Page 1 of 5 previous | start | next Printing? Use this version

By Steven Gulie

Greg Keith is dying. It doesn't matter that he's a poet. The cancer doesn't care about his extraordinary voice. But I do. I want to record the rich sweetness of that sound before it's gone. More than that, I want to digitize it, to make a crisp copy that won't degrade, that will last forever.

I go to another poet, Maude Meehan, who has two "talking books" out, looking for free advice. "I have the master tapes right here," she says. I gape in horror as she pulls out two dusty cassette tapes. She looks at me. "Is something wrong?"

No, no, I assure her, finding an excuse to leave quickly. Everything is wrong. Those little cassettes are amazingly noisy. Play a blank tape into a spectrum analyzer sometime. Look at the waterfall of sound. From 2,000 Hz to 8,000 Hz it's like a blanket. To fix it, recorders "pre-emphasize," boosting the hell out of the higher frequencies, trying to pole-vault them over this wall of noise. On playback, we de-emphasize, squelching the high frequencies. The noise goes away, pretty much. So do the high frequencies we recorded. But we boosted them, right? So it's OK?

Well, sure, if the de-emphasis is exactly balanced to the pre-emphasis, if the filters have incredibly expensive front ends, if thermal noise can be kept to a minimum, if chaos and entropy can somehow be kept from meddling. Right. If we knew how to stop that, we'd just keep the cellular DNA from degrading into cancer and ask Greg to read that piece aloud, One More Time.

I remember buying my first boom box. It had a five-band graphic equalizer. "Does it have Dolby?" I asked. The salesman flicked the rightmost slider on the equalizer to the bottom. "There you go. Dolby." OK, he was an asshole. But he was right.

Analog tape just doesn't cut it. For those of us in the computer industry - I work for Apple - analog anything can begin to seem clumsy. With analog audio recording, it's not just the tape noise, but the way it's encoded: the tiny magnetic charges, each one's strength proportional to the air pressure on the microphone at the instant of recording, each one fading, blurring. The molecules, neatly lined up by the tape head, slouching into disorder, talking to their neighbors about nothing, yielding to noise and chaos.

I go to the sound studio of another friend, Eric Thiermann of Impact Productions. "Use DAT," he said. "Use my studio." Of course. Digital audiotape. The real trouble with analog tape isn't that it fades. Everything fades. But with digital media, lots of signal essentially says "1," little or none just says "0." As long as you can distinguish a lot from a little, you can tell the 0s from the 1s. You can make a perfect copy. When you copy an analog tape, you copy the noise as well, while adding fresh noise in the process. You can't help it. When the original decays, you can't get it back.

I sit in the studio, headphones warm on my ears, letting Greg's voice wash over me.

Of an axe biting into wood all day we say the edge galls.
Chips and divots out of pavement and such we call spalling.
The odd dime-silver atom sticks in the whorl of the thumb.
I count on such erasure, spend that coin again.

Page 2 >>