Angles Of View

One of the most interesting things about Information is that its transmission is essentially organized in two, and only two ways. Despite the enormous variety of available languages, despite the huge selection of available media, when it comes time to send a message, the content will be either analog or digital. And even though these two labels are bandied about in a large number of contexts, it may still be productive to ask

Analog and Digital - Signal or Code?

The world that we perceive is analog. Direct one of our senses we have at any portion of the world and the data available will be unlimited. If we add an instrument whose sensitivity to these data is greater than our own (a microscope or a telescope are two examples), we will still not be disappointed by reality’s ability to present itself in ever greater detail. In fact, except at the level of the subatomic, all of our perceptions are limited only by the sensitivity of the instruments we use to make them. The external world really is continuous at all but the very smallest of scales.

This property of being continuous applies not only to the spatial resolution of the world but also to its temporal dimension. Although we divide a day into 24 equal parts and then divide each of those by 60 and then by 60 again, these units are convenient only to the diurnal rotation of our planet. They say nothing about the actual divisibility of time itself. From millennia to picoseconds, the duration of the world can also be measured at any scale and, theoretically, to any degree of accuracy.

When we wake up in the morning and look out our bedroom window to see what kind of day its going to be, we are making an observation. If we happen to glance at the thermometer tacked to the jamb, we are making a measurement. In either (or both) cases, what we have acquired is information about the weather. This example, intentionally mundane, seeks only to illustrate what information basically is.

Information means some kind of intelligence, some kind of data that has been put into a form that is suitable for communicating. That is, information may be thought of as the content of some signal or message which one person might wish to send to another.

In the analog world all signals are continuous and are continuously variable. If we monitor our conversations with a microphone attached to an oscilloscope we would see that our speech is carried by sound waves which have a distinct, undulating shape. If we modulate our voice, raising or lowering it, speaking faster or slower, these sine waves will exhibit continuously varying wavelengths and amplitudes each dependent, as it were, on our every word.

In the visual world, the most obvious analog signal is the classic photograph. The 3 x 5 snapshot exhibits a certain of amount of information. If we wish to see more, we take the negative from which it was made and create a bigger print, say an 8 x10, which will enable us to discern considerably more detail.

Since the detail within light rays which entered our camera and which exposed our film is infinite, in principle we could enlarge the analog image until its resolution matched the real world original. (In practice, of course, this turns out not to be possible because at a certain magnification the structure or grain of the emulsion [our "instrument"] intrudes and instead of seeing more useful information we see less.)

Whether, then, the recording we’re making of some part of the world is audial, visual, or both, the apparatus we use, if they are analog, will copy their subject in a continuous, dynamic, and highly faithful way.

So. If all of that’s true about analog, why do we need even to talk about digital, the other "way" to communicate information? What possible advantages could digital have and how are they different from analog’s? In answer to both of these questions, consider the following little paradigm.

Two people walking in opposite directions are transmitting information to one another by using walkie talkies. Initially, as the experiment begins, they are separated by only a hundred yards or so and thus are able to hear one another’s words with complete clarity. As they move farther apart, however, the sensitivity and power of the walkie talkies begins to be challenged and, as the signals issuing back and forth become weaker, their clarity is increasingly disturbed by extraneous noise and static.

At first only a word or two in any transmitted sentence may be garbled by the static and hence the receiver can do a pretty good job of filling in what she didn’t in fact hear. But as the distance between the two instruments is increased still further, the static becomes so bad that maybe only one word in ten gets through and thus no dependable extrapolation is possible. The signal to noise ratio has been destructively inverted and neither person can make out the other’s messages. Both are at a loss for words.

But what if instead of words which, after all, are strings of analog sound, they struck upon the notion of signaling one another by flicking the "transmit" button on their instruments on and off? Although each instrument can no longer reproduce words sent by the other, they both can still detect on/off signals.

Now it is extremely important to notice that when communication resorts to a binary vocabulary of on/off much more has changed than merely the "language" of the transmission. The strings of words sent back and forth through the two walkie talkies were the signal - other than their transduction, they required no further translation. But when a signal is digitized something quite new and very important happens to the communication process - the signal is encoded into a series of electrical pulses which, after being transmitted, must be decoded if they are to be understood. What this really means is that the signal is no longer being sent at all and instead of a string of actual words being sent we now utilize a string of zeros and ones which, by their unambiguous simplicity, can stand for the signal while they quite robustly withstand the noise.

A signal which we wish to send from point A to point B doesn’t need itself to be sent if we digitize it. Instead what point B gets is a long series of on/offs which, if the recipients at point B know the right rule can be decoded in such a way that a new copy of the signal may be assembled from it.

As a few more illustrations may illustrate, the power of this approach is enormous. If we do not have to record (either audially or visually) the actual content of a possibly very complex signal, we merely have to create a series of small, literal instructions for its exact assembly someplace else. Think of the Pathfinder mission to the Martian landscape and the stunning pictures it returned of the surface of a planet some 170 million miles away....

Digital is so easy because it’s so tedious; it exploits time as its medium of differentiation and spatial as well as temporal information is coded accordingly. Let us suppose that this page is a digitized message and that it’s going to be projected onto a screen which uses a projector which has 480 pixels by 640 pixels (the redoubtable VGA). All the computer beneath it has to do is follow in rigid sequence a series of 307,200 very simple instructions. Starting at the upper left hand corner of the screen the display device gets the very first of the 307,200 sections into which the screen has been evenly divided and does one of only two things: it turns the pixel on (making it white) or it leaves it off (black). And that’s it. (Yes, if it’s a color display, there’s one more part to the instruction: On or Off + what color; but this a quantitative, not qualitative, addition.)

Because a computer can carry out a very large number of instructions as simple as this in a very short period of time, a screen like this one can be written and rewritten more than 30 times every second. And if the image, instead of being a simple page of monochromatic text, were instead to be a fabulously intricate photograph of, the computer would never know the difference because what is being displayed is not the picture - it’s just the pixels, one by one.

And since that is so, this page or that picture could be distributed to thousands of computers feeding thousands of projectors and each and every one of them would be able to produce (but not reproduce) the same image each and every time.

The analog image is a copy (in the true sense) of some original. It has recorded the actual, cursive shapes of the letters on this page and the white spaces beside and beneath them. But the trouble with a copy is that all you can do with it is copy it again and again. And, as each successive copy is made some qualitative aspect of the original is degraded or lost.

A digital image, in contrast, can be edited. Since the "ON" signal from one of our walkie talkies has only a single frequency, all other frequencies can be filtered out, thereby liberating the coded signal from its enshrouding and extraneous noise.

Finally, the two methodologies are distinct not only as a code is distinct from a signal but also as to their resolution. In the analog world this resolution is always continuous and infinite. The digital world never is.