of animals on a videodisc record might expected between the digital audio and be an excellent an,d useful educational and ordinary LP producers. In the field of text entertainment medium for children. The publishing, competition with publishers of dimensions of random access, voice and mo­ printed materials can be expected. In the field tion are endless. of online databanks, there will be competition At this point, however, some thoughts among producers of videodisc records should be given to the "electronic publishing" containing large databanks, and among data­ industry as a whole. The success of videodisc base spinners who sell online services in the will depend on many factors. An important traditional way through networks and time-

1982, Vol.2,No. I 81 ELECTRONIC PUBLISHING REVIEW

sharing computer centers. For the producers tional publishing it will open many others in of databases, once they have made appro­ the new field. The publishing industry has priate arrangements for record production, it interesting times in sight. .. will make no principle difference whether they provide their subscribers with magnetic (j) Use of videodisc as timetables. Railway tape or videodisc up-dates. and flight timetables are similar in nature to The sharing of labor in electronic publish­ the previously discussed electronic directory, ing will be similar to that in the classical in that they are made up of masses of refer­ publishing industry. There will be a separate ence information up-dated about twice a year, videodisc record production industry on the and printed in bulky books. If they could be analogy of printer's shops. There will be elec­ stored electronically, large amounts of tronic publishing houses who will be respon­ materials, especially paper, could be saved. sible for the content of the videodisc records, One possibility would be to store timetables similar to traditional publishers. There will be on interactive videotex frames. This approach new type of author with a new type of skills: already functions perfectly on a small scale in addition to writing skills they will require with existing videotex technology. However, skills in video, voice, and also in videodisc similar to encyclopedias and electronic direc­ controller programming. The subject of tories , the volume of information to be stored videodisc authorship is obviously much presents a major barrier for the presently­ broader than this, and requires more concen­ known videotex systems. Thus using video­ trated study. disc players for this purpose, either linked A final thought here: videodisc is a new online to videotex networks as third party medium suitable for expressing ideas, computers or as stand-alone systems, would feelings, and emotions. The emergence of a be a feasible solution. Online access through new medium is often accompanied by the the videotex network could have an additional emergence of a new type of art. The advent of provision for subsequent online reservations the motion pictures triggered off the cine­ when a suitable train or flight is found and matic art, the development of radio broad­ free places are still available. Obviously, this casting led to the emergence of radio plays, sort of interactive traffic could not be per­ and the art of modem journalism was made formed by stand-alone videodisc devices. possible by the development of new printing Therefore, it seems that the information technologies. And the novels of T. Mann, retrieval type of usage of interactive videotex W .S. Maugham and L. Tolstoi could not have systems - which is at present the main usage been disseminated without printing. In a category of videotex-like systems - can be similar way, it is likely that videodisc tech­ taken over in many cases: by teletext or cable­ nology will be the basis for a new type of crea­ text if a full channel be put at the disposal of tive art based on video, audio, and text such services (around 100,000 videotex information and exploiting the mixing and frames can be rotated) or by videodisc as programming capability of the videodisc shown here. The main strength of interactive controller. videotex lies in its interactive nature, and Selling videodisc records will be a specia­ from the information retrieval point of view, lized new business with strong resemblance in its capability for frequent and quick update. to the selling of books, traditional records, films and games but requiring concentrated interdisciplinary marketing. 5.3 Videodisc as Too/for Education All in all, the "electronic publishing" We have seen what an excellent tool for industry is an exciting new field offering new information transfer the videodisc player is. opportunities and challenges to producers and Education is also a sort of information trans­ sellers of information as well as to buyers. fer: in a structured way, information is "fed" While it might eliminate some jobs in tradi- into the memory of a person being educated.

82 1982, Vol. 2, No. I ELECTRONIC PUBLISHING REVIEW

Theoretically, videodisc is an ideal tool for videodiscs will be excellent educational education. However, its acceptance in an devices. Many of the applications for video­ educational environment depends upon user discs discussed in previous chapters could impressions: Is it easy to use? Does it provide play major roles in education. new functions compared to devices used for The automated dictionary, for instance, similar purposes? could do much to promote language training In order to " measure" the "interface and education - perhaps more than any other quality" of videodisc, let us compare it with device before it. It will help to enrich users' the seven interface performance factors as vocabulary in the language, will promote defined by Card et al. [ 12]; spelling skills, word recognition, and aid in - Time: How long does it take a user to word search, synonym search and the like. Its accomplish a given set of tasks? capability for easy word retrieval could revo­ - Errors: How many errors does the user lutionize the translating process. make, and how serious are they? So-called "programmed course books" - Learning: How long does it take a novice have proved to be useful teaching tools for to learn how to use the system to do a given set language training. On the analogy of pro­ of tasks? grammed course books, programmed courses - Functionality: What range of tasks can a on videodiscs - so-called Computer Assisted user accomplish with the system? Instructions (CAI)- could be made available, - Recall: How easy is it for a user to recall perhaps containing voice and motion picture how to use the system for a task that he has not scenes in addition to text. The flow of the done for some time? course would be pre-programmed: the student - Concentration: How many things does a learns a new chapter supported by text, video user have to keep in mind while using the and audio information; a test then determines system? whether or not he has understood and - How tired do users get when they use the mastered the material. If he has not, the system for extended periods? material is repeated, this time in more detail and perhaps from a different angle. Subse­ Fox et al [8] name four additional factors that quently, a new test decides whether or not the must be considered when evaluating inter­ student may go on to the next chapter. Of faces of educational systems: course, should the student satisfactorily pass - Variety: Does the interface provide the test first, he may immediately go on to the information in a variety of ways; for example, next chapter. Supplementary tests could be does it offer color ratherthan black and white, used to allow particularly advanced students flashing images, speech output, motion to skip certain chapters. pictures? Programmed courses (CAI courses) could - Curiosity: Is the user's curiosity to explore be designed for practically any field: mathe­ piqued by the interface? matics, language, biology, engineering, pro­ - Fun: Does the user find the device fun to gramming, art, geography, and so on. The use? principle of Computer Aided Instruction is - Adaptability: Does the device adjust to the described by James Martin in The Wired user's level of competence as the user Society [20): becomes more experienced with the device? Computer-assisted instruction has Satisfying these demands can present a been used for a wide variety of people difficu:t design problem, but it seems that it and a wide angle of subject matter. It can be done for videodiscs relatively easily. has been used at most levels of educa­ In an educational environment it is of special tion, from very small children to grad­ importance that the last four factors be rela­ uate students and professionals. When tively easily fulfilled. This indicates that the system is well designed, the pupils

1982, Vol . 2, No. 1 83 ELECTRONIC PUBLISHING REVIEW

are captivated by the terminals; they tion. With computers, the process is learn at a fast rate with a high level of pupil-centered, not instructor-cen­ retention and finish each session with tered , and the machine adapts its pace a sense of accomplishment. The com­ to that of the student. Dull students can puter is programmed to respond to ask for endless repetitions without them sensibly, with infinite patience, embarrassment. Quick students or and with timing designed to maximize those who already partly know the the reinforcement of the information material can skip a segment - with in their minds. The pupils leave the the machine questioning them to terminals stimulated and often mentally check that they do in fact know it. fatigued. However, a badly designed Students who want more practice in a system can be immensely tedious and certain area can obtain it. can leave the pupils with their patience strained to breaking point, thinking Writing these elaborately structured that it is a worthless gimmick and that teaching programs requires much time they can learn much better from a and considerable skill. book. To write well for this medium requires considerable care and talent. An outstanding programme for com­ The writer must understand the subtle puter assisted instruction is a work of psychology of the medium. great art . A strong sense of style is Some topics lend themselves well to needed. This is a new medium quite computerized teaching, but others do different from any that have preceded not. Subjects involving large amounts it, and needs its own rules concerning of routine details or facts are natural style. No doubt a style guide will be for computers, as are subjects with developed in time, and it will probably elaborate but standardized logical change with time just as style in other procedures. It is ideal for teaching media, such as movies, has changed. spelling, simple mathemetical tech­ Computer-assisted instruction may be niques, the mechanics of foreign lan­ much more susceptible to change guages, statistics, computer program­ because of the intimate two-way inter­ ming, electronics, and so on . It would action, which no other medium has. have more difficulty teaching philoso­ phy, carpentry, basic principles of There are three levels at which comp­ calculus, or appreciation of music, uter assisted instruction can operate, although even here it could assist a differing in channel requirements. human teacher. The lowest level displays alphabetic Given a suitable subject and skilful text on the terminal. This, like all uses programming computer teaching can of computer dialogues with text, have significant advantages over con­ letters, and numbers, has low bit trans­ ventional classroom teaching. Teach­ mission requirements compared with ing in a classroom is almost always digitized telephone transmission. instructor-centered . The students have The medium can be more effective to proceed at the speed and level of and stimulating when pictures are complexity that is dictated by the used. The second level can display instructor. The brighter students, find­ still color pictures on the screen. Some ing the pace too slow, are bored; those systems have used two screens, one who are not so bright often become for text and one for pictures. Substan­ lost or fail to understand part of what is tially more bits are required to store or said. Any student, bright or dull, can transmit pictures, but the requirement miss sections through lapses in atten- is low compared with the transmission

84 1982, Vol. 2, No . 1 ELECTRONIC PUBLISHING REVIEW

capability of cable television. does not wear out, its life-time being practi­ The third level uses movies. An cally unlimited, make this technology educational movie will stop at inter­ particularly suitable for use in archives. The vals to show still frames and ask the major hindering factor at present is the students questions. The response to requirement for a high degree of production the questions will determine what precision and the clean atmosphere of an frames are shown next and whether the "operating theater". machine repeats or skips segments of As pointed out previously, the fact that movie. The selection of the next videodisc cannot be deleted and re-used is movie segment may take ten or twenty hardly a disadvantage for this application, in seconds, and this selection delay is light of the low costs of producing records and hidden by showing still frames. the fact that the final goal is to archive data which normally should not be deleted. The lowest level can be achieved by any in­ In many cases large amounts of data need teractive computer system: interactive video­ to be archived. Statistical data for countries, tex systems such as Prestel, for example, firms, towns, factories etc, can be stored and would be ideal for this. The second level, archived by this medium on a monthly or additionally showing high quality freeze yearly basis. Mass measurement data frame pictures, will be attained in the next gathered during experiments, such as during a years by Picture Prestel. The third level, at flight to the planet Saturn, must be archived, least on a long-term basis, will only be possi­ duplicated, and distributed as well. More­ ble through stand-alone videodisc systems. over, they must be protected from being over­ Perhaps sometime in the future fiber optic written. For this and similar purposes video­ link based interactive videotex systems will disc is a suitable medium. be able to compete with the videodisc in this Source and object code programs are also field. ideally suitable for storage on videodisc Continuing the theme of suitable videodisc records. applications in teaching and learning, refer­ Using a videodisc output device working in ence books on videodisc could serve educa­ computer centers with digital optical recording tional purposes as well. techniques as described in Section 3 and An analysis of movement - e.g., when shown in Figure 14, it will be possible to save learning ballet or skiing - can be made easily the content of the system and of library mag­ using videodiscs. No other type of video netic discs on videodisc on a regular basis. At device presently in use would allow such a present, system and library disc savings are full analysis of movement (normal motion , mostly made on magnetic tapes - usually slow motion, picture by picture, forward according to the grandfather-father-son motion, reverse motion, etc). principle; i.e. three subsequent generations The variety of videodisc applications for are stored at the same time. By using space­ educational purposes is endless and here an saving videodiscs one could build up a system attempt is made only to point out some of its and library savings archive containing snap­ potentials; a comprehensive mapping of all shots of the computer center system over educational applications cannot be given. many years. Where this is required, videodisc technology could be extremely useful.

5 .4 Videodisc for mass storage and archives of data and programs 5 .5 Videodisc in office automation and docu­ The fact that a videodisc record can contain ment filing such vast amounts of coded data, being the In the view of Barrett [23], ideally an opti­ most condensed form of mass information mum mass storage and document filing storage developed to date, and the fact that it system should ideally have the following

1982, Vol. 2, No. I 85 ELECTRONIC PUBLISHING REVIEW

characteristics: c) Retrieval - recall based on various a) non-erasable archival medium attributes of the data b) fast access time d) Composition - changing form, appear­ c) high data rate (for input, processing, ance, and structure of the data as and when retrieval and output) required d) low cost e) Distribution - recognizing multiple needs e) volumetric efficiency or group actions f) data and/or image storage f) Communication with geographically dis­ g) expandability to at least 10- 15 bits capacity. persed users and sources. The requirement for an non-erasable archiva­ Of course, computers and word processing ble medium would exclude magnetic tapes machines do the lion's share of the above because of their sensitivity to temperature, work. Communication can be carried out humidity, and electrical/magnetic interfer­ either via online telecommunication channels ence. (Annual re-copying is often resorted or offline (e.g. by sending diskettes or video­ to). From this point of view, microfilm, disc records by mail). However, for office holograms, and any other inexpensive write­ automation functions such as permanent once, read-often media are preferable. (Micro­ storage, analog or digital optical recording filmed evidence is now acceptable in a court systems seem to be extremely useful. Sys­ of law, whereas magnetic tape is not). Mag­ tems like the previously described Philips netic tape was originally intended for short­ DOR system or the Document Filing systems term data capture and manipulation; how­ announced by Toshiba (DF-2000) could play ever, its erasability is undesirable where long­ an important role here. term integrity is sought. The Toshiba DF-2000 image recorder/laser A typical access time for magnetic tape file system which will soon be available on mass memories is 15 seconds; microform the US and Japanese markets, will be the first retrieval systems available can achieve 5 videodisc technology office system on sale in seconds, depending on capacity. Require­ the US. It will consist of a high resolution ments vary with the application, but 5 laser scanner input device (8.5 seconds input seconds would probably be adequate for time per document), a high quality laser document-retrieval systems; a requirement printer ( 14 seconds for printing a single docu­ easily met with the videodisc technology. ment), a central computer (4.5 seconds for In discussing the storage potential of video­ retrieving a given document and calling it up disc, Harry Collier [22] points out that one on screen) and an analog optical disc storage videodisc record device (capacity of 10,000 document pages = 35 magnetic tapes per record). The system is expected to cost = 500 microfiches only around US $50,000. A record will cost = 182. 5 feet of microfilm $140. This cost level will make these systems = 50,000 A4 pages in image format highly competitive with other systems, such = 500,000 A4 pages in coded alphanumerical as those based on microform. format.

Office automation systems require the 6.SUMMARY AND CONCLUSIONS following basic functions in addition to docu­ [l] Optical disc technology is a revolutionary ment filing: new technique that will have a major impact a) Recording - the acquisition and input of on both information technology and the elec­ information tronics industry. We see at present the emer­ b) Storage - temporary or permanent gence of analog storage technology for video­ retention discs with the tendency of storing growing

86 1982, Vol. 2, No. 1 ELECTRONIC PUBLISHING REVIEW

amount of digital information in a "pseudo" potential; for instance, the inter-changeability digital form . Videodisc with digital storage of records from different systems could not be technology will also be entering the market assured.Standardization is needed not only in soon . Digital systems based on the PCM the physical characteristics of videodisc (Pulse Code Modulation) technique allow records, but also in content, programming, sound and video signals to be recorded in high methods, coding of information, error quality at a relatively cheap cost. Also , recovery, and so on . another digital technique is used to store vast [4) Videodisc technology will be applied in amounts of coded digital information, such as revolutionary new ways: interactive TV, an of coded text, data, and computer programs. electronic encyclopedia containing textual, These two digital coding techniques will be audio, and video information, automated used in future vidoedisc systems, allowing dictionaries, online information banks, time­ flexible mixing of audio, video and coded tables, electronic directories, and other types textual information. In principle, mixing of of electronic reference devices. media will also be possible on analog video­ disc . The memory unit price for optical [SJ Videodisc technology has great potential storage systems - which we believe will be in the area of entertainment and education . It the basic technology to be used in videodisc can be expected that this technology will - is far below that of all other presently considerably change classical teaching known and used techniques, especially on the methods . mass produced (pressed) videodisc records. 161 Videodisc could have a significant impact However, this very positive cost factor might on online information banks. Online informa­ be negatively influenced if data accuracy tion service centers as we presently know problems with mass produced videodisc them could lose their biggest customers, who records cannot be solved cost effectively. might install their own videodisc data centers [2) Market penetration and the overall success at low cost.On the other hand, service costs of the videodisc technology will depend on for online information service centers opera­ the availability of appropriate videodisc pro­ ting videodisc storage devices could drop. grams. Present difficulties lie in the produc­ [7) Videodisc will have a major impact on tion technology of records. As experience videotex technologies as we presently know is gained, however, it is hoped that these them. The information retrieval function of problems can be overcome. videotex will be particularly affected. Some Another problem is that the production of applications, such as encyclopedias, are the laser-technology based videodisc requires better suited to videodisc. As we have pointed a completely new technology and new record out in other studies [ 17], [ 18], the success of production and tools. Accumulation of the videotex-like systems will depend on the capital needed to begin production might scope of their application. At present, how­ be a prohibitive factor in a time of economic ever, videotex systems primarily support recession. Industrial policy factors, i.e. 'who information retrieval, an application type is backing or not backing what' might be which might be especially affected by video­ crucial to the success of this new medium. disc technology. Where frequent information [3] The development of uniform videodisc update is required, videotex technology will standards is urgently needed. There are continue to play a leading role; where huge already a few different systems, although the amounts of information needing relatively technology is still in its infancy. The elabora­ infrequent update are required, videodisc tion and approval of standards will be a long technology is preferable. Videodisc storage and painful process, perhaps taking years. devices will be particularly useful for replac­ But without standards the production of ing magnetic disc storage devices in videotex videodisc records cannot fully exploit its information centers.

1982, Vol.2.No. I 87 ELECTRONIC PUBLISHING REVIEW

[8] Videodisc technology has a "dualistic [12]S.K. Card, T.P. Moran and A. Newell: nature" with both consumer electronics and 'The Keystroke-level Model for User information technology aspects. A conver­ Performance Time with Interactive Sys­ gence of video and information technology tems.' Report SSL-79-1, Xerox Palo will form a solid base for the emerging Alto Research Center, Palo Alto, CA., "videomatics" era. 1979. [I 3]Electronic Publishing Review: 'Pergamon launches patent drawings on videodisc' (news item), Electronic Publishing Review, 1981, 1 (2), p. 71. References [14]H.A. Maurer, W. Rauch and I. Sebes­ [l] F. Barry: 'Videodiscs - too late for the tyen: 'Teleplaying in Videotex'. IIASA gravy train?' New Scientist, 1981, 91, Working Paper (in preparation), Laxen­ p. 1264. burg, Austria: International Institute for [2] R.M. White: 'Disk-storage technology,' Applied Systems Analysis. Scientific American, 1980, 243 (2), [15]S.C. Cheek and D. Simpson: 'Common pp. 112-121. musical notation, computers and graph­ [3] P. Isaacson and J.E. Juliussen: 'Window ics,' Computers and Graphics, 1980, 5, on the 80's,' Computer, 1980, 13 (I). pp. 87-91. [4] D.J. Theis: 'An overview of memory [16]Sony Corporation: ' Digital Audio Tech­ technologies,' Datamation, 1978, 24 (I). nology'. Sony Brochure, 1981. [5) J.E. Juliussen and W.J. Watson: 'Prob­ [17]H.A. Maurer, W. Rauch and 1. Sebes­ lems of the 80's: Computer system organ­ tyen: 'Videotex Message Service Sys­ ization.' The Oregon Report: Proc. tems'. IIASA WP-81-113. Laxenburg, C onf. on Computing in the 1980' s. Austria: International Institute for App­ Portland, USA: IEEE Computer Society, lied Systems Analysis, 1981. 1978. [18]H.A. Maurer: ' Bildschirmtext- [6] I. Sebestyen: "'Videomatics": the con­ aehnliche Systeme' . Report B 11, June vergence of video and information tech­ 1981 . Graz, Austria: I nsti tut fi.ir nologies.' IIASA working paper (in prep­ Infonnationsverarbeitung Technische aration). Laxenburg, Austria: Inter­ Universitat Graz. national Institute for Applied Systems [19]W. Hoekstra: 'Megadoc: a document ar­ Analysis. chiving system based on digital optical [7) S. Nora and A. Mine: The Computeriza­ recording,' in Electronic Document De­ tion of Society.' La Documentation livery (J. Page, ed.). Proceedings of an Frarn;:aise, Paris, 1978. Exhibition and Workshop on Docu­ [8] M.S. Fox, D.J. Bebel and A.C. Parker: ments, Luxembourg, December 1980. 'The automated dictionary,' Computer, Oxford: Learned Information, 1981. 1980, 13 (7). [20]J. Martin: The Wired Society. Englewood [9] H .R. Collier: 'The politics of European Cliffs, New Jersey: Prentice Hall, 1978. information: ritual suicide?' Monitor, (21 ]AT & T Long Lines: The World's Tele­ Sept. 1981 , no. 7. phones: A Statistical Compilation as of [ lO]Foel Stratte, McClure: 'French telecom­ January, 1980. Morris Plains, New munications: digital technology and the Jersey: AT & T, 1981. Telematique program,' Scientific Ameri­ [22]H. Collier: 'The videodisc in information can, 1980, 243(3),pp. 25-43. retrieval'. Lecture given at the EUSIDIC [I I ]United Nations Department of Interna­ Annual Conference, 13-15 October tional Economic and Social Affairs 1981, Bern, Switzerland. (Statistical Office): Statistical Yearbook [23]R. Barrett: 'Developments in Optical 1978. United Nations, New York, 1979. Disc Technology and the Implications for

88 1982, Vol. 2, No. 1 ELECTRONIC PUBLISHING REVIEW

Information Storage and Retrieval ' . The [27]Philips Press Information: Optischer British Library Research & Development speicher mit Diodenlaser zeichnet 1010 Reports. No. 5623, Boston Spa, UK, bit auf einer JO-Centimeter-Platte auf. 1981. Netherlands: Philips Press, 1978. [24]G.C. Kennedy et al.: 'An optical disc [28]Predicast Inc .: Data Storage Devices - replaces 25 mag tapes ,' IEEE Spectrum. Market Report GR-34. Cleveland, Ohio: February 1979. Predicast Inc ., 1981. [25]F.L. Greenagel: 'Arete - a 3000-year­ [29]E. Sigel, M. Schubin and P. Merrill: old word for the latest in electronic pub­ Videodiscs: the Technology, the Applica­ lishing,' Electronic Publishing Review, tions and the Future. White Plains, N. Y.: 1981, 1 (3), pp. 177-182. Knowledge Industry Publications, 1980. [26]New Scientist: 'Slow start for laser video­ [30]Personal discussions on digital optical disc factory,' New Scientist, 1 October recording with J. W. Klimbie from Phil­ 1981. ips Data Systems, Netherlands and U. Glhofer, Product Manager, Philips Data Systems, Austria.

1982 , Vol. 2, No . I 89

RELATED PUBLICATIONS

Maurer, H., W. Rauch, and I. Sebestyen (1981) Alphabetic searching in videotex systems. Electronic Publishing Review 1(3):217-223. Also issued as IIASA Research Report RR-82-11. Maurer, H.A., W. Rauch, and I. Sebestyen (1981) Videotex message service systems. Electronic Pub­ lishing Review 1(4):267-296. Also issued as IIASA Research Report RR-82-10. Maurer, H. and I. Sebestyen (1982) Unorthodox videotex applications: teleplaying, telegambling, telesoftware, and telecomputing. To appear in Information Services & Use.