Is multimedia really a solution without a problem? Can sound, music, full-motion video, and graphics effects increase your productivity? Do you really need a talking spreadsheet that tells you "I'll be back" upon termination? Even the big multimedia-related vendors have trouble answering questions such as these. That's why their response usually sounds more like a "why-ask-why?" commercial for Bud Dry--an answer that tastes good, but is less than filling.
New products--hardware and software--are cropping up faster than developers, users, and even standards committees can keep up with them. Like it or not, multimedia is reaching critical mass.
On the upside for developers, multimedia programming can be downright fun, albeit at times confusing. Which video file format should you use? How can you support all of the available sound cards? What are the best compression schemes? Can a DSP really increase performance without sacrificing flexibility? To help you answer this ever-growing list of questions, we're bringing you Dr. Dobb's Sourcebook of Multimedia Programming, a special issue of Dr. Dobb's Journal focusing on multimedia software development. Typical of the tricks you'll uncover inside are Scott Anderson's techniques for morphing and Neil Rowland's C++ class library for encapsulating Windows' low-level waveform audio services. Other authors examine digital-video file formats, audio-compression techniques, and programming animation under Windows. You'll also read about the VESA committee's efforts to create a standardized software interface for audio, and--in the spirit of the original Dr. Dobb's--a feature on how to roll your own RS-232-based sound system. By the time you've finished reading Dr. Dobb's Sourcebook of Multimedia Programming, you'll be on your way to deciding what's hot and what's not when it comes to multimedia applications development.
For instance, among the "what's hot" items is QSound, a virtual-audio technology from Archer Communications (discussed by John Ratcliff) allows left and right panning within a 180-degree "soundscape." The sound technology was first used by the film industry in the movie Willow, and has been used by recording artists such as Paula Abdul and Madonna. Game developers have now begun to include QSound in CD-ROM games. Although the QSound algorithms are patented, Creative Technology has licensed the technology from Archer and is providing, free of charge, a QSound API to registered Creative Labs/Technology developers. Sierra Semiconductor has also licensed QSound technology and is currently providing it in their Aria chipset.
Another area that's warming up is speech recognition, and several companies are clamoring to tap into what some experts project will become a one-billion-dollar industry by the end of the decade. At the high end, IBM has ported its RS/6000-based, continuous-speech technology to the 486 and Pentium. Included in its product line is a developer's edition codeveloped with Carnegie Mellon University that sells for $299 and includes a 1000-word vocabulary. Interestingly, both Creative Labs and Sierra Semiconductor have added speech-recognition technology to their product lines. Sierra's Aria Listener is a low-cost speech-recognition engine included as part of its multimedia chipset that also includes the Aria Synthesizer for general-MIDI support. Creative also has a Windows-based speech-recognition engine called VoiceAssist. As with QSound, Creative is providing a VoiceAssist API, which is available for $99.95. Just as interesting is a recent agreement between Borland and Creative Labs to bundle an OWL 2.0-compatible VoiceAssist API in Borland C/C++ 4.0.
Of course, all of this new capability comes at a price. To keep up with emerging technologies, the Multimedia PC Marketing Council has upped the ante on the definition of a "basic" multimedia system. The new MPC level 2 specification defines a multimedia PC as having a 25-MHz 486SX CPU, a minimum of 4 Mbytes of RAM (with 8 Mbytes recommended), a double-speed CD-ROM drive capable of transferring data at 300 Kbytes/second, a 160-Mbyte hard disk, and a VGA+ adapter with a display resolution of 640x480 and 64,000 colors. (For more information, contact the Multimedia PC Marketing Council, 1730 M Street NW, Suite 707, Washington, DC 20036.)
Given this relatively high cost for entry, the technical hoops to make it work, and the lack of clear productivity gains, you have to wonder if multimedia is worth the effort.
Recall that in the early days of the PC color added little to the immediate productivity of users, yet the CGA quickly became standard fare. Obviously, if history is any gauge, appearances are indeed important. Now, sight and sound allow multimedia developers to involve users in ways not readily possible before. What you see and hear is what you get.
Michael Floyd
Executive Editor
Copyright © 1994, Dr. Dobb's Journal