Seybold Digital World Conference

European Multimedia Bulletin

London, 1991


Seybold held their second Digital World conference June 4-6 in Beverly Hills, California, USA. About 600 attendees participated in a single track of conference sessions which ran most days from 8:30am to 9pm, while another 1600 visited the Demo Center (manufacturer's exhibits and demos).


Seybold observes that more, and someday most or all, information and entertainment is being prepared and/or delivered to it's consumers in digital form. The goal for the conference was to provide a forum for discussion and interaction among members of the various disciplines which are being affected by these changes, and to increase awareness and coordination between disciplines. To this end, the conference supplied an impressive list of well known and/or well regarded speakers from a diverse set of areas (some of whom are mentioned below).


Although the "Digital World" moniker was quite broad, most speakers and participants tended to focus fairly practically on their own areas - quite often multimedia computers. Other popular topics included video games, cable and interactive television, CD-I and related formats, computer graphics, computer-based video editing, film and television production, smart telephones, image compression, education, and marketing. None-the-less, the cumulative exposure to many of these subjects was large and several speakers looked somewhat farther afield.


Many conference attendees were policy makers for their companies, and Digital World was unquestionably a good place for making high- level as well as specific technical contacts. After three very active days, I came away tired but much better informed and with a good deal of follow-up to do. Taken as a whole, the conference goals were probably well met.


The Opening and Closing Sessions were two of the most interesting. In different ways, each demonstrated the breadth of the conference. The first speaker was Steve Jobs, co-founder of Apple, founder of NeXT, and part owner of Pixar. He made a number of observations with which I am inclined to agree. He observed that, by consumer standards (VCRs, etc.) "most of what you're going to see is pretty crummy", current products are "the early stepping stones in getting to zero".


On multimedia specifically, Jobs said "I think MultiMedia is a $0 billion per year market" - MultiMedia is an "enabling technology". This issue characterized a split that I saw in much of the rest of the conference, between those who focused on the technology of MultiMedia and other digital systems and who thought that with a nice enough technology a market would develop; and those who thought that one must start by looking at present and anticipated consumer needs and desires, find places where value can be added, and then work back towards present and future technology to meet those needs. I think of these, as the "push" (technology-driven) and "pull" (customer-driven) schools of thought.


"Pullers" included Bob Stein, co-founder of The Voyager Company, a major publisher of interactive videodiscs and software (referred to during the conference as "content"), and a number of the many representatives of the major Japanese consumer electronics companies in attendance. Based on his compelling-by- dint-of-practicality presentation, I'm inclined to say that Peter Blakeney of IBM, who wore an open-collared shirt and sweater during his evening session, is a "puller". Tyler Peppel, manager of the media integration products group at Apple, seems to be mostly a "puller" and coined the term "technofascination" to describe a common motivation among "pushers".


The most obvious "pushers" included Rob Glaser, general manager of the multimedia systems group at Microsoft, who did a song and dance on the MPC (Multimedia PC) standard and demonstrated the Tandy MPC with a 25MHz 386 and preliminary software (scheduled for Summer release); and Harry Fox, president of Advanced Strategies Corp., who seemed to be the primary booster for the MPG (MPC software developers' Group). I believe MPG is part of the MPC Marketing Council which was recently formed under the Software Publishers Association and administers the MPC logotype. Both of these fellows felt that "the market exists NOW" and that the primary obstacle to sales is education of the consumer and retailer - clearly a "push" approach. When Denise Caruso, editor of Seybold's new Digital Media newsletter, in essence accused Microsoft of believing in industry standards so long as they were Microsoft's, Rob Glaser allowed that Microsoft "could do a better job of not always prefacing our listening by proving how smart we are", but his entire answer did not auger well for Microsoft's working cooperatively with others.


Not only was the drive for multimedia somewhat unclear, Jobs observed that "I can't figure out what it (multimedia) is". He suggested that multimedia is primarily "adding full motion video and high quality audio" to existing PCs/workstations, and suggested calling it that. (While that may be correct at the most basic technical level, I feel that useful distinctions can be drawn between voice and music, animation and video, etc. We ought to be able to operate both more powerfully and more naturally by allowing manipulation of the underlying abstractions of the different types of information.) Of course, when Jobs demonstrated NeXT's current multimedia mail system (mildly enhanced over the system of a year ago), he repeatedly used the term "multimedia". He also spoke in favor of standards, separating the delivery medium from the data representation, and interpersonal computing as the goal of the next decade.


The second speaker was Jim Clark, CEO of Silicon Graphics. The first part of his talk focused on the technology for using digital data in the home. He started with "The Chip" - a 300-500 MFLOP MIPS-based CPU with lots of memory bandwidth and interconnectivity for parallel processing, to be used for video, compression, graphics, and sound, aimed at the consumer market (implies high volume implies cheap) and ready in 3-4 years. Several jaws dropped when upon questioning Clark reported that, while the American semiconductor manufacturers he had approached hadn't responded, NEC and Toshiba were aggressively interested and have provided $5M (each?) and 25-35 engineers to design and fabricate the chip! (Privately, DEC implied that they are developing a 200MFLOP chip, to be ready in about two years.) Coming down on the TV side of the very smart TV vs. very graphics household computer debate, Clark sees one of these chips implementing a digital cable TV box, including all decoding functions, and with 3D games, yellow pages, and other functions added in by loading code into the chip from the cable on a per application basis. He also sees the chip as a building block for workstations, further benefiting from economies of scale.


Clark would prefer to see the "Baby Bells" (American regional phone companies) as providers of the next generation digital services, but under the present state of regulation he doesn't feel that they have the incentive to run high-bandwidth connections (optical fiber) into homes. The primary alternative is the cable television providers, who effectively have monopolies in each service area. Ideally, Clark would like to have phone and cable companies competing to provide these services.


Trip Hawkins, president of Electronic Arts, a self-professed TV addict, quoted Marshall McLuen - "TV will make the classroom obsolete" by making the classroom boring. He made the delightful observation that people use the TV remote control a lot because "its all the interaction you have" with conventional television, and referred to multimedia as "the interactive medium", "the medium of doing". He envisions the TV, with a less than $500 "multimedia box", as the gathering place for the family of the future. Hawkins predicted that games and simulations will become realistic enough that adults will become interested, that many household activities will funnel through the TV (e.g. checking who's at the front door), and that the multimedia TV will be used for home video movie editing, catalogs with digital video that will let you display clothes on yourself before buying, and a variety of other functions.


Hawkins claimed that the required bandwidth is only 1M pixel/ second, that CD-TV and CD-I won't provide enough interaction (and that compression is not of great importance). Basically, Hawkins feels that video games are an important, fundamentally useful learning medium for children, with a teacher on the side as a coach. Upon prompting, he admitted that multi-person involvement was desirable, but he didn't seem to have great sympathy for the idea that actual physical interaction with others might be one of the major draws of arcades ($7 billion/year vs. $2 billion/year for "Nintendo"-type and computer games, by his figures). By an informal show of hands, 25% of the audience indicated they would want such a box; that figure rose to 50% after additional discussion and some prompting.


Setting a stake at the extreme end of the spectrum Nolan Bushnell, inventor of "Pong" (reputed to be the first video game) and founder of Atari, said "what we all know" - that the ultimate delivery platform is total control of all senses. Stepping back from that, he observed that conventional television has too low a bandwidth for a person, that "games have a serious problem - they are contentless", "fundamental teaching in schools hasn't changed - and its too damn slow", and "at home we have production values - at school we have boredom". He has also observed his children getting excited about and using a multimedia atlas and encyclopedia (presumably Commodore CD-TV), and suggested that kids will spend 1/2 their school time at monitors and 1/2 with teacher, effectively reducing class size by 50%.


There was an on-going "controversy" over the use of JPEG vs. MPEG. The MPEG standard is essentially finalized and was designed for motion, while JPEG was intended for still images. However, most of the computer-based video companies were using JPEG compression for their motion video, and MPEG didn't seem to be getting very much respect. The JPEG users usually pointed out that they need every frame compressed individually to support frame-accurate editing, while MPEG computes two out of three frames as "deltas" from the previous frame, with a data rate of about 1.25M bits/ second. Although I did not have the opportunity to carefully examine JPEG and MPEG compression, my initial impression is that MPEG may not be of high enough quality for many applications, including broadcast video. (Reports are that discussions on an MPEG-2 standard is already underway.). Didier LeGall, of C-Cubed, did confirm that C-Cubed will be making MPEG compression and decompression chips, which will at least give MPEG a fighting chance..


Of course, some people sidestepped this question. JVC announced their JVC Extended compression method, which provides a higher data rate than MPEG (4-5M bits/second). C-Cubed will be doing the JVC chip, which they say will also support JPEG and MPEG. The evening before Digital World, Apple officially announced their Quicktime operating system software to support motion video, and it got quite a bit of attention at the conference. Several presenters commented very favorably on the system, and several vendors displayed Quicktime products.


In other sessions of particular interest, Gary Demos at DemoGraFX described essentially all of the proposed and existing HDTV standards, while Charles Poynton of Sun Microsystems covered many of the current and future issues in HDTV system design and development. Steve Arnold of Lucas Arts discussed the uses of digital video in feature film production. He displayed Industrial Light and Magic's animation of the pseudo-pod sequence from "The Abyss", and digital matting from the rooftop scene of "Backdraft" and "Terminator 2". (In the scene from "Backdraft", a fireman runs along a collapsing, burning roof. The building, a burning scale model roof, and the person running were all shot separately and then digitally matted.) Diana Gagnon gave a surprisingly interesting presentation on Interactive Television. The "Interaction TV" system was particularly interesting, providing a reasonable array of TV and yellow-pages (telephone directory) style listings, as well as remembering and using user preferences. Gagnon also discussed the "Hi OVIS" (??) system, tested in Japan about 10 years ago, which did essentially everything anyone might expect interactive television to do in the next 5 to 10 years; Anyone with an interest in interactive TV should look into this system. Unfortunately, no one from FROX attended the show, probably because their first public showing was at the Consumer Electronics show in Chicago that same week.


In the Closing Session Andreas Bechtolsheim, architect of the original Sun workstation and co-founder of Sun Microsystems, described a "breakthrough product" which Sun ISN'T building; Mark Weiser, head of the computer science laboratory at Xerox PARC, looked down the road at a truly digital world; and Jean-Louis Gassee, founder of Be Labs and ex-president of Apple Products Division, rather affably summed up the entire conference.


Bechtolsheim described a "rental only", reasonable cost, intermediate resolution, digital video system which "could be on the market in two years" delivering high quality movies on video to the home. Some of his starting points were that reasonably priced HDTV is too far away; digital media is here, but digital transmission is a bottleneck - not least because of FCC regulation; and that "standards are good, but great products are better". The system would display 960 x 540 square pixels, 24Hz progressive scan with a 72Hz refresh, scalable for lower resolution, and be delivered on CD4 - a blue light CD with four times the capacity and data rate - with compression yielding a data rate around 300K bytes/second. The first system would be a portable player with six inch display at $1200, followed by a home deck for $600, and a 28" digital display for $2000. The development of this system would also allow convergence of many of the parts and systems used in consumer electronics and workstations.


Obviously, this system is a natural for a consumer electronics company that also has control of a lot of software (movies, music videos) and sounds extremely feasible. By an informal show of hands, 95-100% of the audience would buy such a system if available two years from now. This was by far the best response to any product idea discussed. (The next closest came after a very interesting talk on interactive Television, when 80% expressed an interest in a box with all monthly costs paid by advertisers and initial purchase price below $200; the "bidding" started at $1000, but that only got a 5-10% response.) [Author's note: If any such company would like to build the digital video system described in the two year time-frame, I would be pleased to discuss heading the project.]


Mark Weiser's charter at Xerox PARC is to consider ideas that are at least seven to ten years out. He described a world of "invisible computational surround" or "ubiquitous computing" where computers are pervasive, surrounding you at all times, but are rarely noticed or used in the current fashion. He described computers evolving into an Invisible Technology, like literacy or electric motors in the present day (they are powerful and pervasive, but almost never noticed for themselves). He envisions hundreds of one inch Post-It-like computers per person per office; tens of one foot notepad or magazine computers per person per office ("an antidote to windows!"); and one three foot or more "whiteboard" computer per person per office. All these computers are linked by wireless local area networks on a variety of scales from part of a room to an entire building with 100,000s of connections. These computers are linked to position sensors, and the entire set-up allows even simple systems to have a good deal of information about their environment and anticipated usage, allowing them to provide services relatively transparently and with much less explicit user control required. Weiser suggest that we will inevitably end up with these types of systems, because of the way people work. PARC has prototyped the 1" and 3' systems (a "whiteboard" with "three button chalk"). He also suggested that "meta-standards" - standard, interpretable methods of describing and implementing standards - would be an important component of evolving, heavily interconnected systems. (Postscript provides a basic form of meta-standard, in that other languages can be implemented in Postscript.) This concept is very important, in that it allows systems with many components and connections to evolve over time without either obsoleting older elements or constraining newer elements to be backwards compatible; it can also allow systems to be deployed usefully and connected with other systems before interconnection standards exist.


The next Seybold Digital World conference will be held June 21-25, 1992 at the Beverly Hilton Hotel, Beverly Hills, California. For information, contact Seybold Seminars, +1(213)457-5850.




Jim Albrycht of Digital Equipment Corporation (DEC) described their Community Multimedia Networking project. DEC has designed and implemented a system for running Ethernet over a standard cable TV channel. This system will allow the existing cable TV installed base (cable into 60% of the homes in the US) to deliver interactive digital connections to the home with minimal overhead. DEC has the system fairly fully worked out, with filtered gateways between different local and remote networks, an installation scheme starting with Universities and Hospitals and progressing to businesses and individuals, and an upgrade path to community and then personal fiber-optic links. This system sounds highly practical, well thought out, and essentially ready for installation (Albrycht runs it in his home/neighborhood). Albrycht reports that, at a recent Congressional HDTV Task Force hearing, chairman Ed Markey of Massachusetts said "The Congress of the United States will make the final decision on the next- generation television infrastructure for the United Status. We would be derelict in our duties as Congressmen if we make an infrastructure to make better couch potatoes. You can be assured that we will make both a toy and a tool to make the United States more productive."

Adair & Armstrong showed a MAC II and LaserDisc based application called "Smart Money". Smart Money is intended to help high school students (16 - 18 years old) learn how to manage money "hands on". Aided by video, sound, still pictures, and spreadsheets the students collect a paycheck every two weeks and select, pay for, and pay off housing, food, furnishings, and other necessities and luxuries. I am generally unimpressed by most current applications, but I believe students would enjoy using and would practically benefit from Smart Money. Adair & Armstrong are looking for partners or alliances to help get Smart Money into the schools.

Verbum showed their two CD-ROM interactive, multimedia version of their magazine "Verbum Interactive 1.0". The "magazine" contains interactive columns and articles, animation and multimedia art, music, and an "interactive roundtable discussion with multimedia industry leaders". Many advertisements take the form of demo programs (e.g. Adobe, Letraset), and the disc contains a multimedia products and services database. The MAC II version, requiring 5M and 8 bit color or better, will be available Summer `91; PC/Windows is scheduled for end of `91; and a CD-TV version is proposed for 1Q92.

JVC announced their WO CD-ROM Drive and Disc, a write once CD- ROM system that will sell for about $2500. JVC, and a number of other companies including Young Minds, offer CD-ROM premastering software.

IBM showed their 3 1/2 inch 128M read/write magneto-optical drive, which should be officially announced shortly before you read this. The disk has a 66mS average access time, can read 384K bytes/second, and write 128K bytes/second. The SCSI drive will sell for $1500-2500, and disks will cost $75. The disks are less than twice as thick as a standard 3 1/2" floppy and make a very attractive package. A drive ran a multimedia product description continuously during the show.

Computer and hard-disk based video editing was one of the most popular product catagories in the Demo Center. Avid (who might be called the industry leader, with 250 systems in the field), SuperMac, Digital F/X, and Light Source all showed systems that are either shipping or under development. Many of these systems display a reduced resolution (as low as 100x100 pixels) and/or a reduced frame rate (as low as 2-5 frames/second), although without special hardware assist. Avid will be coming out with a high resolution system in October `91. Using JPEG they expect 20K bytes/frame and close to 3/4" video tape resolution. Fluent and Avid are discussing a low-end PC-based version of Avid's product, using Fluent's just announced video / audio / realtime compression / video windowing board.

Quickdraw-based products shown included MacroMind Director, Light Source's MovieTime, Diva's Video Shop, and SuperMac's Spigot and ReelTime. Quickdraw was officially announced the day before Digital World.

Clarity Software showed a multimedia mail system for the Sun SPARCstation under X/OpenLook. It provides a set of functions similar to NeXT's multimedia mail, along with a number of useful sorting and searching functions and the easy integration of new applications. Mail will be compatible across DEC, HP, and Silicon Graphics systems which will be supported by 1Q92. It looks like a very nice system which I would very much like to see Sun adopt as a standard part of their window system release.

Commodore and their CD-TV system were notably absent or keeping a very low profile. There was one attendee from Commodore registered. Likewise, there was no sign of Phillips and their CD- I system, which is reported to release later this year but not support full motion video until sometime in `92.



About the Author: David Reisner is president of Synthesis and David Reisner, Consulting, a Southern California-based firm supplying services to the US and International markets. Mr. Reisner has been consulting for more than ten years in areas such as software and computer systems architecture, programming environment and user interface design, digital audio systems and signal processing, product design, custom research and reports, and multidisciplinary problem solving (generalist / synthesist). Past projects include hard-disk based sound recording and editing, biological "fingerprinting", a clipboard computer (in 1980), a portable source-level debugger for Ada, and killer whale training using sound projected underwater. Synthesis / David Reisner, Consulting, dar at



David Reisner  -  Synthesis  -  info16 <at> <this domain dot com>