When we think of computing, we assume that digital information is the paradigm of technological perfection, in comparison to old analog devices, in which the information degrades along the time. This is not correct, and in some cases far from the truth. We trend to consider the digital technology as the most accurate way to transmit, store and reproduce any kind of information, and take for granted that the analog information is inconsistent.
Regarding the degradation of the information that occurs in magnetic tapes or audio vinyl, it is only on account of physical reasons. The magnetic heads steal some of the information of the tape every time they play it, and the record player needle erodes the disc surface. The point is that the information stored in them is in principle very good, it is realistic and natural (although not free of some imperfections). With digital media, the information doesn’t degrade by continuous playback, but the material in which the information is stored can degrade itself, and much earlier than its supposed lifetime, beginning showing some defects. What’s more, digital information always seems to be coldly perfect, unnatural:
Everything we perceive is analog information, a continuous signal detected by our senses and processed by our brain. However, digital information is not continuous, it is a binary representation of the information. For example, when we listen to music from a CD, the digital information must be converted to analog signal the loudspeaker can play and we can perceive.
When we store information using analog devices, as audio tapes, cinema, they record the information in a continuous way, and the limitations come from the ability of the media material to pickup all the information contained in the signal. The analog problems are associated with the progressive wear of the media as we play it. However, the analog signal can be as good as digital information, and somehow always seems to be more realistic and natural.
When we store digital information, we must convert it to bits, concrete units. The problems come when we have to decide how many bits we will use to capture the signal information. For example, if we decide to use 1 bit to store colour information, we will only be able to represent 21 colours, that is, black and white. If we use 8 bits, we will be able to represent 28 colours, that is 256 colours. As the human eye can perceive about two million colours, this is perceived by us as defective or inaccurate information. Anyway, we can decide to use 32 bits and represent 16 million colours, but, is everything we see only colours?, what about hue, brightness, etc?. Apart from the ability of the monitor to show the binary colours once they are converted into analog information. So, digital information is not necessarily perfect.
In addition, if we try to store information along the time, like music or movies, apart from the number of bytes to store the information, we will have to decide how many snapshots of the analog signal we will take per time unit. This is also known as sample rate. A sample is a digital image of the analog information occurred in a single moment. So, depending on the number of digital “snapshots” that we take per time unit, the information will be more or less accurate, although it will never be equal to the original signal (that would need a infinite number of samples).But one of the most important problems comes from the D/A and A/D conversion. The devices that convert analog information to digital signals are far from being perfect. They are machines, so they are limited by their construction and the quality of their materials. So, they are actually unable to capture all the available information. For example, a picture is not only its colours, there are other vaguer factors, like brightness, contrast, saturation, hue, angle, etc. that are very hard to accurately summarize as digital 0’s and 1’s. The first problem is that we may lack of the technology capable of detecting such subtle information.
The second problem is that, even in case we had such technology, storing the information in the necessary detailed way as to reach our senses’ precision would need a huge amount of bytes, and therefore a big storage media, extremely expensive. Examples are the Audio CD and DVD. They have no capacity to store all the information our ears or eyes can detect, but they can accommodate enough data to make us feel that the music and video are of a “very good” quality.
Another problem is the D/A (digital to analog conversion). When we capture digital information, we break a continuous signal into small pieces. However, in order to output that information to the analog form we can perceive, a digital-analog converter must rebuild the digital pieces to create a continuous signal, that will never be equal to the original. Besides that the conversion must be done in real time, so there has to be a balance between the necessary accuracy rebuilding the digital signal and the necessity to output the information quickly enough, or handle the contingency of any error during the conversion.
Because the digital transmission is a weight of errors. Every digital circuit must have an error correction module aside: It is very easy to get a 0 instead of a 1 and vice versa. This must be at least detected, so the bit can be sent again, or even repaired by guessing which was the signal sent. The digital and analog conversion is full of errors, not so much that the information is broken, but enough to cause a perceivable distortion (the quality of the converter and its components is critical). Then, the error correction circuits work by detecting the error, and requesting again the information (something usually impossible), or replacing the incorrect bits with the most likely ones, creating a simulation of the real information.
So, we can see that digital information is more a trick to our senses than perfect information. It’s like cinema, in which we are not watching a continuous film, but about 30 photos a second, creating the impression of motion pictures. The digital information tries to reach the threshold from which our senses cannot detect the sample rate, or the digital accuracy of 0’s and 1’s used. However, reaching that threshold has proved to be more difficult that it seemed. Our senses are extremely precise and perceive a wide range of information, and be able to trick them requires a great technology and knowledge we haven’t acquired by now. The best example of this challenge is a digital movie. It is relatively easy to distinguish between a conventional movie and a digital one. We can even detect any digital image added to a conventional movie. “That is made by computer”, is the typical comment.
The day we can’t detect the trick, we will be able to say that digital technology has reached the goal. That is, we can consider digital information perfect when we can’t detect its imperfect nature.