As Kodak rolls out Photo CD and Tandy introduces the Video Information System, the CD industry grows up

When Sony and Philips developed the compact audio disc, they had one purpose in mind: to create a new medium that would deliver the highest quality stereo sound, unprecedented dynamic range, and freedom from the background noise that plagues tapes and records. Taking advantage of existing technology, the Sony engineers did most of the work in less than nine months.

The development team never thought about using the CD as a computer storage device — or indeed, about using it for anything other than music. Much of the technical specifications seem to have been chosen by default, not forethought.

The sampling rate (44.1 khz) was chosen because Sony had digital tape decks (built around video recorder components) that already sampled at this rate. The size of the disc was chosen so that it would be as small as possible, but still large enough to hold Herbert Von Karajan’s interpretation of Beethoven’s Ninth Symphony. “Random access” meant being able to access individual tracks within a few seconds.

However, as soon as audio CDs became successful, people began to look for ways to use them as an inexpensive computer storage medium. Since the objective was to piggyback the massive scale of production for both CD readers and the discs themselves, the computer industry “adapters” could not change any of the physical characteristics of the discs. The data storage capabilities and the data transfer rates were locked in stone. What they could (and did) change were the ways in which information is recorded on the disc.

The most important difference is in the allowable error rate. A one-bit error every few seconds is unnoticeable in a musical passage, but it would render computer data utterly unreliable. The mathematics of error-correcting codes had already been well developed, largely for the telecommunications industry; by using sufficient redundancy in the code, any desired level of reliability could be achieved. So the computer inventors cheerfully sacrificed some of the compact disc’s vast storage capacity for error-proofing, and some more for the directories and file-header structures that a computer needs to make sense of any data.

Thus was born compact disc – read only memory, or CD-ROM. Its designers, like the earlier Sony engineers, had one goal in mind: a cheap medium for distributing software and data files. With strong leadership from the dominant player in the personal computer industry, Microsoft, the data format was standardized in short order (well, relative to other computer standards) and CD-ROM’s acceptance seemed assured.

One good turn … But like crabgrass in the flower bed, further refinements of the basic format sprang up almost overnight. The fundamental problem was that different constituencies were trying to adapt CD technology to different purposes. The mainstream of the CD-ROM “industry” has been (and remains) distribution of text databases. Most of the hype, however, has focused on sexier “multimedia” applications that require the user be presented with text, sound and motion concurrently.

The Holy Grail for many people in the industry has been to sell this technology into the home, either in the form of a home personal computer that will support CD-ROM-based multimedia, or via CD-ROM “players,” which will hook up to the consumer’s television set.

A hodge-podge. So different groups, with different priorities and objectives, began to develop new ideas for how best to use this cheap storage medium. Their legacy is a hodge-podge of competing and incompatible flavors and formats of CD-ROM. Each new format allowed the basic digital medium to address some new niche:

• The original High Sierra standard (which became ISO 9660) was designed with the ms-dos file system in mind. (No surprise there, given Bill Gates’s level of involvement.) Independently, Apple adapted the basic data layout to support the Macintosh filing system, HFS. Not to be outdone, the Unix community had developed its own format to support the Unix file system. Fortunately, the Unix and Mac software architectures are flexible enough to accommodate foreign file systems such as 9660 by simply installing a software driver. The principal consequence has been to make extra work for publishers who want their content to be readable across platforms.

• Microsoft, naturally, wanted the home multimedia player to be based on Microsoft technology. For a long time, it saw CD-ROM-based multimedia as a means of getting a (Microsoft) PC into every home. However, the enormous diversity of PC peripherals and configurations has not provided a single, consistent “target platform” for people interested in developing multimedia titles. Working with Tandy and other hardware manufacturers, Microsoft attempted to remedy this by coming up with a standard MPC (Multimedia Personal Computer) specification, which required a Windows/MPC-compatible CD-ROM drive.

• CD-ROM/XA (extended architecture) enabled the interleaving of CD-audio segments with computer code segments on the same track, allowing continuous sound within an animation sequence of a game program, for example. Currently, most CD-ROM players automatically recognize the difference between audio and computer data, steering each to its appropriate jacks on the back of the cabinet — stereo output or serial connector.

• Philips’ CD-I used the basic CD-ROM data format. But its developers wanted a consumer-entertainment appliance for animations, games and “info-tainment.” They chose a nonstandard operating system and processor chip, which meant that CD-I software will not run on any personal computer — and vice versa.

• Philips, however, has been the first to demonstrate high-quality, full-motion video playback on enhanced CD-I players. This may give it an important leg-up in the interactive multimedia player market. (See Vol. 2, No. 1, p. 17, for more on Philips’ recent announcements.)

• Kodak’s Photo CD makes just one wee change to the basic CD-ROM/XA format: it allows the disc to be updated by adding new information at a later time. The trouble is, the controller chip inside most current CD-ROM players will refuse to read past the information that was written in the first recording session. (Remember, CD-ROM was designed for bulk distribution, with all the information gathered and organized beforehand.) “Multisession” drives without this restriction are just now coming into the market. Some older XA drives can be “upgraded” with the addition of new driver software.

• There are a host of other formats, including CD+G, CD+MIDI, CDTV, the proprietary formats for the Sega and Nintendo game machines, and others.

This summer saw two significant events that are evidence of the growing maturity of the interactive CD-ROM industry: Kodak launched Photo CD into the consumer market and announced significant commercial enhancements to the format. And Tandy introduced a direct competitor to the Philips CD-I home multimedia system called Video Information System, or VIS.

The strategies and targets of these two companies are quite different, but each is displaying a savvy about the technologies it is developing and the markets it is trying to conquer (see related stories, p. 5 and 14).

The industry grows up. After 10 years of this mess of initials, the industry that developed around the compact disc seems to be maturing. Either through market acceptance or sheer will, certain CD formats are solidifying their positions, while others have fallen by the wayside. For their part, the survivors are increasingly willing to acknowledge benefits of alternate formats and to work through compatibility issues.

That doesn’t mean you can now use every CD technology with your Mac or PC or in every CD player. There is still the small matter of software that’s needed to make sense of incoming data. But the incompatibilities among the many CD-based media are slowly resolving themselves. While some platforms, like the CD-based game machines, are intended to be single function, most CD-based devices, including the next generation of “plain old” CD audio players, will soon be able to handle more than one format. In most cases, the incompatibilities can be overcome if it’s worth someone’s while to write the code.

AFTER A THOUSAND FLOWERS BLOOM

Technology seems to move in waves; first there is a rush of inventions, none of which are quite compatible with the others, as people explore and exploit the potential of a new medium. Later comes the consolidation, as people figure out how to make one gadget handle all the various alternatives (and as the less-popular inventions are ignored and die off). CD-ROM seems to be entering that phase now.

The CD-ROM as an interactive, mass-storage medium had two challenges facing it from the start: First, it was never intended to do anything other than deliver sequential music through a home sound system. Once everyone in the computer industry decided that the cost savings inherent in piggybacking a mass-produced consumer product were too irresistible to pass up, CD-ROM had to overcome the market fragmentation that occurs when a technology is developed by a huge number of competing interests over the course of many years. The latter issue now seems to be resolving itself, and compression technologies are, for the moment, compensating for the former.

The CD world has matured. Only one invention, the blue laser (which, because of its finer beam, would be able to pack much more data onto a disc), is even visible on the horizon. Several companies will offer double-speed drives that will spin the disc at twice the speed — and thus double the rate at which data comes off the disc (and halve the playing time), but this represents a tweak rather than a fundamental change. Other companies are trying to develop magneto-optical technology to ultimately replace the CD-ROM.

All of these represent only distant threats; the whole point of CD is low-cost access to content, and blue lasers or faster drives aren’t going to be cheap for a long time. In the meantime, the “good ol’” CD-ROM is finally starting to come into its own.

Peter Dyson, Jonathan Seybold, David Baron