Despite serious errors in judgment, it’s too soon to tell

Jan Timmer, chairman of Philips Industries, certainly cannot be faulted for lack of vision. Though most of us can’t hope for more than one big, good idea in a lifetime, the man who made his career launching the compact disc revolution had Vision No. 2 more than six years ago. He saw great potential for a CD-based consumer electronics device that would bring interactive programming into the home and turn people’s attentions away from the miserable offerings of the ubiquitous idiot box.

Wagging tongues. As a result, Compact Disc-Interactive (CD-I) is today a well-worn phrase that industry wags roll off their tongues with a knowing, haughty chuckle. But like it or not, CD-I was the buzz of the Consumer Electronics Show in Chicago, and Philips Industries–purveyor of the product that’s supposed to revolutionize consumer electronics–has promised that the first CD-I retail units will finally be shipped in October 1991.

The easy bet to place is that CD-I will be more fizzle than sizzle. Philips and its CD-I software subsidiary, American Interactive Media (AIM), have asked us all to suspend our disbelief about CD-I for longer than most of us have been willing to bear. Though initially excited, many of us started feeling burned by the endless hype.

Not holding our breath. The specifications were first announced five years ago, in May 1986, at Microsoft’s International CD-ROM Conference in Seattle. CD-I was to be released in 1987. But it wasn’t–the first fully functional version of the spec wasn’t even ready until November of 1988–and still hasn’t been. Philips also said CD-I would cost less than $1,000, the magical price point for success in the home market. Today its list price is $1,400. It was also going to have full-motion, full-screen digital video capabilities built in, but now digital video will be optional–at a price still undisclosed by Philips.

Many people who’ve been holding their breath for this product finally turned blue and passed out. When they came to, they started looking elsewhere for the Multimedia Player That Would Change the World.

It’s not unusual. Bernie Luskin, president of AIM in Los Angeles, insists that CD-I isn’t late–that perception, he says, is a result of “uninformed expectations.”

In fairness, this lengthy a gestation period is hardly unusual for a new hardware platform, even one that’s based on an old, weird chip (a 16-bit, 10-MHz Signetics 68070, a superset of the 8-bit Motorola 6809), 1 MB of RAM and CD-RTOS (for CD Real-Time Operating System), also old and proprietary, and first developed about a decade ago by Microware of Des Moines, Iowa.

The hangups mostly hinged on the very new problems of integrating media types–video, graphics and sound–and getting them to run in some kind of synchronized fashion via a CD drive’s pathetic data rate. Encoding schemes for color, video and audio information were known prior to CD-I, but there was no custom silicon to implement them. Philips had to do that as well–a fairly rigorous piece of work that stole a hefty chunk of time in the process.

Kvetch away, but it is a standard
So today everybody is kvetching about CD-I. It’s too late, it’s too expensive, it doesn’t “do” full-motion video, it’s slow, it’s hard to develop for, the titles are boring, and Nintendo’s already in the home.

It’s heeeere. But Timmer’s idea was to do one thing that everyone says this industry needs, and that Philips has proven it does exceptionally well–set a standard, and an international standard at that, for an interactive device that would play any standard CD made anywhere in the world.

Philips did this by devising a specification now known as the Green Book, with Sony’s collaboration (though Matsushita was along for the ride), which went much farther than simply mapping out the “I” in CD-I. It was constructed so that all of today’s standard CDs worldwide would be able to play in the CD-I box–everything from audio CDs to CD-based games to CD-ROM/XA and, more recently, Kodak’s new Photo CD.

The box was thus designed as a component, and anyone who wanted to could get a Green Book and build one, thus helping “commoditize” the technology. In fact, many Japanese companies have CD-I players in development.

A little worrisome. One thing of concern, however, is Philips’s MPEG chip deal with C-Cube Microsystems, announced at CES, to provide the MPEG full-screen, full-motion video decompression module for CD-I. (Without it, CD-I can only display a small portion of a screen of video at less than the requisite 30 frames per second.)

Philips won’t talk about how much it will cost–a big concern to developers who need consumers to make an extra purchase to take advantage of full-mo. “It has to be reasonable,” Jerry Calabrese, VP of CD-I marketing for Philips, acknowledges. “We’re not going to do something dumb. We’ve got a lot invested in CD-I.” But many think it’s a big mistake to ship the first CD-I players without it.

In addition, C-Cube’s first big design win, in the Next Computer System, backfired when C-Cube couldn’t satisfy Steve Jobs with a stable enough chip. It’s doubtful, given the difficulty of getting robust decompression onto a chip, that Philips will be able to meet its fourth-quarter ship date for the video module.

Stellar design. Often overlooked in the feeding frenzy of criticism is the design of the CD-I box itself, at least the Magnavox-labeled player Philips was demonstrating at CES. Making it the same size and color (black) as a standard stereo component was obviously important, but the control device on a player with “interactive” in its name was key.

Since remote devices are today the closest that consumers get to “interactivity” with their TVs, such controls for interactive players are a critical juncture point. The control device shown with the Magnavox CD-I player is elegant, small and easy to hang onto.

As seen in the photo at left, the remote uses a small joystick to maneuver the cursor around the screen. Buttons surrounding it function as selectors; switches for TV and CD-I, play, pause, stop and volume control are absolutely obvious.

They’ve got the fever. Another critical piece of the puzzle, which Timmer et al. learned from CD audio, was the importance of titles.

When Philips decided CD audio was the greatest thing since sliced bread, it was not a far reach for it to tap its PolyGram music label and say, “Convert your archives to CD; let us worry about marketing them.” Skeptical audiophiles, faced with a massive library of titles on a new medium, weren’t hard to convince after they gave a listen.

But there is no equivalent of the audio fanatic, or a massive library of anything, in the interactive media business, so it was up to Philips to create a new titles market from scratch. It attempted to do this by founding AIM (and sister companies in Europe and Japan), chartered to build a titles base for CD-I. What Philips did right by founding AIM, it also did wrong by how it set up the organization.

Where AIM went awry
It’s a nasty business, this pioneering. Through no real fault of its own, other than impatience to start the ball rolling, AIM started pitching the CD-I platform to developers without any real tools to help them. Horror stories abound; many developers attest to having spent most of their time (with money from AIM keeping their companies afloat) creating the tools to create the titles.

Every producer I spoke with said that until about six months ago, this was absolutely true, though Luskin insists it isn’t so. What changed in that period of time was the availability of a new compiler from Microware, and the acquisition of a tools company called OptImage (more on them later).

Artistic control. In addition to a dearth of tools, AIM has suffered (and suffers) from its demand for a high degree of control over the end result. Scripts and projects were reviewed to the nth degree by all levels of AIM bureaucracy (which one wag said made the French government seem like a paragon of efficiency); as a result, many producers found their projects unconscionably delayed by anywhere from one to two years.

In addition, its boilerplate contract actually gave AIM the right to pull a project at any time and give it to another developer, the equivalent of taking Super Mario Brothers from Nintendo and letting Sega finish it. Luskin says this degree of control was necessary with many developers who had no experience creating image-and-sound-based applications, and was not the case with “key developers we felt we could trust.”

But even experienced AIM-funded developers often found themselves and the minute details of their projects under intense scrutiny from many layers of AIM bureaucracy. Some developers indeed left the fold, disgusted; others had their projects pulled. Luskin says AIM only rejected “inadequately or ineptly presented” proposals, creating what he called “consternation” among some in the developer community.

In perpetuity throughout the universe. Other clauses in AIM’s boilerplate are equally less than inspirational, requiring that developers relinquish all optical-disc publishing rights to AIM “in perpetuity throughout the universe”–meaning that AIM need never pay a dime of royalties on any CD-based version of a title, whether 100 or 100,000 are sold.

Of course, such clauses are subject to debate and change, but inside sources confirm that many small developers were either too naive or too broke to invest $5,000 or $6,000 in hiring a good entertainment rights attorney to negotiate their contracts.

Not all developers think AIM cuts a bad deal. Luskin estimates an average title probably costs between $200,000 and $300,000; developers say it’s hard to get outside investment money for an unproven platform.

“I certainly think AIM right now is one of the only games in town,” says one. “When another company comes in and says, ‘I’ll spend $100 million per year developing optical media stuff,’ things will change pretty quickly. But good developers who aren’t happy about the ownership of (royalty) points are being silly.”

That scenario, by the way, may not be far off: rumors abound that Japanese consumer giant JVC is chatting up CD-I title developers.

You can’t buy love. Luskin says AIM’s up-front investment in titles has already “significantly improved” the genre. In any case, some developers who are signing on with AIM or continuing to develop products for the company are using its money as an investment in the learning curve for future products they can develop without AIM’s financial support. Keep this in mind when you see the first generation of CD-I titles.

What AIM’s doing right
Absolute dedication to the cause. Despite setback after setback, Philips and AIM continue to spend massive amounts of money–the conservative estimate is $250 million–on titles development, and they are moving forward to rev the hardware.

Tools. Part of that investment has been to correct a critical weak point: tools. To that end, Philips recently purchased a company called OptImage, based near Microware in Des Moines.

OptImage’s Balboa developer toolkit is a complete development package for standard-issue titles (though developers pushing the envelope, as they say, still find themselves doing some custom hacking).

Cinemaware and Garry Hare. In addition, AIM realized the long-term value to bringing some “names” in-house. AIM recently scored by hiring David Riordan, maker of fine interactive movie products under the auspices of his old company Cinemaware, after Cinemaware dissolved. He now runs an AIM division called Point of View and is busy building new CD-I titles.

Garry Hare, president of Fathom Pictures and another highly regarded developer, was recently made managing director of European Interactive Media, AIM’s sister company, and will facilitate the development and distribution of titles overseas.

Nintendo, too. Philips also recently announced a surprise agreement with video game giant Nintendo, whereby Nintendo has granted Philips a license to develop and market video games on CD for Nintendo’s new Super Famicon and Super NES 16-bit home video game hardware. They’ll work together to develop a unique format for Nintendo, based on CD-ROM/XA.

With Philips’s “backward compatibility” strategy, this means the Nintendo CD-ROM/XA games will also play on CD-I. AIM also licensed Nintendo characters such as the Super Mario Brothers and will develop and market CD-I titles based on them. This was a smart move–video games are an obvious impetus for a multipurpose home electronics purchase.

Titles. For those who believe that a massive archive of titles is what will drive the market, AIM has funded interactive products in subject categories from art to gardening to music lessons to children’s educational entertainment. “Big name” developers include Rand McNally, Time-Life Books, The Smithsonian Institute, Children’s Television Workshop and ABC’s Wide World of Golf. Most people agree that 50 titles will be ready when CD-I is released in the fourth quarter.

Too soon to say
Those of us on the inside looking out may be too jaded to know a good title when we see one, though I saw nothing in the Philips display at CES that would make me run out and buy a CD-I player in October.

But during a panel called “Where are the markets?” at last month’s Digital World conference in Beverly Hills, there was general consensus that the “killer application” was a windmill that was probably best left for the computer industry to tilt at, or for those who have lucre-filled fantasies of some player, somewhere, that could repeat the (relatively) instant success of CD audio.

Blind men and elephants. Larry Lowe, a former applications design engineer at AIM, says public perception of CD-I is like the fable of the seven blind men and the elephant: each believes that whatever he grabs–the tail, the ears, the leg–defines the beast. In the same way, the variety of CD-I products makes it impossible to pigeonhole, or to forecast its future.

There are two obvious bottom lines. One is whether dealers will believe in the concept and/or understand it enough to stock CD-I players on their shelves. (By CES, more than 2,000 dealers had agreed to stock CD-I players and titles.) Even more important is whether consumers will believe there is some added value in being able to “talk back” to their televisions. For the sake of everyone who wants this business to move ahead, we’d better hope the answer is “yes.”

But it’s vital to keep in mind that CD-I will not be an immediate, runaway success. Many factors, not the least of which is the fact that consumers don’t have a reference point when it comes to interactivity, will contribute to what’s bound to be slow growth.

Will Philips last until CD-I-2? Then there’s the question of whether the lack of motion video decompression and titles that, to date, are underwhelming will be fatal to CD-I. Will there be enough CD-I fanatics, as there were Macintosh fanatics in the early days, to carry CD-I through on word of mouth until video and good titles appear? And will Philips’s financial staying power prevail until CD-I-2? It’s too soon to say.

Even if CD-I bombs, however, it won’t be the end of the world for interactive technology. It will succeed sooner or later, with either this generation of products or the one after. The challenge is to try to pinpoint what’s required to make interactive products compelling enough to move them into the mainstream. And that is what we’re waiting for.

Denise Caruso

THE WRONG TECHNOLOGY
We can do better than CD-I

The first-generation CD-I will not have more than very modest success in the market. They are expensive, and consumers will not find the titles compelling. It is possible that second- or third-generation players with full-screen, full-motion video and lower prices will be much more successful, though I am not willing to bet on it.

From a technological point of view, it would be a shame if CD-I were successful. The technology is based on an oddball and obsolete processor, a godawful operating system and too many proprietary bits.

True, the hardware could be replaced with more modern, more open hardware, but the legacy of the original hardware and the original operating system will handicap CD-I forever.

We would be much better served if this kind of product were based on computer hardware and software that is shared with the mainstream computer industry. This would allow consumer products to benefit from the title-development tools available in the computer industry, and both industries could benefit from common advances in hardware and software technology.

In the end, despite Philips’s commitment to the project, and despite support from many Japanese companies, I will be surprised if CD-I is the next hot consumer product. The technology is not quite ready. The market is not ready. Besides, technically speaking, we can do a lot better.

- Jonathan Seybold