There are certain words these days that are being used to sell just about everything.
“Light” (or, horrors, “lite”), is one of them, appearing on everything from beer to slightly-less-greasy-than-usual potato chips; “cholesterol-free” is another; “green” is a third, and afourth, and the one I want to talk about, is “digital.”
Digital dashboards, digital TV, digital audio, digital watches– digital is becomingsynonymous with high technology. But what’s so special about digital-this and digital-that? Just what is digital? And how does something become digitized?
Well, the word digital comes from digit, as in numbers. To digitize something, then, is to turn it into numbers; more specifically, to turn it into binary numbers– the numbers that computers use. The whole point of digitization is to enable computers to manipulate the digitized information.
The binary system of numbers is called that because it has only two digits: 1 and 0. These correspond to the “on” and “off” of an electrical switch, and when you get right down to it, that’s all computers are, millions and millions of tiny on-off switches wired together.
Now, it might appear difficult to convert, say, Beethoven’s Fifth Symphony into a string of ones and zeroes, but that’s exactly what happens in the process of creating that increasingly common example of digitization, the compact disk.
The changes in audio recording offer probably the best examples available of both digital technology and its alternative, analog technology, so let’s take a closer look at just how that music gets from the studio to your ear.
LPs and cassette tapes are both analog recordings: “analog,” because, on an LP, the depth and straightness of a continuous groove varies in a way that closely matches the variations in pitch and volume of the original sound– it is therefore “analogous” to the original. On a cassette tape, those variations are recorded by the varying strength of a magnetic field. In both cases, the more closely the variations in the recording match the variations in the original sound, the better the recording. However, any distortion introduced at any point is a permanent part of the recording.
In a digital recording, however, the original signals are stored as binary digits, and turned back into sound only at playback, eliminating several intermediate steps and the opportunity for distortion.
But how can a continuous stream of information, like the varying intensities of pitch and volume produced by a vibrating violin string, be reduced to a series of numbers?
The answer is, it can’t– not perfectly. If you think of an analog recording as a smooth line drawn on a graph, a digital recording is a collection of dots. The dots may be so close together that they look very much like the smooth line, but there are still gaps, however tiny, between them. So how come digital audio doesn’t sound fragmented?
It’s because of the ability of computers to manipulate binary digits– and to do so very, very fast. Sound that is being digitized for a compact disk is sampled at intervals of 25 microseconds, or roughly 40,000 times per second. At each sampling the intensity of the sound is assigned to one of about 65,000 levels and is described by a string of 16 binary digits. At 40,000 samples per second, our ears and brains are simply not able to detect that any information is missing. As well, on playback, an electronic filter “smooths off” the corners of the abrupt jumps from one encoded level to another, recreating a continuous sound.
Still, those “gaps” in the recorded information are still there, and in some cruder forms of digitization they’re quite noticeable. Simple computer graphics, for example, have a distinctly “grainy” look caused by the fact that only widely-spaced or too-large parts of the image have been digitized. The more digits you use to describe something, the better it’s going to look or sound.
However, computers can even overcome the shortcomings of insufficient digitization to a certain extent. They can fill in some of the gaps with new strings of digits obtained by averaging the values assigned to the nearest samples. They can correct errors– remove flyspecks from a photo, for example, or distortion from an audio recording– by throwing out values that are obviously wrong. This kind of enhancement is one of the great advantages of digital recordings, most obvious in NASA’s photos of the planets, which are, of course, transmitted digitally from spacecraft to Earth, and then are “computer-enhanced.”
If there’s one hazard to all this digitization, however, it’s the flip side of the ease with which computers manipulate digitized data. Early photos from the Viking lander on Mars showed the sky to be a lovely shade of blue. It turned out it was really pink. Scientists had had the computer make the sky blue because that was what they expected; only when they re-checked the original data did they realize their mistake. Today, magazine photographs are routinely digitized and then touched up. Don’t like that house in the background? The computer will gladly take it out of the picture altogether, or paint it a different colour. Think the model would look better if she were black? Presto! She’s black.
Digitization has made it possible to store and manipulate information in ways our ancestors couldn’t have imagined. It has fueled the information explosion and revolutionized electronic communication of all forms. But don’t let the ads fool you: just because something is digital doesn’t mean it’s better. It just means that somewhere along the line, a computer gave it the once-over.
After all, a compact disk is great, but a live concert is even better. And the concert is analog all the way.
