Paul White explains the basic principles of tape recorders to help you make better recordings.
Despite the many advances in technology in recent years, most recording facilities rely heavily on the same, basic analogue tape recorders that have been part of our lives for the past four or five decades. Paul White analyses analogue.
While I wouldn't pretend that designing a tape recorder is particularly easy, using one is fairly straightforward. Like television, you don't need to know too much about how it works to make use of it. However, there are certain simple rules and principles that, if applied, can make an enormous difference to the quality of recordings made with any level of equipment.
There are two distinct types of tape recorder on the market: the cassette machine and the open-reel. Cassette systems are convenient and relatively inexpensive but, because of their relatively low tape speed, they can't produce recordings to the same high standard as professional, open-reel models. Nevertheless, cassette technology has advanced considerably over the past decade, and even a domestic hi-fi cassette recorder can be used to make very creditable recordings.
It helps to keep things in perspective if we look at the complete recording chain, where we should first examine the role of the microphone. Sound energy propagates as vibrations in the air and the microphone intercepts some of this energy, converting it into a varying electrical signal, whose amplitude represents the original vibrations. The key word here is signal, and I'll be using that a lot in future. Whenever a sound is converted into an electrical form, it is referred to as a signal, not as sound. In the case of a microphone, this signal is very small and must be amplified (made larger) before it becomes useful. Microphones will be considered in some detail in future issues of RM, but in the first instance, the general principles will suffice.
Though complex in their workings, tape machines operate on a very simple principle: an electrical signal (such as the output from a microphone) is amplified and then used to vary the magnetic field in a part of the tape recorder known as the record head.
A system of motors, rollers and guides is used to draw the recording tape across the head at a constant speed and the record head imparts a magnetic charge to the particles that make up the tape's surface as it passes. In other words, the original signal from our microphone has been converted into an equivalent magnetic signal which is stored on tape. If the tape moves at 1 7/8 inches per second, as is the case with domestic cassette machines, it follows that the information contained in one second of sound is spread over a piece of tape 1 7/8 inches long.
If you were to look at the surface of a recording tape, you'd see that it is made up of countless tiny particles of magnetic oxide, each one of which is capable of retaining magnetic charge. The recorded signal is really the statistical summation of the magnetic charge on all the particles passing over the tape head at any instant in time. This is the main reason that higher speed professional machines produce higher quality recordings — during each second there are more magnetic particles passing over the head, which produces a statistically more accurate representation of the original signal. Those of you versed in basic statistics might deduce that a wider tape would also result in a more accurate recording for the same reasons — and you'd be right. This is an important consideration when we come to discuss multitrack recording.
To replay the tape, it is rewound and then drawn over the replay head, which works in exactly the opposite way to the record head; the changing magnetic field on the tape produces an equivalent electrical output at the replay head, which can be further amplified until it is powerful enough to drive a loudspeaker.
In domestic machines, the same head is often used for both recording and playback, though it obviously can't do both jobs at the same time. A further head is used to erase or wipe the tape of unwanted recordings; this is known as the erase head, and it works very much like the record head except that it records a very high frequency signal, out of the range of human hearing. This re-charges the particles on the tape's surface and erases the previous signal — something like the tide washing away sand castles! This explanation is oversimplified and omits several technical details, but from the operator's viewpoint, it will suffice. Figure 1 shows how the heads are arranged in a typical tape recorder.
The loudspeaker functions in exactly the opposite way to the microphone and converts the electrical energy back into mechanical energy in the form of air vibrations. This completes the recording chain from original sound, through sound storage to sound replay. Figure 2 shows the simplified recording chain from microphone to loudspeaker.
However, the quality of reproduction relies on every stage of the process being linear; in other words, whenever a signal is amplified or converted from one form of energy to another, the conversion must retain the exact nature of the original signal. In real life, this degree of perfection is unattainable and some distortion is inevitable. Even so, the amount of distortion introduced by well-designed equipment is so small as to be inaudible.
When making a tape recording, there are two main enemies — noise and distortion. Neither can be eliminated completely, but correct use of the equipment can ensure that both are minimised.
When I talk about noise during recording, I don't mean the musicians, (though they often qualify). I mean the background hiss due to the electronic circuitry and, more importantly, hiss due to the recording tape itself. Because the signal recorded onto tape is statistical in nature, there is always some random element, which manifests itself as a background hiss. Even blank, unrecorded tape creates a hiss if you play it, as all the randomly-charged magnetic particles conspire to create a random electrical signal — hiss. Referring back to the 'sand' analogy, tape noise can be visualised as the roughness of the sand when the tide has washed away all other features.
The point to remember is that tape hiss is more or less constant, so the way to minimise it is to ensure that the signal you do want to hear is recorded a lot louder than the hiss you don't want to hear. But as in most aspects of life, there's more to it than that, otherwise you'd simply record the highest imaginable signal onto tape and swamp the noise into insignificance. The sad truth is that there's only so much level you can record onto tape before you run into another set of problems. There comes a stage when the magnetic particles on the tape's surface are fully charged, or saturated, and simply can't hold any more. In consequence, if you try to record a higher level signal than the tape can accept, the signal will be distorted, and when you come to play back the tape, this will be quite audible, making the sound appear fuzzy, indistinct and generally unpleasant. The skill is in setting the record level as far as you can above the hiss level but low enough to avoid distortion creeping in on any loud passages. For this reason, most tape recorders are fitted with level meters with the overload region marked in red.
Having painted what must seem to be a grim picture, I should emphasise that the distortion problem isn't as serious as it might appear — it doesn't all happen at once, but creeps in progressively as you increase the signal level and push the record level meters further into the red. In practice, short excursions into the red overload area of the meters will be inaudible, but with any tape machine, it's wise to make a few experimental recordings and then listen to the results. After a short while, you'll develop a feel for how much signal level you can get away with.
One tip is to record bright sounds, such as cymbals, at a much lower level than you'd record more normal sounds. Bright sounds tend to overload the tape at a much lower level and distortion may be evident even before the meters go into the red.
Before leaving the subject of noise and distortion, it should be made clear that tape isn't the only culprit; every electronic circuit ever invented suffers from noise and distortion and the user must ensure that the signal being processed sits somewhere in the usable window between the two. This useful window — between noise at low levels and distortion at high levels — is generally expressed as dynamic range, and is measured in decibels or dBs. The significance of this must be taken onboard, as it crops up time and time again in all aspects of audio engineering from basic tape recording to broadcast, editing, record and CD production and signal processing.
As mentioned briefly in last month's roundup of cassette multitrackers, most modern tape recorders are fitted with something called noise reduction, which is a great help in keeping hiss down to acceptable levels. The most famous name in noise reduction is Dolby, their consumer 'B' or 'C' systems being found on nearly all modern hi-fi cassette decks. The other noise reduction systems you might come across when working with home recording equipment are dbx and Dolby S, though they all have one thing in common — they must be used both during recording and during playback. In technical terms, these are known as encode/decode systems, because they do something to the signal when recording and then do exactly the opposite when playing it back. In theory, this leaves the signal in its original state but reduces the level of hiss.
The way they achieve this apparent miracle is to record vulnerable, low-level signals at a higher level than they normally would be, making them strong compared to the fixed level of tape hiss. During replay, these low-level signals are restored to their original level and the noise is pushed down with them. Louder sounds are not processed, as they are loud enough to hide the hiss anyway. The technical operation of these systems is rather more complex than might at first be imagined, but the end result is a significant reduction in tape hiss.
Most hi-fi cassette machines are stereo, which means that two sets of signals are recorded side by side — one to feed the left-hand loudspeaker and one to feed the right. This is achieved by using a divided tape head, so that two stripes or 'tracks' of recording are laid side by side along the tape.
You may also recall that cassette tapes can usually be turned over to 'play the other side', but this doesn't play the back of the tape, as the term might lead you to believe. In fact, the recording only happens on one surface of the tape and this is divided up into tracks — two running one way and two running the other — rather like the lanes on a dual carriageway. There is no physical division on the tape itself, of course; where the recording takes place on the tape is determined purely by the geometry of the record head. Figure 3 illustrates this arrangement.
Whatever type of recorder you have, check your machine's manual to find what brand or type of tape is recommended by the manufacturer and either stick to that or buy a reputable equivalent. Stay with brand names you know, such as TDK, Sony, 3M, Ampex, Maxell, Scotch, Memorex, BASF and so on, even though they might cost a little more than 'cheapo' brands. All the better machines use 'Type 2' tape, which is sometimes also referred to as chrome-type. Though it costs a little more than standard ferric tape, the results are cleaner and brighter. Resist the temptation to use cheap tape from the market, unbranded tape, or computer cassettes, as these all too often sound dreadful and quickly shed brown oxide all over the tape heads and guides.
Tape should be stored in a clean environment, out of direct sunlight, and at room temperature. Avoid leaving tapes in the car, as direct sunlight can heat them considerably; also take care to avoid damp or dusty atmospheres. Handled with care, a good tape will last for many years and can be reused dozens of times.
Now we've covered the basics, next month I'll be looking at a typical 4-track recorder.