Home Music Music News

Computers That Make Waves

Digital synthesizers take to the road

Muhammad Ali, Stevie Wonder

American heavyweight boxing champion Muhammad Ali (center) holds a microphone as he listens to singer Stevie Wonder, who sits at the keyboards, Chicago, Illinois, circa 1980.

Adger Cowans/Getty

Open your window and listen for a few minutes. You may hear cars rolling by or dogs barking or airplanes zooming overhead. Maybe you’ll hear crickets or your neighbor’s kids. Or maybe just the wind. Those subtle, complex sounds — and an infinity of other real and imagined noises — can be at your fingertips with the newest generation of digital synthesizers.

“We are literally on the edge of a significant technological change in music,” says Dartmouth Professor John Appleton, inventor of the Synclavier digital synthesizer. “Sophisticated technology has simply not been available to the performer. It’s been in the laboratories and universities, but not out on the road.”

As computers get smaller and less expensive, computer-controlled digital synthesizers designed for the working musician are beginning to appear on the market. They’re still expensive — the Fairlight CMI costs $27000; the Synclavier, $14,000; the Roland Micro-Composer, $5795; and the Crumar, $1650 — but they’re already turning up in rock contexts. Stevie Wonder is carrying a Fairlight on his current tour, and Led Zeppelin’s John Paul Jones recently bought one; Herbie Hancock owns a Synclavier; Tangerine Dream’s arsenal includes a Micro-Composer.

To the musician who knows how to use them, digital synthesizers offer unparalleled flexibility and precision. As with digital recording, the sound is treated as a string of numbers; for digital synthesis, those numbers are generated, manipulated and stored by a computer. Digital synthesizers are flexible because any sound can be turned into numbers, or “quantized.” And since those numbers can be stored and reproduced exactly, the sound stays as programmed, without distortion.

The first commercially available synthesizers — including Moogs, ARPs and other familiar brands — were analog. In analog synthesizers, sound is generated by a varying voltage stream that moves through the machine, where it is shaped by such devices as filters, amplifiers and attack and decay controls (or envelope generators). These control devices are inexact, and creating a complex sound is difficult or, at best, unpredictable. Like any electronic signal, the voltage stream is subject to distortion, and anything from atmospheric conditions to an infinitesimal twist of a dial can cause a sound to drift. The only way to store an analog sound is on tape, with corresponding loss of fidelity.

Digital synthesis offers numerous theoretical advantages. According to synthesist-composer Larry Fast, who is currently experimenting with digital synthesis at Bell Labs: “In a digital synthesizer, rather than directly manipulating an actual voltage stream, all manipulations are carried out in a computer as math functions. Filtering and amplitude changes and enveloping are math formulas. At the last possible stage in the synthesis process, when you have a string of numbers that define every instant along the wave form, those numbers are turned into voltages for an analog signal — because loudspeakers, and your ears, are analog devices. All kinds of ragged, irregular, complex wave forms can be created simply by assigning different numbers to each point on the wave form. And those complex wave forms are much closer to what the real world is like than we’ve ever gotten before.”

The advantages of digital synthesizers don’t end with their capacity to create abstract sounds. Once a sound has been quantized, the computer can reshape it in as many ways as a programmer can imagine. “When a sound is stored in the machine,” Fast says, “it’s just a bunch of numbers, and the machine doesn’t know where it came from.” It is possible to program an entire piece of music, then instruct the computer to completely shift its proportions — to stretch out the sound or radically condense it. In the Fairlight CMI, the computer can interpolate between two different sounds, creating a smooth dissolve (over any length of time the player specifies) between, say, the sound of breaking glass and the tolling of a gong.

Fast, who plays the Fairlight on Peter Gabriel’s upcoming album, considers it “the best of all the synthesizers I’ve come across.” For state-of-the-art technology. it’s a surprisingly unassuming machine. Instead of the array of modules, dials and patch cords that make most analog synthesizers resemble the cockpit of a DC-10, the Fairlight consists of three innocent-looking boxes. One is the computer, which includes a slot for its Floppy Disc memory. The second box is the master keyboard, which can take on eight slave keyboards, and the third is the Graphics Monitor, similar to a cathode-ray-tube computer terminal, complete with an alphanumeric. keyboard that looks like a typewriter.

Sounds can be created by drawing in wave forms on the Graphics Monitor with a light pen, in a process something like using an oscilloscope in reverse: the screen displays a graph, and the player draws in the pitch or decay characteristics of the sound he wants, in minute detail. According to Bruce Jackson, who represents the Fairlight in the U.S.: “Musicians don’t like to think too abstractly, and without the knobs to twist, they could feel intimidated. But this gives more physical action to the machine.”

The Fairlight also includes a live microphone; the computer will digitally analyze up to one second of sound, which can then be played up and down the keyboard or otherwise manipulated in the computer through typed-in instructions. Besides rendering the Mellotron obsolete — with one note from any instrument, the Fairlight can re-create an entire range — the live microphone allows real sounds to be blended with physically impossible ones.

“There are some things that analog machines can do right now that digitals can’t,” Fast points out. “Not all the digital machines are capable of doing everything, on the basis of holding costs down or because the program is difficult or hasn’t been developed yet. Something simple like a glide between notes — a portamento — is very easy on an analog machine, but it’s extremely difficult to program on a digital machine. But, in fact, anything in a digital machine becomes pretty much as easy or hard to program as anything else, so the very complex things are close to the easy things. It’s just that at some point, you have to draw the line on how expensive that beast is going to be. And when you get some real fast sixteen-bit microprocessors, which are on the horizon right now, there’s going to be a big price break.

“In the long run,” he concludes, “I don’t think there’s anything, in theory, that an analog machine can do that a digital can’t.” 


Powered by