Multimedia has arrived. I read it in PC Computing. What? You're not convinced by this authoritative testimonial? Sigh. We live in an age of cynicism.
Personally, I'm so jaded I don't even trust cynicism any more. Talking with a musician/programmer recently convinced me that, for all the hype and hoopla over multimedia, there are real technological issues here, and there is a real market.
Or rather, there are several technologies and several markets. And the markets--or the various views of the market for multimedia stuff -- don't map in any very useful way onto the different technologies involved.
It's all rather messy, which may help to explain why the fluff flags go up whenever the subject is raised. The musician/programmer I mentioned admits that, for his company, today's multimedia market isn't necessarily tomorrow's multimedia market.
That musician/programmer is named Roger Powell, and he's the manager for audio applications for Silicon Graphics. Talking with Powell set me to thinking about, and then to reading up on, the multimedia phenomena: the emerging standards and the forming markets and the players who appear to be molding this clay, and the very pragmatic question one keeps coming back to: What is this stuff for, anyway?
Powell's story is more or less the standard computer meets music, music goes on the road, computer finds music again story. Originally a professional keyboard artist, Powell got involved early on with the Arp synthesizer company. "I've always had this dual interest in music and technology," he recalls, "and when personal computers became available I gravitated toward them and began learning programming as well as making records." Make records he did, with Todd Rundgren's group Utopia, but he never got far from computer technology. When S-100 machines arrived on the scene in the late '70s, Powell began assembling the hardware and writing assembly language sequencers to control analog synthesizers.
"Things got a little more sophisticated when the Apple II and the IBM PC came out, and there were actually good C compilers," he told me. "And then MIDI was invented, which provided a protocol and some hardware to actually talk to these instruments." Around 1985 he released a product named "Textures," one of the first MIDI sequencing programs for the IBM PC. He was still with Utopia then, still working as a professional musician.
When Utopia quit touring, Powell got into engineering heavily, working for a Boulder, Colorado company that was producing digital audio workstations, machines targeted at the professional audio production market. He worked there "for about four years, primarily writing user interface code in Microsoft Windows, which [ran] on a computer that connected to the workstation." When he got a call from Silicon Graphics describing what they had in mind for audio, he soon found himself working there.
It's easy enough to see why Powell is doing what he's doing. Like Steve Wozniak building the Apple I or Bill Gates writing Altair Basic, he's engaged in work that is largely indistinguishable from what he would call play. He's doing it at least in part for his own satisfaction.
The analogy gets stressed when you press it. What Woz and Gates did turned out to be eminently marketable. Nobody needs to defend or define the market for personal computers or development software. It's a different story with multimedia.
It's tempting to ask, even if there is a multimedia market, why is there a multimedia market? Bill Gates is one of the people I've heard answer that question, essentially, in this way: Because today's machines have more power than can be justified if all you're going to do is run spreadsheets and word processors. There needs to be a power-hungry application to justify the new hardware. A story well suited to a cynical age.
The product that inspired PC Computing to announce, in its December 1991 issue, that multimedia had arrived is a program from Macromind called "Action." Action is a Windows application that builds on some emerging standards for multimedia delivery platforms and on multimedia extensions for Windows from Microsoft. It represents one answer to the question, What is multimedia for? In the Action paradigm, multimedia is advanced presentation software. The user developing a presentation can use Action to add sound, video clips, animation, or interactivity. More impressively, Action lets one add sound, video clips, animation, and interactivity to presentations: It facilitates the simultaneous, synchronized presentation of information in different media.
That's one definition of multimedia: advanced presentations. The Silicon Graphics path to multimedia suggests a different definition, one based on the democratization of professional audio and video production. That view grows out of the company's existing customer base, and it explains why Silicon Graphics hires people such as Roger Powell. But, as we shall see presently, Silicon Graphics doesn't have just one view of multimedia.
But back to Macromind. Macromind is worth watching. Besides Action, Macromind also makes multimedia tools for the Mac; in fact, it is the multimedia company, more than any other software company. Its name isn't Macromind any more, though, since it merged with another big media company, Paracomp. Farallon used to be another player in multimedia, with a set of sound tools, a base on the Mac, and a serious commitment to Windows. But Farallon recently decided to concentrate on its connectivity products, and sold off its sound tools to--right, Macromind-Paracomp. The company is gobbling up the multimedia applications market for Macs and PCs.
But an application software company isn't in the best position to create the standards. Apple and Microsoft have taken serious steps toward establishing standards for multimedia.
Basically, multimedia requires tools and standards for sound, video, animation, music, and why not 3-D graphics while you're at it. Multimedia systems need to be able to store and retrieve video and sound in real time, support acceptable and/or professional quality sound and video, deal with existing formats and devices, edit, synchronize--there are many problems and questions.
Apple's QuickTime is both software and a standard for time-based media: video, sound, animation. It includes compression and a standard format named "Movie" for digital video and sound. On existing hardware, QuickTime Movies are going to be small, slow, and monophonic. Not exactly Michael Jackson, but the fault is more in the current state of Macintosh hardware than in QuickTime, which is generally regarded as pretty spiffy. QuickTime is software, so it can make almost any Mac multimedia ready, by this limited definition.
Microsoft led the way in creating a standard for multimedia-ready machines in the DOS/Windows universe. The Multimedia PC (MPC) standard is pretty minimal, but the existence of a standard let Microsoft get on with extending Windows to support multimedia. Microsoft is turning technology licensed from Macromind into a movie format akin to Apple's. MPC is essentially a delivery platform standard.
The first MPC-compliant machine announced was Tandy's five boxes, introduced last summer, each containing the requisite CD-ROM drive, audio circuitry, Windows with multimedia extensions, and stereo sound drivers. Without monitors, the machines start at $2599.
The Apple-IBM multimedia venture, Kaleida, hints at higher bar, both for authoring and delivery. The most salient feature of Kaleida appears to be a media scripting system that will evolve from QuickTime and be usable on DOS, OS/2, and Mac systems.
And then there is the definition provided by Silicon Graphics' Personal IRIS line, in particular the Indigo.
Indigo is breaking new ground for Silicon Graphics, but gradually. "The premier customer for Silicon Graphics," Powell told me, "is the creative or technical professional. Our traditional markets have been the scientist's desktop, but we're also very strong in the entertainment industry, from the computer animation standpoint. Our machines are used quite heavily in the film industry for doing animation and effects. 'Terminator II,' Michael Jackson videos, that sort of thing."
Indigo puts some of this professional video and audio capability on a machine that compares favorably in price with heavily decked out, high-end Macs and PCs. The $8000 entry price for an Indigo can quickly ramp up to $10-12,000 with additional disk drives and add-ons, but for media development, a lot of what you'd want is already there. Powell paints the picture of the short-term market, which is the existing customer base. "The low-end machines can populate the desktops and connect to the larger machines, which might be render servers. Things that take an enormous amount of processing you might offload to the bigger machines, but a lot of the work can be done at the desktop [with an Indigo]."
But what about emerging markets? I asked him. He agreed that the Indigo was viewed a wedge for opening new markets for Silicon Graphics, particularly digital media authoring, "because the price factor puts you in the category of a high-end Mac or PC, and the capability is far greater than you get with those machines. Those machines are pushing the envelope on the processors installed in them and the peripherals that you can plug into them. You start plugging cards into a Mac to try to get it to be a digital audio workstation ...you're sucking up all the power of the machine and you have no expansion path."
If he's right, it puts an ironic twist to the argument that the multimedia market exists because the machines are powerful enough to support it.
Is he right? I waded through descriptions of reasonable development systems based on PCs, Macs, and Amigas, culled from press releases, magazine articles, and conversations with developers. Here's the picture I get:
It's possible to put together a Mac, PC, or Amiga system that displays color 3-D animations; gives the user control over on-screen motion video and lets the developer capture and edit video; accepts and presents CD-quality sound; and has software tools for multimedia integration, painting, and 3-D modeling. I come up with something like $12,000 for a minimal Mac system. You can get into multimedia development for well under $10,000 on an Amiga or a PC, but you'll definitely be cutting some corners.
The Indigo system that checks all the same boxes will probably run closer to $12,000 than to the entry $8,000, but in general you're getting more with the Indigo. It's built around a RISC architecture; is ACE-compatible; has an audio subsystem that isn't affected by CPU load; and has a video bus that cuts the costs of video cards because video and graphics share common back-end circuitry. The operating system is IRIX, which is SVR3 UNIX with 4.3BSD extensions such as TCP/IP and NFS network protocols. IRIX implements the X11R4 window system, Motif, and Display PostScript.
A look at the back of the case shows how much is built in: a stereo digital I/O jack for input and output from or to a DAT deck, CD player, or MIDI device; stereo line-in and line-out jacks for analog I/O; a microphone input jack; a stereo headphone jack; two Mac-compatible serial ports; a SCSI II port; an Ethernet port; and a bidirectional Centronics parallel port.
The figures are for system, memory, monitor, necessary cards, disk storage, and software. Video cameras, scanners, and other such devices are extra.
The Indigo system strikes me as a gorgeous machine to play with while trying to decide what the multimedia market is.
Copyright © 1992, Dr. Dobb's JournalAfter Utopia, What?
Lights, Camera, Action
Defining the Development System