The History of Interactive Compact Disc-based Multimedia
The following text is a chapter from the book The Shape of Things to Consume -- Delivering Information Technology into the Home, written by Alan Cawson and Leslie Haddon of the University of Sussex and Ian Miles of the University of Manchester. The book was released by Avebury in 1995 (ISBN 1 85972 052 8). The authors gave permission to use this text. All opinions are solely those of the authors, not the ICDIA.
brief history of optical disc technology
CD-based multimedia technology derives from research on the application of the laser to the recording and retrieval of information from a spinning disc. In the recording process tiny pits are burned into the surface of the aluminium disc, which is then protected by a plastic skin. On playback, differences in the refraction of the laser between the pits and the shiny surface are detected and converted into a signal which can then be processed in different ways according to the information content.
The original research programme began in the late 1960s. At its central R&D laboratories in the Netherlands, Philips engineers identified three possible uses for this technology, and began parallel projects concerned with data storage, video and audio (see figure).
The first of these was concerned with mass data storage for professional applications, and has been marketed as the Megadoc system. The second was the Video Long Play project, which later became LaserVision and finally LaserDisc. At the same time a joint venture between MCA and IBM in the United States — Discovision — was working on exactly similar lines, and the final specification for LaserVision was jointly agreed between Philips and Discovision and launched onto the consumer market in Britain in 1982. A double-sided 12 inch optical disc stored video and soundtrack in analogue form, and up to 72 minutes of video could be recorded on the disc.
The launch of LaserVision in Europe coincided with the surge in sales of videocassette recorders (VCRs), and it managed to achieve only a very limited penetration of the domestic market. However, the ability of the technology to permit random access to any point on the disc allowed for the coupling of the LaserVision player to a computer so that video sequences could be accessed interactively. In May 1985 Philips announced a shift towards professional markets and interactive applications for LaserVision.
The research on the application of laser technology to audio recording followed a slightly different path, but again the parallel research being conducted by other companies resulted in collaboration and the pooling of technology. Whilst the formidable storage requirements of moving video pictures constrained LaserVision to analogue formats, recording audio alone permitted the signal to be stored digitally which allowed for much more accurate reproduction than was available from long playing records or audio cassettes. The disc could also be smaller (12 cm rather than 12 in) and still hold considerably more than an LP record (72 minutes rather than 50).
While Philips were developing CD along a parallel path to LaserVision, Sony were busy adapting LaserVision technology for audio applications. By 1976 Sony had achieved an optical audio disc with 30 minutes playing time on each side; two years later they had stretched it to 2.5 hours per side. Philips announced their own ‘compact disc’ format in July 1978 for launch in the early 1980s. The following year Philips and Sony signed a co-operation agreement, and submitted a common standard to the Japanese Digital Audio Disc Committee. At that time it was known that several other companies were developing digital disc technology, including RCA, Pioneer, AEG-Telefunken and JVC. Sony had strengths in digital-to-analogue conversion techniques, and had developed a method of error correction based on the Reed-Solomon code, and Philips had unrivalled expertise in laser technology. The final jointly agreed specifications (known as the ‘Red Book’) were released to licensees in 1982, and the first products reached the market in 1983.
CD-Audio, unlike LaserVision, achieved dramatic success in the consumer market, and became the fastest selling product in the history of consumer electronics. Philips had learned from its earlier failure with its own VCR technology (the V2000 system) that there were enormous risks in trying to push its own product against competing formats. The decision to collaborate on the CD-Audio standard was the immediate consequence of this failure, and the need to collaborate to avoid this fate was to become almost an article of faith in shaping Philips’s strategy on the development of other products within the CD family, including CD-i.
Both Philips and Sony appreciated early on that CD could easily be applied to data storage, and they agreed a provisional standard for encoding data on compact discs — CD-Read Only Memory, or CD-ROM — in October 1983. This product was envisaged as a computer peripheral which would be developed principally by third party suppliers within the computer industry. The Philips/Sony standard (known as the ‘Yellow Book’) covered the physical aspects on recording on the disc and error correction; it did not extend to file structures and file handling, so that initially not all CD-ROM discs were compatible. A full standardisation was achieved by an ad hoc industry group in the USA (the High Sierra standard) and adopted by the International Standards Organisation as ISO 9660.
CD-ROM has developed as the most significant technology in the new industry of optical publishing, and is widely used to distribute large databases, principally to professional users, but since 1992 the market has taken off for individual PC and Macintosh users. The 550 megabyte capacity of the disc permits some 100,000 pages of A4 text to be stored on a single disc, with any part of the text selectable within one second. CD-ROM players have been marketed to existing personal computer users as an add-on, but in 1990 Sony launched a hand-held unit, the Data Discman, which uses a liquid crystal screen to display data retrieved from an 8 cm disc containing 300 megabytes of text, and later the Bookman using conventional 12 cm CD-ROM discs.
During 1984 Philips and Sony began independently to work on another
derivative of CD which would combine audio, text and graphics. The two
companies joined forces to develop a draft standard at the beginning of
1985, and later that year Matsushita joined in to work on the development
of integrated circuits. The first public announcement
of the new product — Compact Disc-Interactive — was made at the first industry
conference convened in March 1986 to promote CD-ROM in the United States.
A provisional standard (the ‘Green Book’) was issued in May, followed by
a full functional specification of the system in March 1987. CD-i discs
and prototypes were demonstrated to licensees in June 1987, and the first
working samples of players were distributed to developers in Autumn 1988.
CD-i is an interactive multi-media technology; that is, it combines sound and pictures (including still and moving pictures and of course text), and allows the user to control what information is retrieved from the disc and how it is displayed. Unlike television or film, which are linear media in that they unfold in a fixed sequence predetermined by the producer or author, interactive technologies allows consumers themselves to determine many aspects of how the material is ‘read’, although within the limits built in by the software designers. As such CD-i resembles the newspaper which can be read in the order and manner desired by the consumer. In this section we will briefly outline in technical terms how this interactivity is achieved, noting that technical choices are often constrained by marketing goals, and vice versa.
CD-i is an all-digital medium; that is, unlike LaserVision, all of the audio and visual material is encoded on the disc as digital information. This aspect of CD-i allows for world-wide compatibility of software, since the visual information is independent of any of the existing television transmission standards, and permits the CD-i disc to be perceived as fully compatible in the same way that the LP record, the compact audio cassette and the CD-Audio disc are, but importantly not computer disks or programs.
The initial specification of CD-i allowed for very limited full-screen full-motion video images because of its prodigious storage requirements — even 550 megabytes would allow only for five minutes of full-screen video. The original designers of CD-i coped with this limitation by restricting motion video to a window which occupied only a fraction of the screen, but subsequently, as we shall see below, sophisticated techniques for compressing video pictures have been developed. The decision not to release CD-i in its original form, but to wait for the perfection of the video compression technology, was prompted by the announcement by RCA in 1987 of a potential rival to CD-i which combined CD-ROM with video compression. Digital Video Interactive (DVI) comprised two chip sets which could be used together with an IBM personal computer and CD-ROM player to display up to 72 minutes of video which had been compressed (using a mainframe computer) onto a standard CD disc.
CD-i is a derivative of earlier CD-Audio and CD-ROM formats, and some of the technical parameters of those formats have been carried forward into CD-i. CD-DA specified how the audio information was to be stored on the disc, as well as basic parameters such as disc size, linear speed of rotation of the disc, the laser read-out mechanism, and the detailed format by which data can be recorded on the disc. CD-ROM extends these basic CD system specifications by dividing the basic data stream into discrete units called sectors, each of which contains 2352 bytes of information. Any sector can be specified as one of two modes: a mode with less user data but more error correction, which is essential for accurate storage of text and numerical data; and a mode for audio and video information with less error correction permitting faster transfer of data from the disc. The CD-ROM specification permits audio and data to be stored on the same disc which opens up limited multimedia (text and sound) applications. A further enhancement of the CD-ROM specification, announced in 1988 as CD-ROM XA (for extended architecture) specifies how graphics can be coded and interleaved, thus permitting the development of applications within the PC world of multi-media applications originally designed for the standalone CD-i system.
From a technical point of view the CD-i specifications are based on CD-ROM, but the conception of the product derives from CD-Audio, that is, as a world-standardised product. As with CD-Audio, any CD-i disc can be played on any CD-i player. Given that part of the original product design was that CD-i players would plug into domestic television receivers (and thus not require the purchase of a specific monitor), video information must be handled in a way which is independent of the three television standards (NTSC, PAL and SECAM). Such compatibility — a basic principle of CD-i design — required that CD-i be specified as a combination of hardware and system software. Besides this, the standard (‘Green Book’) allows for the mixing of CD-DA and CD-i tracks on CD-i discs, and requires that all CD-i systems are able to play CD-DA discs.
There were three further basic principles — or as Philips engineers put it, logical requirements — of the system. The first of these was self-contain-ment: all that is required to play a CD-i disc is contained in a single unit which plugs into a TV set; software designers must not assume the presence of other peripherals such as a floppy disk drive — all the software to run the application must be contained on the disc itself. Secondly, CD-i should connect to and build on existing mass produced consumer electronics products. CD-i players use the basic drive mechanism of the CD-Audio player, and clearly this requirement is intended to permit rapid progress to high volume manufacture for the consumer market. Finally, the system should be made as ‘future proof’ as possible, allowing for future enhance-ments of the system whilst always maintaining compatibility with existing systems.
In the CD-i specification audio, graphics and text can all be recorded on the disc in different ways to suit different applications. Each has different space requirements, so that, for example, a combination of moving video and high quality sound will make heavy demands on disc space. If the quality of the audio is reduced, or video confined to a portion of the screen, there is more space for graphics, and so on.
The brief technical overview given above demonstrates how fundamental choices in the design of the system follow from the basic principles adopted by Philips and Sony at the outset, which in turn were the sediment of experience of both successful (CD-Audio) and unsuccessful (V2000 VCR) product innovation. CD-i shares with CD-Audio the characteristics of a software-dependent product; that is, the hardware is the delivery mechanism for the application contained in the software. CD-Audio had the advantage of an existing software base (in the form of the back catalogues of the music companies) which could quickly be converted to the new format once the publishers were persuaded of the market potential of the product. Besides that, Philips owns PolyGram, one of the largest music companies in the world, so that at least one company was already converted. The experience with other similar products such as record and tape players also suggested that as the market developed the proportion of total sales accounted for by software rose constantly.
Thus it was clear at the outset that the production of software would be a key determinant of the success of the product. The problem faced by Philips was that no-one had any experience of writing interactive multimedia software. Some of the relevant skills could be found in the computer software industry, but other equally important skills involved in handling video and audio were more likely to be found in the broadcasting industry. CD-ROM’s orientation as a form of electronic text publishing had in the beginning been clearly towards major book publishers and database providers, for whom CD-ROM’s advantages could relatively easily be compared to their existing activities. Such publishers were among the actors likely to be interested in multimedia publishing, but to what extent were they likely to emerge as first movers? As we shall see later on, the problem of securing software development was the most critical part of the introduction of CD-i and entailed the creation of a series of joint ventures in which, in effect, the risks of entering this new field were underwritten by PolyGram and Philips. The uniqueness of CD-i lies not in its software-dependence, but in the specific character of the software itself, and the complexity (and cost) of the tasks involved in producing it. The temptation for any innovator in such circumstances is to exploit the proprietary nature of the technology, and seek to maximise the rent obtainable from the value added to the technology in the software production process. It is perhaps not surprising that Philips at first succumbed to that temptation by seeking to retain proprietary control over the authoring tools necessary to develop CD-i titles. At first it was envisaged that Philips itself would establish the studios to which software producers would come with their programme material, so that in effect Philips would be a partner in every piece of CD-i software. Later it was recognised that this would have a crippling effect on the growth of the software industry, and that more was to be gained by trying to maximise the number of players by lowering the barriers to entry. Thus in 1990 Philips introduced a range of hardware options and developers’ tools to encourage small software houses to enter the industry.
Software-dependence has other aspects besides the problem of ensuring
software availability when the hardware is launched. The ‘product’ is the
combination of hardware and software, but in this case the identity of
the product was thought to be more strongly influenced by the nature of
the launch software. Because of the greater availability of the relevant
expertise, Philips had decided to locate almost all of the early software
development in the United States, where Philips and PolyGram had formed
American Interactive Media (AIM), based in Los Angeles. AIM had itself
negotiated joint ventures with a wide range of developers in order to produce
the thirty or so software titles which Philips had estimated to be the
minimum necessary for a successful launch. Having done this, the marketing
strategy for positioning the product was in effect constrained by what
would be available, although efforts had been made to produce a range of
titles that encompassed ‘serious’ material such as encyclopaedias as well
as games. This aspect of software-dependence will be discussed more fully
It is perhaps misleading to speak of CD-i as a ‘product’ at all. It is better understood as an enabling technology, which permits the development of a number of different product configurations according to intended applications and markets. The most important feature, to which we will return again and again, is hardware-software interdependence, but also significant is the distinction between consumer and professional applications, which leads to separate but overlapping requirements. Moreover, the definition of the ‘product space’, and the more precise parameters for potential products within that space, is a continuous process which does not come to an end at the launch of the product.
Like many electronics companies in the late 1970s, Philips was faced with the prospect of a convergence between the previously discrete technologies of computing, telecommunications and consumer electronics. Computers were increasingly being linked through telecoms networks, telecommunications switches were becoming giant computers, more and more consumer products were embodying microprocessor control. Moreover the advent of the home computer at the end of the 1970s, and its emergence as a mass market product in the early 1980s, suggested a range of possibilities (i.e. a ‘product space’) for new information technology products targeted at the home.
The most common designation for this product space is ‘interactive multimedia’, or, increasingly, just ‘multimedia’ — that is, the simultaneous and complementary use of audio, textual and visual material which can be stored and then retrieved interactively by the user. Different companies have settled on different generic descriptions for this product space, like ‘hypermedia’ which Apple prefers, or the more cumbersome ‘AVCC’ (audio, video, computers and communication) which is favoured by Matsushita.
Philips’s organisational response was to set up a new corporate group,
Home Interactive Systems, intended to develop new products based on the
opportunities offered by technological convergence, outside the existing
product division structure (which had separate divisions for audio, video
and data systems). The project to develop interactive compact disc was
located within HIS, and was initially conceived of as only one element
of a home system. The component parts of the system
would allow for a number of different functions, and consumers would be
able to build up their own network. A senior executive in HIS described
in 1985 what the consumer could expect to have in the near future next
to the stereo system and the video:
It was, however, recognised that the ‘computer’ element would need to
be disguised as a result of early experience of consumers with home computers.
An American HIS group was formed at Knoxville in January 1984, and as its
first task it analysed the home computer market in the US. It used a clinical
psychologist to conduct in-depth interviews with owners and potential owners
of home computers:
The need to dissociate interactive CD from the home computer quickly became a fundamental design principle. In the early discussions it was recognised that the large majority of consumers found computers difficult to use, and part of this lay in the use of a keyboard for input. However, since the 1970s, and especially since VCRs became widespread, hand-held remote control devices had become increasingly popular. The design solution to user friendliness lay in a combination of familiar hardware (the remote control) and software which took advantage of the sophisticated graphical user interfaces such as that developed for the Macintosh computer. To reinforce the identification with familiar consumer electronics products, and further distance the product from the computer, the screen icons often chosen in software programmes are the now standardised symbols from tape and video recorders: > for play, >> for fast forward, << for fast back,  for stop, and so on.
A further development of the user-machine interface arose directly from the requirements of software development. One of the first joint venture agreements signed by American Interactive Media was with Children’s Tele-vision Workshop, the producers of the phenomenally successful children’s television show Sesame Street. The objective was to produce a title which carried the dual education/entertainment nature of the television show into an interactive format. Pre-school children would be able to learn letters and numbers by interacting with the same characters that they knew from the screen. But CTW thought that the target group of very young children would find the hand-held prototype remote control that Philips had designed in Eindhoven difficult to use. As a condition of the contract CTW insisted that Philips finance the $200,000 development costs of an entirely new trackerball remote control the size of a briefcase which children could have on the floor in front of them, or operate from their laps.
Professional hardware product development
Philips’s initial strategic assessment of its technology was wedded to the idea of a proprietary asset which it would control — albeit by necessity in partnership with Sony — in order to maximise the expected income. The dominant theme was monopoly, where control of a smaller market was preferable to a minor share of a large market, even if in absolute terms the latter was greater than the former. Even after the idea of monopolising all software production was abandoned, this mode of thinking dominated the decision to develop authoring tools only within the chosen operating system of CD-i. In this way it was expected that software developers would have to come to Philips for a complete hardware and development software package. Thus the first professional CD-i players, which were released to potential developers in 1988 in order that they could evaluate the technology, were complete development platforms, rather than designed as peripherals for existing computer equipment (see below).
The announcement of a rival interactive multi-media technology, DVI (Digital Video Interactive) by RCA in 1987 forced a reconsideration, but not at first an abandonment, of this strategy. DVI comprised a chip-set which would handle the decompression of previously compressed video images, so that 72 minutes full-motion full-screen video could be stored on a compact disc. It was designed to work within the IBM PC environment, so that potential software developers would be able to make use of their existing equipment. This context led Philips to the view that it was unlikely to mount a serious challenge to CD-i, which was at that time firmly targeted at the mass consumer market. But when Intel, a major semi-conductor manufacturer and supplier to IBM, bought the rights to DVI in 1988, Philips began to take the potential challenge more seriously, especially as Intel invited developers to consider mass market applications for DVI. As late as mid-1989 Philips was continuing to insist that it had no plans to market a CD-i board so that software developers working within the Macintosh or IBM environments could develop CD-i and other multimedia applications whilst using the same installed base of computer equipment. Product management within Philips remained wedded to the idea that CD-ROM XA would provide the bridge between the PC world and CD-i. As we shall see, however, by mid-1990 the position had changed completely and Philips were offering CD-i tools designed to match whatever equipment developers were already using.
The decision to abandon exclusivity was market-led. In order to provide a wide range of software, applications developers had to be encouraged to switch to the new format. The longer the lead time between concept and launch (which as we have seen was lengthening all the time), then the more real was the prospect of competing technologies, and the less credible a monopolistic strategy was likely to be. So from the initial idea of marketing professional players, but being the only source of software development, Philips found itself entering the market as a producer and seller of hardware and software for applications development.
The first actual hardware product sold to potential developers was a
three-box player, designed in Eindhoven but manufactured by Kyocera in
Japan, which began to be available from September 1989. The CDI-180 player
worked with the CDI-181 Multi Media Controller to play CD-i discs, and
could be hooked up to the CDI-182 Expansion Module, which in addition to
adding floppy disc storage, provided interfaces to conventional computer
equipment such as printers, hard discs, modems and MIDI equipment.
The CD-180 was not sold by itself, but as part of a ‘starter pack’ to include
some development software. The rationale for doing that again reflects
Philips’s proprietary and paternalistic inclinations:
The second level of entry into the development of CD-i software was what Philips refers to as a ‘personal publishing system’ which allowed for the development of complete CD-i titles which could then be pressed onto a disc. The initial product launch included two versions — one for the PC and one for the Macintosh — and further versions were announced for Sony and UNIX workstations. The PC version was a complete system costing £50,000 to include a dual processor computer and 766 Mbyte hard disc; the Macintosh version was an upgrade kit for £10,000 which comprised an emulator card and software and required standard audio and video capture cards to provide the same functionality as the PC version. It was envisaged that these products would sell to large corporations deciding to adopt CD-i for in-house applications, such as training or parts information, but who did not themselves intend to become CD-i publishers to a wider market.
For those who did intend to become CD-i publishers, Philips designed a third level of entry which was intended to be run as a network of workstations, so that the various tasks of audio and video processing could be done simultaneously by different people connected to the system. (In Levels 1 and 2 only one person can use the equipment at any time). The heart of this was a Sun SPARC workstation, which could be linked via an Ethernet to additional Sun workstations, or to PCs or Macintoshes used for video and audio processing.
The hardware players which were intended for sale to institutional users,
as opposed to software developers, took some time to define, since there
continued to be uncertainty as to the likely uses for CD-i. Part of the
appeal of CD-i for institutions, however, was expected to be that it would
be a mass market product rather than a product for professional users only.
In this respect CD-i is the equivalent of VHS rather than U-Matic in terms
of video recorders, and compared to current interactive videodisc technology,
it would be cheap — £1,000 rather than £3,000. The only technical
difference between the consumer machine and the base-level desktop professional
machine would be that the latter is likely to have a floppy disc drive
and more interfaces. Above the base case, the principles rather than the
details had been formulated. These relied on a threefold distinction drawn
from types of applications:
These products signified ‘a radical shift away from the big centralist studio concept’ towards a design philosophy for professional products which is ‘more in tune with the real developers’ community.’ It was the culmination of the evolving awareness that CD-i had to be sold first to software developers if there was to be any hope of it reaching a mass market, and is a clear example of how marketing considerations impacted on product design during the pre-launch phase.
Consumer hardware product development
Product design issues specific to the consumer version of the CD-i player can be conveniently discussed under the headings of technical issues — those relating to the configuration of the hardware for domestic applications; user interface issues — relating to how interactivity is put into effect; and physical design issues — the appearance and ‘image’ of the consumer product.
The overriding technical issue concerns the significance of incorporating full-frame full-motion video (FMV) into the consumer product. As we have seen, the original technical specifications of the CD-i standard allowed for a limited amount of FMV (5 minutes), but in practice applications developers tended to incorporate part-screen FMV in their prototypes to permit disc space to be conserved for stills and computer generated moving graphics (as well as text). The announcement of the DVI chipset in 1987 presented a direct challenge to the initial assumptions, and initiated a crash programme by engineers in Philips (and Matsushita — see below) to develop a video compression technology to match DVI. The perceived necessity of meeting the challenge from DVI was sufficiently overriding to set back the proposed launch date of the product by over three years. The compression algorithm was fixed in April 1989, with production samples of the chips due in Autumn 1990 and full production was due to start in early 1991.
The development team at Eindhoven considered the possibility of launching CD-i without FMV, and then offering an upgrade to consumers when the FMV chips were ready. This option was rejected for the consumer version, but as we have seen remained part of the programme for professional models. The fundamental design principle of full compatibility would have been compromised if the upgrade path had been followed, since discs incorporating FMV would not have been fully playable on early machines. The simplicity of the message to consumers (all discs are playable on all players) would have been lost. But this consideration was later overruled by the pressing need to market the product, once Commodore launched its own rival CDTV product. Philips decided to delay the final design for the FMV chip until the international standards body’s (ISO’s) Moving Pictures Experts Group (MPEG) adopted a standard for video compression technology. CD-i was launched in the UK in the autumn of 1991 at £699, with players containing a slot in which a promised £130 FMV upgrade could be fitted.
Much more uncertainty, however, surrounded the relationship between CD-Audio players and CD-i players. The basic laser assembly and the motor drives are common to the two, and it would be feasible to offer CD-i as an attachment to CD-Audio players. Indeed the inclusion of digital input/output sockets on top-of-range CD-Audio players makes enhancement a possibility for at least part of the existing stock of CD-Audio players in consumers’ homes. This option was strongly supported by A.D. Little in their report to Philips UK, and they suggested that such CD-Audio players be marketed with a ‘CD-i ready’ label — a route that was in fact chosen for certain CD-Audio discs (see below). Philips’s own product plans for CD-i players did not (as of March 1990) include provision for such a product, but managers argued that they could quickly be developed if a demand was foreseen.
Another issue which divided the design team concerns the relationship between CD-i and other optical disc technologies, and whether a single product might be offered which is capable of playing a range of discs. The CD-i specifications ensure that all CD-i players will play CD-Audio discs, but not LaserVision videodiscs. In 1988 Philips ‘re-launched’ their LaserVision consumer technology as ‘CD-Video’ offering 12 cm discs containing CD digital sound together with up to 5 minutes of (analogue) video aimed especially at the pop music market. The CD-V range also included 8 inch and 12 inch discs, and Philips and other companies produced combination players to accept all sizes of CD-Video discs. Whilst CD-V failed to give the boost that Philips had hope for to its flagging LaserVision sales in Europe, sales of combination players in Japan and the United States, especially by Pioneer, have been strong. Recognising that the attempt to link a failed consumer product (LaserVision) to a successful one (CD-Audio) had itself failed, in 1990 Philips announced that the name CD-V would be dropped in favour of Pioneer’s term ‘LaserDisc’. Perhaps not wanting to add to this confusion, Philips had no plans to market a combination CD-i/LaserDisc player at the time of the A.D. Little report, but only a few months later a combination player was planned for introduction to the consumer market as its second CD-i product a few months after the launch.
A further technical issue in CD-i product design was a clear illustration of the conflicting pressures from technical and marketing staff. The intention was to market the CD-i player for use in conjunction with the domestic television set, rather than as a complete unit with its own monitor. In addition to allowing for its introduction at the target price of $1,000, this design embodied the image of the product as a way of enhancing television, rather than as a new kind of home computer. The technical issue concerned how the connection to the television was to be made. Video recorders and satellite tuners use the aerial socket (RF), as do most home computers. Picture quality is improved if video signals can be connected directly to the TV circuitry without RF modulation, but this requires both sets of apparatus to be equipped with special sockets. Recently televisions have been sold with SCART sockets, but it was estimated that in 1990 only 10 per cent of televisions in British homes had SCART connectors. CD-i engineers wanted to maximise picture quality; CD-i marketers recognised that for consumers to have to buy a new TV set to play CD-i would constitute a barrier to adoption. In 1990 the issue was still unresolved, but launch plans suggested that CD-i players sold in Europe would not have an RF connector. Had this happened, ‘take it home and plug it in’ would not have been possible for the majority of British consumers in the crucial first few years. In the end, the marketers prevailed, and launch models did have RF connectors.
The hardware aspects of the design of the user interface were much less problematic, especially as the keyboard was ruled out almost from the beginning as a means of ‘interacting’ with CD-i. The technical constraint is that the user must have some means of moving a pointer across the TV screen and registering a choice. Graphical user interfaces on modern com-puters solve the same design problem through the use of a mouse or trackerball, and early discussions of CD-i included these as possible user interfaces. But the PC-user works at a desk close to the screen, where adjacent flat surfaces allow for a mouse or trackerball to be used. The first CD-i users would be using a television set, often in the living room, where there is no convenient surface for a mouse to work on. As we have seen, the infra-red remote control was chosen early on as the interface to consumer CD-i players. The design of the remote control was the same size as those used for TVs and VCRs, but included a small thumb-operated joystick which can be used to move the pointer on the screen; buttons are used to ‘click’ on the desired object. Once this decision was made, it remained only to settle such issues as size and complexity — as we have seen even the simplest version was considered by software developers as too complex (and small) for operation by young children. Design mock-ups available in 1990 suggested a relatively uncluttered model with, in addition to the joystick, buttons for stop, pause, play, fast forward and fast reverse. The launch model’s remote control was almost unchanged from these mock-ups.
A further user interface issue concerns the design of system software which controls what the viewers sees when they turn on the machine. The professional version, like a PC, reports on the configuration of the system and its status. The prototype consumer version ‘booted up’ to a menu which asks whether the user wants to play a CD-i disc, a CD-Audio disc, or perform some other function. The third alternative is argued to be necessary to cope with future enhancements — for example using the CD-i system as a controller for other equipment such as the VCR, the hi-fi system, the satellite tuner etc. It is not difficult to envisage the CD-i player becoming the graphical interface for coping with currently frustrating tasks such as setting the time-shift controls on a VCR. It is clear that the design of the appearance of the user interface will quickly evolve, but at the beginning it can be used as a way of identifying the technology through the use of the CD-i logo on the opening screen, and of course different manufacturers may wish to differentiate their products from each other by designing the opening screen in different ways. Given that there is little room for product differentiation through technological diversity within such a tightly defined standard as CD-i, the issue of differentiation through different designs of the user interface becomes critical for hardware manufacturers.
The evolving paradigm for the physical design of the consumer player has reflected the developing consensus of the ‘image’ of the product — principally that it should be distanced from the home computer. In the early 1980s, when the first target was hobbyists, home computers were sold as much on the basis of their technical specifications as on their potential applications. But experience with marketing, and the shakeout amongst manufacturers, firmly identified the importance of the software dimension, and in particular the importance of games as the key application. The lessons for CD-i were apparent: hide the technology, and stress ease of use, compatibility and the content of the software. But further: stress the broad educational and entertainment angle, and avoid the identification of the product with games playing. In part this reflected a view within Philips that games playing was not a ‘serious’ application for the technology, and in very early brainstorming sessions at Eindhoven the seriousness of purpose was emphasised by the shocked reaction to a suggestion that interactive soft pornography might help to open up the market for CD-i in the way that it had for VCRs.
The success of CD-Audio strengthened the view that CD-i should be designed to appear as a machine for home entertainment, like the CD-Audio player and the VCR, but not like the computer. Product mock-ups prepared by Philips’s industrial design centre were variations on the theme of a matt black case the size of a VCR, in keeping with the prevailing (in 1990) dominant design paradigm for audio-visual products including hi-fi. The front of the casing was to be kept as simple as possible with an LED panel and the minimum possible number of buttons. The dominant theme for the consumer is familiarity — a ‘natural’ extension of the consumer electronics products already present in every home. This ‘matt black’ orthodoxy seemed to be as strong for the consumer product as the grey (IBM colour) casing was for the professional model. The former was designed to be distanced from the image of a computer; the latter was designed to symbolise it.
The concept of the ‘professional image’ also extends to the means by which the disc is inserted into the player. Philips’s market research suggests that the shiny CD sold in a ‘jewel case’ is a very important part of the appeal of the format to consumers, and helps to justify the price differential between CDs and vinyl discs or tapes. As with CD-Audio, the consumer will put the disc in a tray on the machine. But the professional version uses a ‘caddy’ which transfers the disc to the machine without the disc being touched by the user. This is reckoned to protect the disc, since institutional users will not handle the discs with the same care as domestic users, but it also helps justify charging such users some 30 per cent more than mass market customers for the same product in a different case.
What is interesting about these design decisions is that they have developed
through market experience of what are considered to be paradigm products,
rather than through direct consumer research. It was the market
success of the infra-red remote control, rather than ergonomic research
with consumers, that led to the consensus on the user interface. Likewise
it was the market success of CD-Audio and the mixed fortunes of
the home computer which was the justification for settling on the black
box design. Future product design, or, rather, the evolution of product
design as market experience with early CD-i products becomes available,
is more likely to reflect market research amongst early purchasers. Tentative
product plans from both Philips and Sony include portable versions with
built-in screens, and the early focus on the living
room is likely to give way to distinct products for different locations
within the house (the study, the bedroom, the kitchen) as the marketing
strategy becomes more refined and feeds back more directly into the design
process (i.e. through actual sales rather than projected sales). Sony had
demonstrated in 1990 a prototype CD-i Walkman-type product with a 4 inch
colour LCD screen, and was already marketing a miniature CD-ROM player
(the Data Discman, later the ‘Electronic Book’) which uses 8 cm CD-ROM
The cardinal principle that the fate of CD-i as a consumer product will lie in the quality and mass appeal of the software is deeply embedded in the CD-i community. Evidence from the experience of other software-dependent products appears to confirm this, and reinforces the need for first-mover hardware manufacturers to become involved in software design and development. Market data from Japan indicate that the value of software sales outstrips hardware sales threefold for VCRs, sixfold for Nintendo computer games, and by ratios of 25:1 for CD-Audio and 30:1 for LaserDisc. Thus the issue for consumer electronics manufacturers is not only that software development is necessary to open up the market for hardware, but that a presence in the software market is a pre-requisite for obtaining a satisfactory return on investment from the new technology.
CD-i is posing far more complex problems for innovators than did the
VCR or CD-Audio. In those cases there was a readily available supply of
software in the form of feature films and sound recordings which could
easily be transferred to the new medium. By its nature multimedia involves
bringing together a range of skills locked into different professional
communities: programming and computer graphics skills from the computer
software industry; editorial and presentational skills from the book and
magazine publishing industry; and audio-visual skills from the film and
television industry. The key task is the integration of these different
processes, and the key constraint is the reluctance of any individual actor
to shoulder the risk of first entry into a new technology and a new market.
The dilemma is neatly summed up by the president of a small software house
The design issues which arise in software production for the mass market can be illustrated in the production of one of the first CD-i discs — Caesar’s World of Gambling — which was produced by a partnership between AIM and CD-i Systems Inc. of Los Angeles. The idea for the disc originated in the popularity of casino parties in the United States, where people meet in each others’ homes to play roulette, blackjack and other games for small stakes. The disc is designed to simulate the experience of gambling at one of the premier casinos in Las Vegas, and contains a mixture of actual photographs of Caesar’s Palace, together with computer animations of roulette, craps, blackjack and five different slot machines. The game rules and betting are identical to casino practice and the overall aim is to repro-duce the actual experience of gambling.
Visual appearance is assumed to be crucial to the success of the disc, and much of the work was done by doctoral students in graphic design at the University of California. Photographic colour slides were scanned, and then enhanced by computer manipulation in order to sharpen or soften edges, create textures and so on. One of the major limitations is the small size of the television screen: to make text readable at 10 feet takes up a considerable portion of the screen, which reduces the area available for visual effects. The interior of the casino cannot be shown in full, neither can the screen accommodate at one time all of the details of a particular game. The design solution was to make the screen a window on a larger visual plane, so that users can scroll as if they were using a camera to photograph the scene. The eventual solution was the product of successive trials.
The age range of expected users was 8 to 80, which meant accomm-odating poor reading skills as well as older players with limited experience of interactive video games. To overcome the former problem ‘the disc literally talks to the player.’ The user forces a response by positioning a cursor over a ‘hot spot’ and clicking the action button. If unsure, the user can click another button, and an oral description is given of what would happen if the action button were pressed. Ease of use is maximised if the hot spots are large, but accuracy of the visual representation is maximised if the hot spots are small. To emphasise once more the distancing of CD-i from computer conventions, there is to be no printed documentation; instead the disc contains a tutorial mode, but thereafter help messages have to be requested by the player.
Such design rules and conventions have been produced through a lengthy
trial and error process, as have the specific technical skills to implement
them. As the producer of the project emphasises:
Philips’s role has thus evolved into a first mover in hardware, in software
and in the production of software tools for the development of software.
It has played the major part in the construction of the infrastructure
of the emergent industry. But the recruitment of producers and the technical
process of design are only part of the process; an important part — perhaps
the important part — is the content of the software, i.e. what kinds
of applications are likely to prove commercially successful. Here the consensus
of the industry seems to be that market research is of limited use, and
market experience with precursors to the technology is a better guide.
The linchpins of the marketing strategy for CD-i have evolved from the bitter experience of the VCR format battles of the 1980s, and the success of the strategy for CD-Audio. The prime lesson was the need to establish, before any product launch, a world standard for the technology backed by the major consumer electronics producers in Japan in order to pre-empt alternative multimedia formats. As we have seen, delays in finalising the technology, and in particular the agreement of standards for the incorporation of FMV, have opened up opportunities for competitive challenges which are discussed below. The second requirement, which has also contributed to delays, was for an extensive launch catalogue of software. Finally falling hardware prices are expected to take the product well beyond the early adopters into the mass consumer market with a target of at least 20 per cent of homes possessing CD-i.
Yet as a radically innovative technology, CD-i poses very different
kinds of marketing problems to earlier audio-visual technologies. The unique
feature — interactivity — has to be explained to the consumer in
an easily understandable way in terms of familiar concepts, yet the extra
benefits which for the consumer will comprise the incentive to purchase
have to be stressed. In the pre-marketing phase this
led to some clear divergences of approach. For example, the Japanese approach
focused on simplicity:
Creating a professional community for CD-i proved to be a lengthy and
difficult process. AIM’s strategy was to interest large publishing and
media corporations in the US through the joint venture approach in the
expectation that smaller producers would be willing to risk their own capital
when they saw that the industry majors were committed to CD-i. In the United
States early participants included Time-Life, Rand McNally, Grolier, the
Smith-sonian Institution, ABC Television and Children’s Television Workshop
(producers of Sesame Street). In Europe Philips had lengthy negotiations
with Maxwell Communications, Bertelsmann, Hachette, Virgin, amongst others,
but had some difficulties initially in securing publicised commitments.
A year before the Maxwell empire crashed, Ian Maxwell cited a number of
factors as inhibiting the participation of publishers:
The issue is not only one of resources, however; it is also access to content, where major publishers and media corporations own material which is capable of being re-presented in CD-i format as interactive entertainment, interactive movies, interactive encyclopaedias and so on. For smaller companies the question of copyright looms large, since established legal notions of ‘film’, ‘book’, ‘play’ and so on will inevitably be subject to expensive challenge until new legal definitions are agreed which match the new technology.
What potential participants do seem to agree on is the importance of ‘trigger’ applications; that is, ones which successfully exploit both the technology and a market niche in a way which triggers hardware sales and breaks the perennial ‘chicken and egg’ problem of software-dependent technologies. Much-cited examples of this include the Visicalc spreadsheet (and its imitators) which helped to establish the concept of personal computing, and desk-top publishing which was responsible for the popularity of Apple’s Macintosh computer. In the early days at Philips the interactive encyclopaedia was seen as a potential trigger product, probably because it fitted well with Philips’s ‘high-minded’ view of the technology, and the fact that the total hardware and software costs of supplying an encyclopaedia in electronic form should be less than the cost of buying a multi-volume encyclopaedia in printed form. But experience suggests that trigger products are unpredictable, and the advice to Philips from its consultants was that it is unlikely that there will be a single trigger. Both the ‘triggers’ cited above come from the professional context where the software solution met a specific business problem, and it is difficult to think of a comparable case in the domestic market unless parallels can be drawn with the stimulus to the growth of the television industry in Britain in 1953 given by the Coronation, and to the VCR industry by the Royal Wedding in 1981. We should perhaps expect an interactive encyclopaedia of the royal family, which after recent marital scandals could include phone calls and paparazzi photos!
Producers recognise that the interactive element of CD-i is what differentiates it from existing audio-visual technology, but that in order to be understood by consumers it has to be experienced. The evidence that consumers will demand new interactive products such as CD-i, when the opportunity to experience it becomes available to them, came initially from a reading of past experience, especially with the remote control of television and CD-Audio. Later experience with prototypes and demonstration machines was claimed by Philips to have confirmed these expectations. Younger consumers were argued to have cut their teeth on interactive technologies such as computer games, but with a $1,000 product the marketing effort has at least initially to be directed towards their parents. Such factors account for the stress on demonstration in order to put across the specific characteristics of the technology. In Japan nine of the principal consumer electronics manufacturers collaborated in mounting a demonstration of CD-i at the International Garden and Greenery exhibition which opened on 1st April 1990 and attracted 8.5 million visitors in its first ten weeks. AIM and North American Philips announced plans for extensive consumer demonstrations of CD-i at shopping malls, where separate areas for children and their parents would afford hands-on experience for consumers. Philips IMS in the UK has similar plans for exhibitions of CD-i in shopping centres and other public places.
The current thinking about marketing CD-i and the identification of critical marketing issues can be seen in the marketing plans of Philips IMS (UK) and AIM.
The core of the UK approach was to prepare the ground for the mass market launch by concentrating initially on specialist interest groups and developing professional applications which could act as demonstration projects for con-sumers. The assumption was that the early adopters of CD-i would be middle class and middle-aged, but PIMS accept that market research will be necessary to identify target groups more closely. According to PIMS ‘The theme that characterises CD-i’s main properties is one of "SELF-ENHANCEMENT".’ This implies individual motivation to improve in a variety of fields.’ The tie-up between self-enhancement and the presumed middle class profile of the early adopter is common to the US and UK, but was not then backed up by market research.
Mail shots were planned to high-earning credit card holders (e.g. Gold Card and American Express holders) and to the members of wine societies, golf clubs, and music societies. For demonstration applications PIMS was seeking to introduce CD-i into what they call ‘Browsing/Captive Audience Environments’ such as airport lounges or stately homes. PIMS recognised that to some extent they were constrained by the software catalogue available, which comprised mainly American titles re-worked to suit European tastes. One of the titles developed by PolyGram was Wines of France, which could be targeted to wine societies; another was Rules of Golf in conjunction with the Royal and Ancient at St. Andrews, which together with Palm Springs Open developed by AIM/ABC Sports, was publicised to golf clubs. The possibilities for extending the effort towards other groups such as gardening clubs, anglers and so on was seen to depend on suitable software being developed.
PIMS identified CD-i as a ‘slow burn’ market, in which achieving early return on investment in software would be extremely difficult. Underlying this belief was the view that educating the ordinary consumer about the benefits of CD-i would take time, which meant that filling the shelves of multiple electrical retailers such as Dixons and Comet with CD-i players and advertising its existence on television would not be enough. Innovative products were argued to require innovative marketing strategies, which in an interesting throwback to the encyclopaedia salesman, might involve teams of door-to-door representatives demonstrating CD-i in the home. The PIMS UK marketing plan proposed that at first the distribution of CD-i should be confined to 100 or so specialist retailers who would be required to stock CD-i software as well as hardware, and offer adequate facilities for hands-on demonstrations. Only at a later stage would distribution be widened to the multiple retailers and television advertising be launched. The actual launch followed this plan quite closely.
PIMS was insistent that the marketing strategy could only be fine-tuned on the basis of market research in order to identify how the product should be ‘positioned’ in the home. This meant physical position (living room, bedroom etc.) as well as conceptual position (extension of the television, source of entertainment etc.). The research brief covered both hardware and software, including behavioural research on interactivity as well as more conventional marketing questions, including the social class and age profile of early adopters, use of other consumer electronics products, and pricing strategies.
This long-term market building approach was, however, based on the assumption that, at least in the early days, CD-i will have the field largely to itself. The announcement by Commodore in May 1990 that it was to launch a rival multimedia product for the home, CDTV, in September 1990 seems to have caught Philips by surprise, and was the first sign of a coming format battle along the lines of that fought out between Sony, JVC and Philips in the early 1980s. The advent of competition forced Philips to take its product into the High Street multiples earlier than it had hoped, and although it had seen off CDTV by 1993, a clutch of other competitors were in the market-place — including VIS and CD-ROM peripherals for home PCs — and others — such as 3DO’s Opera — were in the pipeline.
American interactive media
AIM was at least a year ahead of Philips IMS in defining its marketing strategy for CD-i, and by 1990 had already conducted an extensive programme of consumer focus group trials and retailer presentations. From this research AIM became convinced that changes in consumer habits and accompanying social trends have prepared a fertile ground for CD-i. The loss of audience of the major TV network channels in the US, and particularly the rapid growth within the video market of special interest videos (such as those featuring fitness, gardening and sports) was seen as part of a trend towards more variegated markets. In turn these were seen to reflect changes in family life, where there is less emphasis on the collective viewing of television programmes, and more on individual-centred activities, or what is known in US marketing jargon as ‘cocooning’. Furthermore, a growing concern with educational standards and self-improvement was judged to fuel the acceptance of appropriate technological solutions. Marketers cited parental reactions to ‘mindless’ computer games, or the resistance to ‘Nintendomania’, as an indication that the time was ripe for more serious applications of computer technology in the home.
Focus group research amongst early adopters of new consumer tech-nologies was claimed to be strongly positive, with few put off by the $1,000 selling price or the $20 - $40 range for CD-i titles. AIM insisted that in their tests consumers enthusiastically embraced interactivity, and found the tech-nology easy to use. The theoretical notion of ‘edutainment’ — that is, the blending of educational objectives and entertainment within the same programme which was among the very first conceptions of CD-i — was claimed to have produced a positive response in the trials. The research did confirm the importance of involving major publishers as co-producers: consumers were impressed by the involvement of major ‘brands’ such as Time-Life and ABC.
These and other findings were built into AIM’s basic marketing assumptions. The role of traditional advertising was seen as limited to supporting the primary strategy of generating credibility through public relations and demonstration events in high profile venues such as Disney’s EPCOT Center, hotels, museums and golf clubs. Crucial to this is again the notion of hands-on experience which cannot be communicated by advertising. Much was argued to depend on the initial positioning of CD-i as a consumer product and the need to avoid creating its image as an advanced form of video game. This could be done by giving it a family orientation, stressing that the diversity of available applications means that the various family members can pursue their individual interests through the same technology. The strategy also involved stressing that CD-i is an evolutionary and not a revolutionary technology, i.e. a logical step forward from the well known technologies of television and CD-Audio. Consistent with the earliest design principles, there was no place in this strategy for selling CD-i as a home computer, or for stressing the futuristic elements of the technology.
Advertising enters the marketing picture only ‘at one minute to midnight,’
although an advertising agency was involved in discussions leading up to
the decision to place CD-i as an enhanced form of television. Ironically
one of the ideas considered and then discarded was to rename CD-i as CDTV
— the initials chosen by Commodore for their competing product. A Philips
marketing manager described the role of the agency in the following way:
One of the most difficult decisions in preparing a new product for market introduction is how to time announcements to different groups. In the computer and consumer electronics industries, products are often announced during the development stage when it is known that rival products are being developed so as to pre-empt competition and persuade software developers and retailers, and sometimes consumers, to delay making any commitment to the rival. But this strategy runs the risk of premature disclosures leading to continual delays — and the sobriquet ‘vapourware’ — which makes acceptance of the technology more difficult when it does arise. Philips’s record in this respect is not a good one: CD-Video was announced before the product engineering phase was complete, which led to the postponement of the launch by a year and much adverse comment in the press. Philips’s announcement of CD-i at the first Microsoft CD-ROM conference in 1986, with a market launch predicted for 1987 led to accusations of ‘spoiling’ the emergent CD-ROM industry. Some commentators saw RCA’s demonstration of DVI at the following year’s CD-ROM conference as ‘Microsoft’s revenge’. The effect of these tactics was to dissuade many software developers from making investments in either technology until a clearer picture of the likely success of each was available. Instead of preparing the market through the steady development of a community of interest, successive postponements in the timetable bred cynicism. Philips’s stance was in sharp contrast to that of its partner Sony, which even long after the Philips launch had disclosed almost nothing about its own plans for CD-i.
By 1988 the seriousness of the mismanagement of the development process was recognised by senior management within Philips. Soon after his appointment in 1987 as head of Philips’s Consumer Electronics division, Jan Timmer ordered a thorough review of return on investment of all Philips’s consumer electronics activities. The HIS (Home Interactive Systems) group, apart from its responsibilities for CD-i, had a general brief to adapt products from the professional sector for the domestic market. It had chosen the ill-fated MSX standard for its home computer, and developed a personal word processor which failed to make any impact against Amstrad’s highly success-ful PCW machines. At this stage Philips considered abandoning CD-i, and it seems likely that this might have happened had not Philips’s Japanese partners continued to express confidence in the product. Instead Timmer chose to reorganise HIS completely, renaming it Philips IMS and completely changing its personnel. By March 1990 not one member of the original CD-i managerial team had survived the reorganisation. Timmer succeeded in persuading Philips’s main board to step up its investment in CD-i at a time when the company as a whole was looking for areas of retrenchment.
The most significant appointment was that of Gaston Bastiaens, who moved
from his post as Plant Manager of the Philips CD factory at Hasselt, Belgium,
to become Director of Philips IMS. One of his first decisions was to stop
making any further announcements about CD-i until a firm timetable for
product development could be drawn up, and this embargo
lasted almost two years until the new range of professional hardware and
development software tools was announced at the Microsoft CD-ROM conference
in San Francisco in February 1990. A major conference attracting 450 fee-paying
delegates was held in London in June 1990, which confirmed the extent of
interest in the new technology amongst the audio-visual and computer software
industries. Equally significantly Bastiaens reversed the earlier policy
of spurning potential professional markets for CD-i, and signed a major
contract with Renault to develop a CD-i system for staff training in Renault
dealers around the world.
The initial impact of CD-i on the market owed much to the range and quality of available software. Software was available at the launch  according in four categories: children’s, games, education and special interest, and entertainment.
We saw that despite the desire to distance CD-i from the image of a computer games machine, games software would comprise a significant part of the launch catalogue. Some of these titles, such as Dark Castle, were conversions from existing computer games, and in such cases promotional material would stress the enhancements that CD-i affords, especially hi-fi sound and later full-screen moving video. What is particularly interesting was the relatively small number of general entertainment titles, given Philips’s intention to stress the commonality between CD-i and existing consumer electronics entertainment products. None of the titles announced at the launch was specifically aimed at women, although children and men are clearly targeted, reflecting the marketers’ views about the likely characteristics of early adopters. At launch there was no sign of the interactive movie in which the producer offers a choice of storylines according to the user’s preference, but several such titles are under development, including some by a Philips subsidiary in collaboration with Twin Peaks director David Lynch. This comparative sloth may be due to the relatively late inclusion of full-motion video into CD-i, but even granted this, the movie industry was conspicuous by its absence among those reported in 1990 to be doing development work on CD-i.
PIMS UK recognised the absence of movie software, especially given the pivotal role that movies have played in the expansion of markets for VCR and satellite television. UK managers argued for the inclusion of linear digital movies (i.e. non-interactive films) which could be available on CD-i discs and playable through CD-i players as an alternative to viewing such films on VCR or LaserDisc. There was, however, resistance to this idea in Eindhoven, on the grounds that it would undermine the aim of identifying CD-i as enhanced television in which interactivity is the principal means of enhancement. Philips’s product development team did, however, design a prototype carousel CD-i player which would seamlessly join two or three 72-minute CD-i discs to accommodate cinema films.
Alongside the launch of CD-i titles, consumers would begin to see CD-Audio
discs appearing in record shops with the label ‘CD-i Ready’ on them. Philips’s
subsidiary PolyGram began to develop a range of music software which contains
text and graphics on the disc inaccessible until played on a CD-i player.
The sleeve notes would include an explanation of what CD-i is, and how
such information as the song lyrics, a discography of the group, biographies
of the musicians and so on can be explored through CD-i.
In consumer electronics technologies, standards have tended to emerge after a process of competition in the marketplace between rival products, with the market leader emerging as the de facto standards setter. Philips was very successful in the 1960s in promoting its ‘Compact Cassette’ technology for domestic tape recorders against the challenge of the American 8-track cartridge, but despite an early lead in VCR technology its attempt to promote its V2000 system as a standard was an expensive failure, almost leading to the withdrawal of Philips from consumer electronics in 1983. Its decision to co-operate with Sony to finalise the development of the Compact Disc was a direct result of this failure, and the alliance between the two firms successfully prevented any rival optical disc recording technology from finding an application in the home. Instead of trying to monopolise CD production, Philips and Sony followed the route that JVC had followed with its VHS videocassette technology by making licences freely available (at a price) and encouraging smaller manufacturers to market their own-label CD players bought in from other firms. It was not the route that Sony chose with its Betamax VCR system, and Sony, like Philips, were losers in that battle.
The mechanism by which Philips and Sony have ‘policed’ the standard has been to insist that licensees stick to the letter of the technical specifications of the system, as laid down for CD-Audio in the ‘Red Book’, CD-ROM in the ‘Yellow Book’ and CD-i in the ‘Green Book.’ For CD-Audio and CD-i the essential principle is that every disc should be playable on every player everywhere in the world. CD-ROM was seen by Philips as peripheral to its main business of consumer electronics, and its Yellow Book specification did not extend to file structures and file handling, which means that CD-ROM discs will not play on any CD-ROM player, although two incompatible de facto standards have emerged around the Multimedia PC (MPC) and Macintosh computers.
The International Standards Organisation’s (ISO) Motion Pictures Experts Group (MPEG) was expected to rule on standards in late 1990, with both DVI and CD-i under consideration, but in the end standards were not finalised until November 1992 — after the US and UK launches of CD-i. Broadband telecommunications also offers the potential for multimedia applications, and the CCITT are also expected to adopt standards in this area. The MPEG-1 standard adopted in the CD-i FMV chipset is broadly similar to the approach adopted by Philips, and is less costly to implement in silicon than DVI, which could spell the end for DVI.
The CD-i licence binds licensees to observe the Green Book standards, which ensure compatibility by laying down exactly how data is handled on the disc. Extra refinements can be added by software developers as long as they observe ‘base case’ rules. One important condition of the CD-i licence is the stipulation that developers must not mix DVI and CD-i on the same disc. Apart from legal action for breach of contract, Philips owns the patents on Adaptive Delta Pulse Code Modulation (ADPCM) — the technique used to record digital sound on Compact Discs. This proprietary control prevented Sony from realising its plan to market long-playing 3 inch CDs for its Walkman range which would have departed from the Red Book specifications. Another means of enforcing the CD-i Green Book standard lies in the proprietary ICs for full-motion video being developed by Philips, Matsushita and Motorola. At least initially Motorola will be the sole source of supply for FMV chips, and any intended departure from the rules could be pre-empted by the withholding of these chips — a far quicker and cheaper means of enforcement than legal action.
The major difficulty for CD-i standards lies not in the hardware, but
in the software. Here the issue is not technical standards but quality
control, and the potential damage that could be done to the market launch
by inferior quality software. Philips and its partners
lack any means for policing software standards, or for conveying ‘best
practice’ in software development:
As we have seen, collaboration between consumer electronics manufacturers was a crucial part of the development of CD-i technology. In addition to the pivotal agreement between Philips and Sony on hardware standards, Philips had entered into a string of joint ventures in Europe, the United States and Japan (see figure) in order to develop hardware and software tools and facilitate software production.
The scale of these joint ventures illustrates the extent to which CD-i as a multimedia technology involves areas of technical expertise (or in David Teece’s terminology discussed in Chapter 1, ‘complementary assets’) which even a major multinational electronics company does not possess in-house. In both Japan and the United States Philips teamed up with major print and electronic data publishers, Toppan and R.R. Donnelly respectively. The development of the prototype professional player was carried out by the joint venture with Kyocera, and the participation of Yamaha and the major record company Pony-Canyon (which has since forged an alliance with Virgin) in Japan Interactive Media was intended to compensate for the relative weakness of PolyGram in Japan compared to its position in Europe and the US. Studer AG is a major supplier of professional studio equipment — including CD-Audio players for broadcasting companies — and the joint venture was formed to produce hardware for CD-i studios. Sun Microsystems were chosen as a partner for the development of multimedia workstations because of their technological strength in high-speed microprocessor applications. Philips-Dupont Optical took over from Philips the management of its CD pressing facilities in Europe and the US. Finally the joint venture with Control Data Corporation gave Philips access to expertise in systems software.
The issue of inter-firm collaboration has become even more critical as delays occurred in the development timetable, and Philips needed to establish a credible programme for introducing the technology in the face of the competitive threat from DVI. After Intel acquired the rights to DVI from RCA’s parent company, General Electric, the prospect of collaboration between Intel, Microsoft and IBM became a major threat, especially after Intel announced its intention to enter the consumer market. It became important to extend the development coalition beyond the Philips/Sony axis, since the recruitment of additional Green Book licensees, whilst welcome, did not represent a major commitment of resources. In 1989 Matsushita announced its participation in with Philips in developing FMV for CD-i, and Motorola — the manufacturer of the 68000 microprocessor at the heart of the CD-i system — was chosen to make the VLSI chips for FMV. The addition of these major partners, especially Matsushita — the world’s largest consumer electronics manufacturer — was claimed by Philips as a major strategic advantage in its competition with rival technologies. Apart from NEC and Hitachi, all the other major Japanese consumer electronics companies were known to have developed CD-i players. At one time it seemed that the two outsiders to the CD-i ‘family’ would be persuaded to join the rival Intel-IBM-Microsoft grouping, but the latter never materialised, and Microsoft joined Tandy in developing VIS as a rival home interactive multimedia technology. As discussed below, by 1993 it was clear that the major competition for Philips would come from the cable TV and telecoms industries, who have quickly grasped the potential of MPEG-1 compression for interactive entertainment.
Thus we can see in the case of CD-i a good example of the tendency for competition in new technology products to take the form of a struggle between inter-firm groups, rather than between single companies, as was the case with colour television and VCR. Coalition building of this kind is partly dictated by market considerations (i.e. market power is critical in persuading retail chains to accept the product), but is also reinforced by the inability of major companies to put together the technological resources by themselves. This of course varies from company to company: Philips feared NEC and Hitachi in particular because of their strength in related technologies. NEC is the major player in the personal computer market in Japan, and is also well-placed in semiconductors, telecommunications and office systems. Hitachi has formidable expertise in VLSI design and semiconductor manufacture in addition to its market strength in consumer electronics. Neither company would need quite as extensive coalition of joint ventures and alliances as Philips has forged in order to mount a strong challenge to CD-i in the marketplace. Moreover as has been shown in recent years by the phenomenal growth (and sudden shrinkage) of Amstrad, major companies can be out-flanked by smaller competitors who can respond more quickly to changes in the market and, by Philips’s standards, achieve astonishingly short lead-times between new product ideas and their launch.
Joint ventures in software production fostered by AIM have, for the most part, been limited to the production of a single title. In the UK, however, a joint venture, SPIN UK Ltd, was formed between Philips and Shell UK to develop CD-i software specifically for the educational and training markets. Shell has had considerable experience of producing videocassettes for use in schools and LaserDisc-based software for in-house applications. Initially SPIN’s major task was to try to persuade the Department of Education and Science to fund a programme to equip secondary schools with CD-i players, and to this end have produced demonstration software for National Curriculum syllabuses in Science and Maths — the two subjects where teacher shortages were having the greatest impact. If successful SPIN had plans to market versions of the software to parents for use at home.
Looser forms of inter-firm collaboration have been employed to try to establish CD-i authoring studios in Europe. New Media is a small company which had been involved in Interactive Videodisc production, mainly for training applications. In 1986 when CD-i was first announced New Media had been working with Grolier on a version of its encyclopaedia as a follow-up to the Philips/BBC Domesday project. New Media produced the first demonstration CD-i disc in Europe, based on its work with Grolier, and the National Curriculum disc for SPIN UK. Despite its pioneering collaborative efforts as the only CD-i studio in the UK outside of Philips IMS, however, the relationship with Philips proved difficult and the company complained (in private) of a lack of consultation by Philips over the latter’s intentions for CD-i. Perhaps such problems are inevitable in the clash of cultures between small pioneering ventures such as New Media and large corporate bureaucracies like Philips.
But distrust is not confined to small companies. Despite the division of labour within the CD-i coalition which gave Philips responsibility for the refinement of the compression technology for full-motion video, Gaston Bastiaens was surprised to discover on a visit to Japan that Matsushita had been working on FMV compression in parallel to Philips’ efforts. When asked why this work was taking place, the Japanese replied ‘Insurance, Mr Bastiaens, insurance.’ In a striking example of intra-firm technology competition in parallel even to this parallel effort, the Matsushita subsidiary JVC has developed independently its own compression algorithm for FMV, which is a re-run of the VHS story when both JVC and Matsushita developed independently different video technologies from which the most promising version (JVC’s VHS system) was selected for commercialisation. For the Japanese firms collaboration seems to be not a substitute for market competition; rather, calculations of competitive advantage are constantly used to evaluate the benefits of collaboration.
Collaboration with Kodak: PhotoCD
Philips joined with Kodak in the final stages of the development of a standard for recording and retrieving very high quality photographic images on compact disc — aimed at both the professional and consumer markets. PhotoCD uses digital compression to store an image in 6 Mbytes, which gives approximately 100 images on a 12 cm disc. The images need not be recorded on the disc in a single session; the consumer can pay the retail photo store to add additional pictures over several sessions until the disc is full. The discs can then be played back on dedicated PhotoCD players (manufactured for Kodak by Philips) or on CD-i players, or on a PC attached to a CD-ROM drive using specific PhotoCD retrieval software. For professional markets — including publishing, medical and photofinishers — different software products are available for use on high speed computer workstations.
Some software developers have announced plans for commercial products which mix Photo-CD images together with text and sound, thus bringing PhotoCD software very close to some CD-i applications. For some educat-ional and reference material Photo-CD offers a very cheap means of develop-ing software, and the agreement between Philips and Kodak to keep the standards compatible means that such products can be sold to the owners of the growing number of CD-i players. Whether large numbers of consumers will buy dedicated Photo-CD players, which although they can play CD-Audio discs, cannot handle CD-i discs, remains to be seen.
The standardisation of the two systems is likely to be of considerable benefit to both Kodak and Philips in establishing CD-i/PhotoCD amongst the raft of competing products and technologies described in a later section of this chapter. The Philips/Kodak collaboration contrasts strongly with the awkward relationship between Philips and Sony, where Sony has gone its own way in sponsoring a rival multimedia format, MMCD (see below), which will not work with CD-i.
The theory of public goods ascribes to government the task of creating conditions which allows markets to flourish which no single market participant will provide. Whether it is constructing physical shelter for commodity markets, infrastructure such as roads and communications, or the legal basis for market exchange, the role of governments has historically been intertwined with the extension of markets. In many cases governments are involved directly in creating public markets for new technologies, especially in the defence sector. In consumer technologies governments have tended to reject any claims of strategic importance made by producers, and for the most part have declined to undertake either a defensive role when industries have been threatened, or a promotional role to reduce the cost or the risk of investment in new technology. Not all governments have followed this path; in France the state owns the biggest consumer electronics manufacturer Thomson, in Europe second only to Philips in size. Moreover, strategic arguments have recently been accepted by the EC and European govern-ments in the case of High Definition Television, where large amounts of governments’ money were used to develop the HD-MAC technology.
In the case of CD-i, Philips has acted as a quasi-governmental actor in terms of creating the infrastructural preconditions for the development of the market. It has subsidised firms to develop software, and created quasi-public resources in the form of studio facilities for smaller firms. The motive for such action has been essentially the same as the justification for governments to provide public goods, i.e. actual or potential market failure. Governments have not perceived CD-i as a strategic technology, despite the recognition that it offers some potential for training applications within the public sector.
Government support is, however, identified as a crucial precondition of the penetration of CD-i into the most significant public sector market, education. In Britain the DES played an important part in the diffusion of microcomputer technology into schools, and for a time offered subsidies to schools wanting to acquire the BBC’s interactive LaserVision Domesday package. Its experience with the latter, however, seems to account for the cautious reception given to suggestions from Philips that it launch a support programme for CD-i. Despite promises that the Domesday disc would be only the first of a range of educational software for the LaserVision system, only one other title appeared, and at £3,000 the hardware costs were too high for schools to afford. Although CD-i promises to be considerably less expensive, the DES has continued to fund development work for a Mathematics project using LaserVision rather than CD-i technology. It has, however made the concession to the CD-i lobby that the mastering of the discs will be done in such a way as to minimise the cost of later conversion to CD-i.
At the European level there has been interest but no firm commitment
to CD-i. EC officials have endorsed the potential of interactive multimedia
for educational and training purposes, and have recognised that its multi-language
capacity could help in the very long term process of harmonising educational
and training qualifications. But up to 1993 the work being done under the
Communities educational programmes like DELTA has been preparatory, and
the EC had not added a public dimension to Philips’s quasi-governmental
programme of infrastructure building, although it did fund some CD-i projects
under the ESPRIT programme. A PolyGram senior vice president expressed
the hopes of the CD-i coalition in the following terms:
Although targeted principally at consumer markets, there are important linkages between these and institutional markets which could help to diffuse the technology. In theory there may be a symbiotic relationship between the two, where penetration in the professional sector diffuses knowledge and skills which in turn help provide the preconditions for consumer acceptance, as happened especially in the United States with the PC. In the case of a product from a single supplier, the co-ordination of diffusion in the two markets is a problem for the firm’s management, and much depends on the internal organisation of the firm. The case study of CD-i reveals how organisational politics within the firm can make such links difficult to forge.
Within Philips CD-i was seen as the ‘property’ of the consumer elec-tronics
division which invented it, and HIS management resisted sharing it with
the Telecommunications and Data Services (TDS) division. The experience
with another technology which crossed the professional/consumer boundary
— LaserVision — also revealed divisions within the organisation:
It is clear that there will be a variety of competitors for CD-i in professional and consumer markets, although in the former each may find a niche in a way that is unlikely in the consumer market, where format battles will eventually produce a de facto standard. The players in each market segment will be different, with competition in the professional sector coming from multimedia PCs and PC add-ons, and in the consumer sector from the computer games and audio-visual industries. The following section reviews the competition according to the different multimedia technologies and the companies which support them. In addition to the products discussed below, which were on the market as of mid-1993, Apple and Toshiba announced in 1992 the co-development of ‘Sweetpea’ — a CD-ROM-based multimedia player which they hoped to market in the US for $750 by mid-1993, and 3DO demonstrated a prototype version of its ‘Opera’ multimedia player which was launched on the US market in late 1993.
Commodore’s rival to CD-i was a fully-fledged consumer product rather than a PC add-on, and was launched ahead of CD-i in 1991. It consisted of the innards of an Amiga computer together with CD-ROM player and interface presented in a black box with an infra-red remote controller. CDTV does not stand for Compact Disc Television, but Commodore Dynamic Total Vision, and it was described on the casing of the first machines as an ‘interactive graphics player.’ The player operates on a slightly amended form of the ISO 9660 standard for CD-ROM, which means CDTV discs cannot be played on other machines. The launch price shadowed CD-i plans closely at £699, later cut to a final selling price of £399. Less than six months after the product was launched in ‘black box’ guise as a television add-on, Commodore changed its strategy completely and re-positioned CDTV as a computer peripheral. The product was re-named Amiga CDTV, and was also made available as a (grey coloured) CD-ROM drive at £350 to plug into the Amiga. In June 1992 Commodore launched a ‘Multimedia Home Computer Pack’ which comprised a CDTV player, a keyboard, mouse and floppy disk drive at £499.
Software was a major factor in determining CDTV’s fate in the marketplace. Full-screen, full-motion video (FMV) was claimed to be available later as an add-on (but without much credibility, since unlike CD-i players, CDTV machines did not carry sockets for FMV chips). Much of the early software comprised converted computer games, which were already available to Amiga owners (at a lower price). Commodore, like Philips, was keen at first to emphasise the ‘serious’ potential of the system and even borrowed from Philips the ugly term ‘Edutainment’. The bundled ‘Welcome’ disc gave a sample of six programmes, which comprise material on the rain forests, sport, space flight, a trip round the Victoria and Albert Museum, a tour of an Egyptian pyramid and a biography of J.S. Bach. Commodore promised 100 titles by Christmas 1991, but in fact there were fewer than this on the market by the end of 1992. The promised range paralleled CD-i titles quite closely, with an encyclopaedia, a world atlas, interactive games and a variety of special interest programs on subjects like cookery. Although titles could be developed more quickly and cheaply for CDTV than CD-i, because of the availability of authoring tools for the Amiga, and the existence of software houses with Amiga experience, this tended to work against Commodore when relatively few high quality titles appeared amongst the offering.
The fate of CDTV, which was discontinued in the summer of 1993, supports the proposition that it is the appeal of the software which determines the fate of the format. The demise of CDTV reduced the number of competitors to CD-i, but news of yet another failed consumer electronics format may deter customers from buying CD-i unless they are reassured by Philips’s image as a blue chip supplier.
CD-ROM and the Multimedia PC
The Multimedia PC is a sub-set of the PC family agreed in 1990 by some of the major hardware and software firms in the computer industry (including Philips and Microsoft, but not IBM which brought out a higher-specification Ultimedia machine). The standard is based on a minimum level of PC (with at least a 386SX processor) equipped with a sound card, and linked to a CD-ROM drive with associated retrieval software. Sales of multimedia upgrade kits began to take off late in 1992, as did CD-ROM discs especially in the MPC format. CD-ROM has some limitations as an interactive multimedia technology, in that in cannot interleave (i.e. retrieve simultaneously from the disc) audio, text and images in the way that CD-i was designed to do. Clever software and software design, or adoption of the CD-ROM XA standard, can partially overcome this to enable CD-ROM and CD-i versions of the same software to be marketed. Compressed digital video can be added to CD-ROM discs, and decompressed through software (as in QuickTime and Indeo (see below) or through hardware chips (as in DVI and CD-i). The very large number of PCs in homes in the United States offers a large market for MPC upgrades, which have reduced in price to around $500 as cheaper CD-ROM drives became available. This solution also offers access to the Kodak/Philips PhotoCD system (see below). The expansion of the range of CD-ROM titles, and the reduction in price of multimedia peripherals for the PC, suggest that the MPC is beginning to become a competitor to CD-i for the home market as well as institutional markets. In 1993 negotiations were underway to upgrade the standard to accommodate faster processors and CD drives. The MPC-2 standard has as a minimum a 486 series processor, and includes provision for dual speed CD-ROM drives which allow for the much faster data rates needed for speedy image handling.
DVI - Intel, IBM and Microsoft
Intel has disclosed its intention of seeking both institutional and consumer markets for products based on its DVI chips, although by mid-1990 only one Canadian company had announced development work on a DVI-based game. The first version of the technology comprised two boards to be used with IBM AT or PS/2 machines or compatibles, for video capture and compress-ion, and for decompression and playback respectively. Moving video compressed in this way can then be used with CD-ROM sound, text and graphics to create the functional equivalent of CD-i. When first demonstrated in 1987 DVI comprised seven boards costing $25,000; by 1990 the cost for the two boards had been reduced to $5,000. The intention was to integrate the DVI chips and the CD-ROM interface onto the motherboard of a PC to offer a true multimedia computer for an additional cost of $500 by 1992, but so far DVI has appeared only on IBM’s high-end Ultimedia PCs. Video compression achieved by the capture board (‘Real Time Video’) gives a relatively poor quality picture; for final products ‘Presentation Level Video’ (PLV) is required which could only be done by Intel at a cost of $250 per minute of video. In fact DVI uses two different compression algorithms for the two levels of video; PLV uses a similar algorithm to CD-i. Alongside the chip sets Intel offer three sets of authoring tools, broadly similar to the different levels of the CD-i tools discussed above.
The chief advantage of DVI is that it can easily (but still relatively expensively) be incorporated into PC-based products, and thus take advantage of the installed base of PCs which might be upgraded. The extra hardware cost of the chipset remains a substantial deterrent, however, and Intel has countered this with a DVI-compatible software solution — Indeo — which works in conjunction with Microsoft’s Video for Windows in exactly the same way as Apple’s QuickTime, giving scaleable video in a window on the monitor screen.
Every Macintosh computer now comes with sound and system software for video compression (QuickTime), but despite price cutting, Macs remain relatively expensive for the mass consumer market. The high end Quadras became the preferred development platforms for multimedia producers, and the quality of the multimedia software tools from third party suppliers such as Macromedia remains ahead of Philips’s own tools for CD-i. Apple seems to be devising its multimedia strategy with institutional markets in mind, and is pinning much of its hopes on hand-held devices, beginning with the Newton which was introduced in 1993. That it had some designs on the home market for multimedia is evident in the luring of Gaston Bastiaens from Philips IMS in 1992 to head the consumer products division at Apple Headquarters, but internal problems led to his departure in 1994.
Fujitsu’s FM Towns and Marty
Fujitsu has had an unexpected success with a multimedia games machine in Japan — unexpected because it is primarily a mainframe computer and telecommunications equipment supplier with little previous experience of consumer markets. The FM Towns product consists of a 32-bit PC with integral CD-ROM drive and FM radio retailing (so far only in Japan) at £1,500. It spent $24 million on promoting FM Towns, primarily to high income families, and commissioned a range of software to include entertainment and instructional titles as well as games, much like CD-i and CDTV. By the end of 1991 sales were reported to have reached 160,000, when the machine was redesigned around the MPC standard including, in addition to its own operating system, MS-DOS and Windows. In early 1993 Fujitsu aimed squarely at the mass consumer market with an FM Towns-compatible derivative, the Marty, a £600 CD-ROM player which plugs directly into the TV set like CD-i and CDTV, and is capable of handling Sony’s Electronic Book format as well as CD-ROM discs. In Japan the launch was accompanied by a high profile television advertising campaign featuring a cartoon frog (Marty) which Fujitsu hoped would catch the public imagination.
NEC’s PC engine
NEC’s proprietary multimedia technology (the PC-Engine console launched in 1987, and the optional extra CD-ROM player introduced in 1988) was the first to use CD-ROM discs for software instead of integrated circuit cartridges. NEC was followed by Sega and, eventually, the market leader Nintendo. and had sold over 3.5 million consoles and 1 million CD-ROM drives in Japan by the end of 1992. The CD-ROM player cost $700, with software titles $20-$25 each. NEC has encouraged software development by marketing a set of authoring tools. NEC at one time controlled 80 per cent of the PC market in Japan, but has lost market share to American companies such as Apple, Compaq and Dell. NEC introduced a version of the PC Engine, called the Duo — known in the US as TurboGrafx — with built-in CD-ROM drive in late 1991.
JVC/Sega’s Mega Drive and Nintendo’s Super NES
Sega, based in Japan, is the main supplier of arcade video games. Following the success of the PC-Engine and FM Towns games players Sega introduced its own home video games player, the Mega Drive or Genesis, jointly developed with JVC, in December 1991. This was an add-on to its games console, and was marketed alongside a range of attachments, including alphanumeric and musical keyboards and a modem. The latter was used to offer a home banking service. JVC introduced its Sega-compatible Wonder-Mega player to the Japanese market, and also developed (with Philips) an interactive karaoke player, Digital Vision, with full-motion video, compatible with CD-i. The market leader in games consoles, Nintendo, was slow to follow suit into CD-ROM-based technology, but eventually launched its Super NES player in mid-1993.
Tandy’s Video Information System (VIS)
Tandy was a pioneer of the home computer in the United States, and is the largest retailer of consumer electronics in the US, under the Radio Shack name. The VIS system was announced in mid 1992 as a direct competitor to CD-i for the home interactive multimedia market. Tandy’s approach to product development, and its strategy for introducing the product to the market differs from Philips’s in several significant ways. Most importantly, while Philips had taken a ‘brown goods’ approach, seeing CD-i as value added television, Tandy has designed its product around existing personal computer technology (much as Commodore did with CDTV). The difference was that Tandy chose as its platform the IBM-standard PC, and worked with Microsoft to adapt the latter’s Windows operating system for VIS. The advantage of this was that software development costs were much lower than those for CD-i, and developers familiar with Windows products could quickly bring out VIS software using the same development tools they use for CD-ROM titles. Conversion to VIS from standard MPC CD-ROM software was inexpensive, and discs could be designed in such a way that they are playable on both MPC and VIS systems. By the end of 1992 around 100 titles, almost all conversions of existing software, were available on the US market. Tandy promised, but did not disclose how and when full-motion video and PhotoCD would be incorporated into the VIS format.
One design feature which differentiated VIS from CD-i was the incorporation in the former of a cartridge system which allowed users to record their responses, or positions within a game so that they could resume in the same place at the next session. Also unlike the CD-i consumer player, VIS embodied an ‘open architecture’ which was designed to permit VIS players to be connected to telephone lines or cable TV circuits. Philips considered, and rejected, such design features for consumer CD-i, but they were embodied in the CDI-602 professional player (but using a 3.5" floppy disk rather than a cartridge) and could feature in a future domestic CD-i player.
In conception and execution VIS was similar to the revised CDTV strategy, but with the advantage of being based on the dominant PC/Windows standard rather than the Amiga. It is too early to say whether the cheapness of producing VIS titles will lead to a flood of low quality titles, as happened with CDTV. The experience of Nintendo suggests that software quality control was critical to its success, and the same may be true for CD-i in that the investment required to produce a CD-i title tends to mean that the production values and standards are higher.
Sony’s electronic book concept: the Data Discman and the Bookman
Despite being a co-developer, and royalty recipient, of CD-i, Sony has done nothing to assist Philips in the development of an infrastructure for software development, nor — apparently — has it developed any CD-i software in-house. Sony’s strategy appears to be to wait for others to make the running on CD-i, and open up a market in Japan for its own Data Discman CD-ROM format — now re-named the Electronic Book. The standard for the EB was set in January 1990, with text and graphics displayed on a hand-held 8 cm CD-ROM player, and the product began to be sold in June 1990. In addition to Sony, Sanyo and Matsushita (Panasonic) have introduced EB players. Support of key Japanese print publishers has been obtained through the establishment of an Electronic Book Committee, and by the end of 1992 there were over 120 titles available in Japan. The original specifications were modified in 1992 to permit sound to accompany text and graphics; the EB-XA players use part of the CD-ROM XA standard with the same 8 cm discs as the original EB.
Despite an outline agreement between Philips and Sony to make the EB system compatible with CD-i, in the same way that Kodak and Philips did with Photo-CD, Sony has not introduced the promised hand-held full-specification CD-ROM XA player, instead launching a non-standard 12 cm disc Bookman, which will not play CD-i or CD-ROM XA discs. The machine uses a different graphics standard for its display from CD-ROM XA. The player is now called the Multimedia CD (MMCD), and has limited compatibility with the MPC standard, in that a modified operating system is available from Microsoft which will enable developers to make discs which will play on MMCD players and standard CD-ROM XA players attached to PCs. As so often, Sony has chosen an independent path.
This chapter has concentrated on CD-i — the most significant of the new interactive multimedia compact disc-based technologies aimed at the home market. In the early days, Philips assumed that the successful pre-market standardisation of CD-Audio, and the partnership forged with the major Japanese firms Sony and Matsushita, would ensure that, like CD-Audio, CD-i would have no competing format to inhibit its diffusion to the mass market. Contrary to expectations, however, the partnerships have proved disappointing, and Philips found itself with a format battle which opened up on many fronts. Not only was there direct competition for CD-i as a consumer electronics product — first CDTV and then VIS — there was competition in packaged media from CD games consoles and the multimedia home computer, and very soon there will be formidable competition from cable TV and telecommunications companies. The original idea of a single standard for multimedia CDs, so that as with CD-Audio, cassette tapes, and LPs, any disc or tape will play on any player anywhere in the world, has long disappeared. Instead, consumers looking at a rack of ‘multimedia CDs’ will face a confusing variety of formats and standards which will require some technical knowledge — at least on the part of the salesman — to ensure that the disc will actually play on the consumer’s equipment.
Philips launched CD-i in the United States in October 1992, and in the
UK the following Spring. One year after launch, and despite the appearance
of further competition, Philips were claiming encouraging early sales of
CD-i — at least as good as CD-Audio in the early years — which would mean
a combined sale of around 70,000 players in the US and UK. In addition
Philips have disclosed that each purchaser of hardware has also bought
six discs in the first year, which compares favourably with an average
of four in the first year of CD-Audio. Over the same
year the availability and sales of CD-ROM titles has expanded dramatically,
from 48 in 1987 to 817 in 1990. According to the publishers of the CD-ROM
Directory, there were 2,212 titles available in 1992. By the time of
the mid-1993 edition, the number had increased to 4,731.
The distribution of discs by format was as follows:
For Philips this would seem to be an encouraging beginning for CD-i, but it must be remembered that CD-Audio had no direct competitor, and for consumers the purchasing decision was relatively simple, since they did not need to learn about what recorded music has to offer, and the benefit obtained from eliminating hiss and scratches was evident. In the case of interactive multimedia none of these factors applies. Even if the product space survives the exaggerated claims made for it, the position facing the consumer will be more complex than for any other past consumer electronics product. The range of products competing within the multimedia product space will be considerable, and the number of products which will join CDTV in the category ‘obsolete’ will steadily grow.
The experience of home consumer electronics products suggests that this format battle will create by attrition a single de facto standard for the home player, but it does not mean the convergence of all consumer interactive multimedia around a single format. Just as the consumer now has a choice of watching a film on broadcast television, on cable television, on video, and now on CD, so will the consumer have a choice of formats for the same interactive multimedia software. There will most likely be a single ‘packaged’ format — and at the time of writing CD-i is the front runner in this race — and software ‘on demand’ via broadband fibre optic cable, whether operated by cable television companies of telecommunications service providers.
Digital compression technology, which made possible the introduction of full-screen full motion video to CD-based multimedia, has opened up a range of new opportunities for broadcasters, cable TV companies and telecoms network operators. The Hughes Corporation launched DirecTV in 1994 eventually to offer 150 channels of television on a DBS satellite. Somewhat later — perhaps two years later — cable TV operator Time Warner, and in competition an alliance of cable TV company TCI and US West (one of the former regional operating companies of the Bell/AT&T system) may launch rival multi-channel services which would have interactive capability. Not only will this mean that viewers can use a keypad to send back messages through the cable to take part in a live programme, it might open up interactive multimedia on demand in direct competition to packaged media such as CD-i and VIS. Time Warner, TCI and Sega are about to launch the Sega Channel, which will offer cable subscribers a range of over 100 Sega computer games if they already own a Sega Mega Drive games console and rent a decoder unit.
Over the past two years there has been a rush by computer companies, consumer electronics manufacturers, cable TV and telecoms companies and audio-visual entertainment companies to form alliances and joint ventures to position themselves for what they see as the inevitable arrival of multimedia in the home. The range of products and services promised adds up to a bewildering maze, and has added a new dimension of complexity to the innovation process described in this chapter.
Click for a full image
This figure represents the range of products and links between companies as of July 1993. Many of these products and services will fail; some at least may succeed in opening up a market for interactive multimedia hardware and software. The time it will take before such a market emerges, and perhaps a de facto standard is set for CD-based media, is likely to be lengthy as the new multimedia software industry learns to master the new technology. It took the motion picture industry several decades to do just that; we would be foolish to believe that a much more complex medium will arrive more quickly.
The experience of innovation in consumer electronics products would seem to confirm the emphasis that Philips has put on developing the infrastructure for software development in its strategy for the diffusion of CD-i. Initially Philips’s managers were naïve enough to think the product could be rushed into the market within months of finalising the technical specifications. When the complexity of the process was eventually grasped, and the necessary organisational changes made within the company, the effort shifted to putting in place what was required to bring into existence a CD-i software development community. The multi-disciplinary nature of multimedia meant that this community would have to combine people (and companies) from different backgrounds, and new skills would have to be developed. This is true of multimedia in general and not just CD-i, and even if the technology of CD-i is overtaken by further developments, and even if CD-i as a consumer product fails to win a mass market, the development of successive products within the interactive multimedia product space will bring with it a new multimedia publishing industry. As a medium of communication, interactive multimedia is so revolutionary that the relevant historical precedents are technologies such as cinema, radio and television, which created entirely new industries. The early moves towards developing multimedia reported in this chapter are like the first hesitant steps of the pioneers of those earlier industries. Unlike Hollywood at the turn of the last century, the multimedia industry at the turn of the next century will be dominated by global alliances forged between multinational electronics companies like Philips, Sony, Toshiba and Matsushita, and other IT and audio-visual producers. The ‘winner’ will have found a successful pathway through the multimedia maze.
3 Not least because Philips was suffering financially from the failure of its V2000 video recorder format against its Japanese rival VHS. The CD-Audio alliance brought together Sony and Philips, the two losers in the video wars.
4 Akio Morita, Sony's chairman, claims that Philips's original specification for CD-Audio was for 60 minutes' playing time, but Morita persuaded Philips to extend the playing time to 72 minutes so that his wife's favourite Beethoven symphony could be contained on a single disc. See A. Morita, Made in Japan, London: Collins, 1987.
7 For a discussion of the VCR format battle between Philips, Sony and JVC, see A. Cawson et al (1990), Hostile Brothers: Competition and Closure in the European Electronics Industry, Clarendon Press, Oxford, Chapter 13.
8 This is based on the 'normal' data rate of 150 Kbits/s, but Toshiba has improved on this by varying the speed at which the disc spins so that 'burst rates' of up to 300 Kbits/s can be achieved, bringing down to less than half a second the median access time, and making CD-ROMs nearly as fast as hard discs. Further innovation has now produced CD-ROM drives which can achieve data rates of up to 600 Kbit/s.
12 One early estimate was that 'it will probably take at least ten years just to produce one really good application.' M. Canter, E. Neumann and J. Fenton, 'Controlling CD-i: Languages and Authoring Systems', in S. Lambert and J. Sallis, eds. (1987), CD-i and Interactive Videodisc Technology, Sams, Indianapolis, p. 141.
13 The drive to create revenue from software development must have been reinforced by an internal review of return on investment in the consumer electronics division conducted by Philips in 1988 which showed a negative return on CD-Audio hardware despite the astonishing success of the compact disc. (Interview, Philips UK, October 1988). In 1990 the expectation in Philips was that the same would happen with CD-i players, but that unless Philips was committed to hardware development the whole technology would lose credibility in the eyes of third parties. (Interview, Philips Eindhoven, March 1990). Thus Philip surmised very early on that CD-i would be a 'razor blades' business, where profits would come from the blades (software) and not the razors (CD-i players), but you still had to make the razors.
14 This can be compared to the difference between the IBM and the Apple strategies in personal computing hardware. IBM deliberately adopted an open architecture to encourage others to manufacture to the standard (the so-called clones), whereas Apple sought (at least until 1993) to prosecute any infringement of its patents.
15 As we discussed in Chapter 2, there are signs of convergence between institutional and consumer markets for IT, and some products sell in large numbers into both markets, as in the case of PCs and PC-based multimedia CD-ROM systems. This market has been labelled the SoHo (Small Office/Home) market.
18 See Chapter 3 where Philips's conception of home automation is described. Later on Philips managers decided that CD-i was most likely to succeed independently of home systems, and indeed the whole home automation concept has been demoted within Philips corporate strategy.
22 L. Arpino in HIS Today, November 1985, p. 9. The research also gave rise to the one product launched before HIS was reorganised and home automation separated from it. That product was the Videowriter, a dedicated word processor which had its origins in the research finding that 50 per cent of home computer purchasers bought them for word processing, and a year later half of those were using their computers exclusively for word processing. This was essentially the same logic that underlay the launch of the Amstrad PCW computers.
25 A.D. Little were commissioned by Philips UK to prepare a study of the potential market for CD-i in 1988. Their report, Interactive Multimedia for Philips UK, was completed in June 1989, when they reported that the idea of a full CD-i plug-in board 'is (still) taboo' although it was under discussion (Vol. II, p. 28). Philips finally marketed a CD-i board for PC, Macintosh and Sun platforms in 1990.
28 To be fully accurate, the first software development products were based entirely on Philips's proprietary technology, and assumed that developers would forsake existing technology to author entirely on Philips CD-i equipment. An interview with an early software developer in Florida in February 1991 elicited the remark that 'We had serial number 001, but it never worked.' He pointed to a piece of equipment gathering dust in a corner, and said that they had quickly reverted to the Macintosh for software development.
31 CD-i was originally scheduled for launch onto the consumer market in Autumn 1987. The timetable set in 1990 was for CD-i to be launched in the United States in Spring 1991, in Japan late in 1991, and in Europe in 1992.
33 Critics in the trade and technical press accused Philips of muddying the CD concept by incorporating analogue video into what had hitherto been an all-digital medium, and for including larger size discs under the term 'compact disc' (CD). The move had reportedly been strongly opposed by Sony.
35 Interview with the Consumer Product Manager for CD-i, Eindhoven, 7 March 1990. This intention was confirmed by the Philips manager in overall charge of CD-i, Gaston Bastiaens, at the CD-i conference in London in June 1990. No such product has yet appeared.
36 The issue is not so pressing in other European countries, especially France, where SCART-equipped sets are more common. The US model will have RF output since almost no sets in the US have any other form of connector. US software developers have to work within these constraints - even more since the NTSC picture quality is inferior to PAL and SECAM.
37 The same stance was initially taken by Commodore with CDTV, but later Amiga was prefixed to CDTV, and the functionality of CDTV as a computer peripheral was stressed in an attempt to market the product to existing Amiga owners (of which there were about 1.4 million in the UK.
40 As reported in interview with one of his colleagues, the suggestion was made by Dick Fletcher, head of New Media, which produced early prototype CD-i software for Philips. In fact CD-i soft porn quickly appeared in the US, and there is now a proliferation of CD-ROM 'pin-up' discs. Philips responded by commissioning The Joy of Sex as a launch title for the FMV cartridge, and signed a deal with Playboy.
41 CDTV used a proprietary caddy system, which added significantly to the cost of a disc. Many consumers would buy fewer caddies than discs, which then involved a fiddly process of swapping discs between caddies.
42 Philips introduced a hand-held model with liquid crystal display in 1992, principally for professional users such as sales representatives. Sony's only model (as of mid-1993) was a similar handheld machine (the 'Intelligent Discman') aimed at professional users. Philips introduced a laptop CD-i player in early 1993 aimed at professional markets.
43 This of course need not be the strategy followed by subsequent entrants into hardware production, who could wait for the market to develop through software sales before competing against the first movers for the hardware market. This is thought (e.g. by Screen Digest, November 1992, p. 252) to be the strategy of Korean manufacturers.
45 Speech by Garry Hare, President of Fathom Pictures of Sausalito, Ca. at the CD-i conference, London, 19 June 1990. Hare was later recruited by Philips to lead the software development strategy for Europe.
47 According to Philips managers it was always expected that the bulk of early activity would be in the US, which reflects the greater interest and willingness to take risks of the American publishing and media industry.
51 Screen Digest estimates that it takes an average of 18 minutes to demonstrate CD-i to a potential customer, which creates some difficulties for retailers of fast-moving consumer electronics goods in stores such as Comet and Dixons. Philips claims that by September 1992, 100,000 people in the UK had witnessed CD-i demonstrations. Screen Digest, November 1992, p. 253.
54 By March 1990, Philips claimed that agreements have been reached with a number of 'major publishing concerns who like to keep things to themselves'. Interview, 6 March 1990. In June 1990 the only projects publicly announced were by Bertelsmann and Berlitz (part of the Maxwell group).
55 Ian Maxwell, Joint Managing Director of Maxwell Communication Corporation plc at CD-i conference, London, 19 June 1990. MCC subsequently secured backing from Philips for a joint venture, Maxwell Multimedia Corporation, which sunk with the Maxwell ship before any titles had been completed.
57 Interview. Screen Digest (December 1992, p. 279) quotes £150,000 to £200,000 as the average cost of producing a CD-i title, which is significantly higher than the average for PC or console games (£100,000), CDTV titles (£30,000-£40,000) and interactive videodisc software (£75,000). The 'typical' cost structure quoted by Philips IMS (UK) is as follows: design: 10-20 per cent; rights acquisition: 0-10 per cent; content creation: 10-60 per cent; content capture: 10-20 per cent; software engineering: 20-60 per cent; and testing: 5-10 per cent.
58 This estimate is based on interviews conducted in 1990. By the time of the delayed launch in 1991 the figure was much higher. Screen Digest (November 1992, p. 251) estimates a total worldwide spend by Philips/PolyGram on title development at $200 million.
59 Interactive Media for Philips UK, Vol. I, p. 64 suggested that encyclopaedia sales will follow hardware sales, and recommended a 'multi-hook' strategy (i.e. different approaches to different segments of the market) for boosting software sales.
60 Yet after an initial flurry of interest, Japanese electronics firms are playing a waiting game, with hardware ready to be released at speed if Philips succeeds in opening up a market. Philips alone launched CD-i in the Japanese market in 1992, and a year later its machines remained the only ones on sale in Japan, apart from Sony's miniature model. Sony seems to be putting its multimedia efforts into its Electronic Books, and perhaps in time its new magneto-optical digital audio technology - MiniDisc - will emerge in interactive multimedia form.
61 The information in the following section comes from the August 1989 draft marketing plan which was generously made available to us by Philips UK. AIM's strategy is based on the presentation by Richard Arroyo at the CD-i Conference, London, 19 June 1990.
62 In fact early purchasers included many more low-income households than Philips expected, thus following the diffusion of VCR more closely than CD-Audio. In retrospect, given the positioning of CD-i as a television add-on, this might have been predicted. Interview with Philips IMS marketing manager, November 1992.
64 In the first year consumers who bought players also bought an average of 6 titles; this compares favourably with the average of four CD-Audio discs bought in the first year by CD player purchasers. Screen Digest, November 1992, p. 253.
65 These claims must be treated with some caution since they were made in a presentation intended to boost acceptance of CD-i. They are presented here in order to show the assumptions underlying the CD-i marketing strategy.
66 Despite this, given the success of Nintendo games consoles in the US market, Philips felt it expedient to sign a deal with Nintendo to permit the conversion of Nintendo software to the CD-i format.
70 This list has been compiled from CD-i News, A.D. Little and AIM. It is not a definitive list of the titles available at the consumer launch, but is intended to show the spread of software in the different categories.
71 More recently there has been some evidence of links being forged between Hollywood and the interactive multimedia development community. Sony Entertainment - the parent company of Columbia Studios - has set up Sony Electronic Publishing, which includes an offshoot, Sony Imagesoft, specifically intended to produce multimedia software from Sony Entertainment films. Sega produced an interactive game version of Steven Spielberg's Jurassic Park, and Total Vision Inc. has acquired multimedia rights to JVC/Largo Entertainment's films. Screen Digest, December 1992, p. 276.
74 This principle was breached by Sony when they introduced the 3 inch CD single, and a range of Walkman-type CD portables which could only play 3 inch discs. Philips opposed the CD-single concept, and reportedly a compromise was reached when Sony dropped their objections to Philips plans for CD-Video, to which Sony had objected because it was not an all-digital format.
77 Many commentators attribute the success of Nintendo in computer games consoles (in taking off where Atari and others failed) to their strict quality control and insistence on manufacturing all games cartridges themselves. The demise of Commodore's rival CDTV system may have been, at least in part, due to the inferior software - often simple conversions of poor computer games software - sold for the machines.
80 See D. Thomas, Alan Sugar: The Amstrad Story, London: Random Century, 1990 for an account of Amstrad's growth. In their report to Philips IMS the consultants A.D. Little pointed out that the competitive position of CD-i would alter greatly if Intel were to join up with Amstrad to produce consumer DVI machines. Interactive Media for Philips UK, Vol. I, p. 94.
81 Interview with SPIN UK, 25 January 1990. The DES failed to adopt CD-i, and Shell lost interest and sold their share of the venture back to Philips, when it was re-named Philips Professional Products International (PPPI).
84 Older CD-ROM drives can only read the index recorded at the first session, but 'multi-session' drives have been introduced and quickly became the standard for new CD-ROM drives. By early 1993 inexpensive multi-session CD-ROM drives were available for less that £200, or slightly more when bundled with Kodak Photo-CD Access software.
88 One of the difficulties of creating consumer awareness of the technology through retail point-of-information applications has been the insistence of the major multiples on making the technology 'transparent'. In the United States there is more competition in the retail sector, and Philips has been able to insist that opening screens announce that 'this is CD-i'.
90 One contender never reached the market at all. Mattel, a major player in the US toy industry, bought a licence to develop Compact VideoDisc (CVD) technology from its developer, SOCS Research Inc. of California. Its intention was to launch a $500 games machine on the US market in 1989, but failed to achieve this and the company abandoned work on CVD technology. Licences for CVD were also taken out by Hewlett Packard and Lockheed, but neither company divulged any product plans, and the technology has disappeared from view.
91 3DO is part-owned by Matsushita, one of the 'core' CD-i firms, and is yet another example of Matsushita hedging its bets on what format will succeed in the future. Matsushita also produces players to Sony's Electronic Book format.