[ Go to February 1997 Table of Contents ]

Optimizing Windows /
John Woram

Now Hear This!
Video is cool, but don't overlook the audio side of your multimedia productions.

Within minutes of installing a new sound card, just about everyone turns to the funstuff folder on the Windows 95 CD-ROM to check out the audio and video performance of the sample AVI files. After that, it's time to produce your own movie masterpiece.

Well, you can't-unless you're willing to spend some serious bucks on the audio/video production hardware you'll need. If you're not ready for that kind of financial commitment, you can still dabble in entry-level multimedia A/V production without mortgaging the farm. Here's how to get started.

First, you need the secret password: "ScreenCam." This neat little multimedia recording applet may be Lotus Development Corp.'s best-kept secret, especially since it's been tucked away in Lotus' SmartSuite package. As I write this, a standalone version for Windows 95 is just emerging from the beta cycle; it should be available when you read this.

ScreenCam essentially turns your PC into a VCR that records every click, scroll and action displayed on your PC's monitor. You can also add captions and your voice to the recording. Not the least of its attractions is price-a mere $129. Since ScreenCam is such a great way to get started in multimedia production, I'll use it here to demonstrate some of the more basic audio/video production techniques, starting with the audio segment.

Perhaps because audio is a recent addition to the personal computer, some people consider it just a nice add-on-a tool that contributes a little something extra to a multimedia production. Well, guess again. In just about any A/V production, the audio is as important as-and often, more important than-the video. As proof, just switch on that well-known multimedia device, the television set. Tune in your favorite channel, adjust the brightness, tweak the contrast and fiddle with the color controls. Then hit the mute button and enjoy. Chances are you'll be left clueless-unless it's a talking heads show and you're an expert lip-reader, or you've got closed captioning. Turn the sound back on, and even without the picture, you'll have a general idea of what's happening-with the exception of sight gags, of course. So even in your simplest A/V production, ScreenCam or otherwise, audio deserves as much attention as video.

Improving microphones

In the past, I've said that the inexpensive microphone included with some sound cards was probably worth about what you paid for it. I may have overstated my case, but in a previous life I was a recording engineer who used microphones that carried four-figure price tags. The bad news is that some top-notch condenser microphones remain up in the kilobuck zone, and a really fancy stereo microphone still goes for more than $5,000. But the good news is that the latest generation of inexpensive bundled microphones is better than its predecessors. You might not want to record the Philharmonic with one of them, but it should be fine for basic production work.

A typical example is the microphone that Creative Labs bundles with its Sound Blaster AWE32 Plug-and-Play sound card. If you've acquired such a microphone recently, you may have noticed a mini-stereo tip/ring/sleeve plug on the microphone cable. You may have also discovered that it won't work with older sound cards. The reason is that-stereo plug notwithstanding-this is a mono microphone. The stereo plug is the tip-off that it's an electret condenser design, which requires a polarizing voltage at the capsule. On that stereo plug, the tip and sleeve convey the audio signal from the microphone to the sound card, while the ring delivers the required 5 volts DC from the sound card to the microphone.

If you plug this type of microphone into a conventional input jack on an older sound card, the microphone doesn't get its polarizing voltage and is therefore dead. By contrast, you can plug a nonelectret microphone into a sound card designed to supply a polarizing voltage without fear of damaging it. Inside the card's input jack, the voltage supply line makes contact with the sleeve on the mike plug, which short-circuits the supply line to ground. Neither the microphone nor the sound card is any the worse for the experience.

Testing, testing ...

Microphone designers strive to design windscreens that do the impossible: Protect the microphone diaphragm from the potentially harmful shock wave of a windblast, while being acoustically transparent to an audio signal. It's not a trivial task. Yet many people think the best way to check a microphone is to blow into it. That's fine if you work in an environmental test lab and need to know how it will stand up to shock and saliva, but a much better test method is talking into it (not such a radical concept when you think of it). Or if you're still mike shy, scratch the protective grille (gently!) with a fingernail.

Once you're sure the mike works, you should find a convenient position for it, with respect to both the user and the computer. Generally, the less expensive the microphone, the more sensitive it is to unwanted windblasts as the speaker pronounces certain letters-especially f and p. As a quick check, hold a scrap of paper horizontally under your mouth and recite the alphabet. (You might want to do this when no one's around-it's not easy to explain why you're talking to a piece of paper.) Repeat the test with the paper above, and finally, in front of, your mouth. As the paper flaps in the breeze, you'll note that each letter generates a distinctive wind pattern that would strike a microphone diaphragm in the same relative position. Since the strongest wind patterns are directed forward or downward, you can protect the microphone against these blasts by positioning it slightly above the speaker's lips.

Noisy fans

Another consideration is the noise of the computer itself, usually generated by its fan(s). Resist the temptation to disconnect the fan, unless you're confident that the computer won't mind working for a while without it. Better yet, move the system unit as far away from the microphone as possible, or place it on a chair. This gets it away from large wall and desk surfaces, which generally deflect the noise toward the microphone.

Before pressing Record, select the minimum sampling rate, sample size and number of channels you need to achieve the desired quality. It might be worth doing the math up front: A sampling rate of 44,100kHz at 2 bytes per sample works out to 88,200 bytes per second, or twice that in stereo. Or, to put it another way, about 10MB per minute for stereo recording, which should give you a rough idea of how many record albums will fit on a hard drive. You can cut the file size by 50 or 75 percent by recording at 22kHz or 11kHz, and in half again by recording 8-bit samples, so it's probably worth a few tests to determine your required minimum. For voice-only recordings, you may not notice much deterioration as the sampling rate is decreased from 44kHz to 22kHz to 11kHz, but the noise-level difference between an 8-bit and a 16-bit sample will always be apparent.

Finally, if you use nothing more than a single microphone connected to the sound card's microphone input jack, make sure you record in mono. There's no point wasting storage space on stereo, since the mono output from the microphone is fed equally to both channels anyway. And depending on your sound card and the software mixer you use, the left/right slider may be disabled or nonexistent. If you need to pan from left to right, then you should select two-channel mode, and you'll need an external mixer if the on-screen slider is disabled.

By comparison, video configuration is kid stuff. If you have no control over your audience's system configuration, then configure your own system for a 640x480-pixel screen resolution. Otherwise, if the audience views your high-resolution production on a monitor configured for a lower resolution, part of your production may be off screen.

Color coordination

As for color depth, there's nothing like a 16-bit high color (better yet, 24-bit true color) video show-and nothing worse than viewing such a production on an 8-bit system. Played back at 256 colors, those lavish color videos generally look like a cheap imitation of an Andy Warhol original. Play it safe by recording at 256 colors or, if you can get away with it, at 16 colors. Not unlike the audio situation, file size grows as screen resolution and color depth are increased, though not as drastically as doubling one of the audio parameters.

Once you've set your audio and video parameters, if you want a nice test of hand/eye/mouth coordination, try to produce your ScreenCam's audio and video at the same time. I suppose it can be done, but certainly not by me-I usually manage to flub the audio while dragging the mouse pointer across the screen and clicking all the right buttons at the right time. By the time I get to "take 27," I figure there must be a better way and, of course, there is. Now, I just ignore the audio until I get the video action recorded properly. Then I watch the playback several times; once I'm reasonably familiar with it, I'm ready to take on the audio track. With nothing else to do, I can just sit back, relax and describe the on-screen action as it happens.

This is easy enough, provided the sound track is nothing more than the signal from a fixed-level microphone input. Things can get tricky, though, if you must make dynamic audio adjustments during the session. Assuming you're watching the finished video, the screen is not available for displaying your software mixer, and therefore you can't make the necessary changes. If this is a consideration, then it helps to have a second system (and an assistant) on hand. Set up that system to handle the sound, and feed its sound-card output into the line input on the main system's sound card.

Next, you can cue your assistant to fade the sound effects in and out from the other machine, while you're watching the show on the main system and the audio controls on the other screen. Assuming its level will remain constant, you can connect the microphone to either machine. If there are considerable music-only segments, then connect the microphone to the secondary sound card and mute it when you don't need it.

Limiting factors

You'll run up against the inherent limitations of on-screen audio control when you vary two or more settings at once. Even with some additional software assistance, audio balancing on a computer monitor leaves a lot to be desired, and so this should be the time to start thinking about an external hardware mixer that will accept inputs from several microphones and various playback devices.

Now, if you'll excuse me, I'll have to continue this some other time-Spielberg's on the phone.

Senior Contributing Editor John Woram is the author of The Windows 95 Registry: A Survival Guide for Users (MIS: Press, 1996). Contact John in the "Optimizing Windows" topic of the WINDOWS Magazine areas on America Online and CompuServe, or care of the editor at the e-mail addresses here.

Windows Magazine, February 1997, page 242.

[ Go to February 1997 Table of Contents ]