The underlying concepts of multitrack recording are explained from the bottom up.
The underlying concept behind multitrack recording is very simple, yet it has completely revolutionised the way in which music is recorded. Paul White explains the essentials of multitracking.
Before delving into multitrack, it is helpful to outline exactly what we mean by the term 'track' in the context of multitrack recording. Last month, we looked at the basics of tape recording when it was explained that, during recording, a so-called recording head is used to make a magnetic 'imprint' on a moving tape which, when replayed, will reproduce the original electrical signal. A single electrical signal recorded via a single head produces one track on the tape, and though this is invisible to the human eye, if we were able to see magnetic patterns with the unaided eye, we would see one continuous track or strip of magnetic information running along the length of a recorded tape. Nowadays, most domestic tape machines are stereo, and for those unfamiliar with the underlying concept, stereo is actually made up of two different recordings, one of which feeds the left hand speaker or headphone and one of which feeds the right. Not surprisingly, these are known as the left and right stereo signals. These two recordings are positioned side by side along the tape in the form of two parallel tracks, and they are recorded using a special head which is really two heads in one. Similarly, the playback head and the erase head actually comprise two heads mounted in the same physical block, one above the other.
In reality, most non-professional tape machines utilise the same head both to record and to replay the tape, which simplifies matters, especially when it comes to multitrack recording, as we'll soon see. However, the erase head is always separate regardless of the type of tape machine.
With a standard cassette, we can turn it around when it has finished and record more music in the opposite direction. This is not, contrary to some beliefs, recorded on the other side of the tape but on the same side, running in the opposite direction, as was explained last month. Again, if we had eyes that could see magnetic recordings, we would see four parallel tracks on the tape, two running in one direction and two in the other, rather like the lanes of a dual carriageway. Figure 1a shows how the track would appear on a standard stereo cassette. It should be stressed at this point that the track layout on the tape is determined entirely by the layout of the tape heads — there is no physical boundary between tracks on the tape itself. The tape is simply a long plastic ribbon with magnetic material covering the whole of one surface. It is impossible to record on the other surface of the tape, as this is not coated with magnetic material.
The concept behind multitrack recording is that the tape machine has the ability to record several parallel tracks across the full width of the tape (anything from 4 to 48, depending on the model), and these may be recorded (or erased) independently of each other, allowing a piece of music to be built up in layers rather than recorded all in one go. As with the stereo machine, the tape heads are actually multiple affairs, with the heads relating to the different tracks stacked one above the other; Figure 1b shows how the tracks run on an 8-track tape. After the first track (or tracks) have been recorded, there has to be some way of hearing these while recording new parts, otherwise there would be no way of keeping the different parts in time with each other. This is no problem, as we have already seen that a record head can also function as a playback head. So as long as the tape machine is designed with this purpose in mind, there's no problem in playing back, for example, track number 1 while recording a new part on track number 2. This process of listening to what we've already done is called monitoring, while recording a new part on a different track is known as overdubbing. These terms will arise regularly and it is imperative that their meaning is understood.
If you think about it, it's essential that the record head is used to monitor the existing recordings, because if a separate playback head were used for the purpose, the new recording would be out of time with the existing recording, due to the short time the tape takes to travel from the record head to the playback head, which may be several centimetres 'downstream'. This can be seen clearly in Figure 2. By using the same head to record and to play back, all the overdubs remain exactly 'in sync'. All cassette multitrackers and the vast majority of semi-pro open-reel multitrack machines use the same head for recording and playback, but when you move into professional circles, the machines tend to have separate playback heads. When overdubbing, the existing tracks are still monitored from the record head in a special sync mode, but the operator has to remember to switch to this mode manually. Semi-pro machines are easier to use as, in effect, they are in permanent sync mode, so the operator never has to give the matter any consideration. Figure 3 shows the monitoring arrangement during an overdub.
Deviating slightly, each track on a multitrack recording is essentially mono, but stereo signals from stereo sources (such as a stereo microphone setup) can be recorded onto pairs of tracks.
Being able to build up a recording in layers offers several real advantages, not the least being that one musician can play several parts in the same piece of music. Furthermore, when the recording is finished, the levels of the different tracks can be balanced to give the required musical result. In fact you can change much more than the levels of tracks — you can change their left/right stereo positioning, you can add effects and you can use tone controls to change the sound. But, as these processes are more to do with mixing than recording. I'll save the full explanation until next month when we cover mixing.
As if this flexibility weren't enough, multitrack recorders also allow us to 'patch up' individual tracks, using a process known as 'punching in' or 'dropping in'. Studio professionals always seem to refer to the process as dropping in, while Japanese multitrack manuals prefer the term punch in. Both terms mean exactly the same thing, so remember them both as they'll crop up frequently. The best way to explain the process is by example. Let's assume that you've recorded a vocal part on track one and everything is fine except for one word, which is pitched flat. Rather than doing the whole thing again, we can punch-in to replace a single word or a single phrase, the only proviso being that there should be a slight gap before and after the section we're going to replace. Essentially, punching in and out is simply a method of putting the tape machine into record without stopping the tape, and most multitrack machines allow us to do this by pressing the record and play buttons together as the tape is running. There is a 'record enable' (sometimes called ready/safe) button associated with each tape track, so that when we do drop in, we only do so on the track we have chosen — the others remain 'safe'.
During the run up to the drop in, the existing recording is monitored, usually via headphones, so all you have to do is sing or play along with it. The monitor mix is usually a mixture of the musical backing plus the original version of the part being worked on. At the right time, the machine is punched into record and the monitor for the vocal track, for example, now switches to the new part as it is being recorded. When the end of the section is reached, the machine is dropped out of record by hitting stop, pause, or — on some machines — play. You can then rewind the tape and see if the new section was satisfactory. If not, you have another go until you're satisfied.
Many of the budget multitrack machines have a footswitch option, allowing the user to punch in and out while playing an instrument that requires the use of both hands. Though any reputable multitrack machine can punch in and out without causing a click on the tape, there is an overlap at punch in, due to the gap between the erase head and the record head. Similarly, when you punch out, there will be a slight gap, as the section of tape between the erase head and the record head will be blank. Some of the more sophisticated (and expensive) machines get around this by using a computerised timing system to switch the erase and record heads at different times to produce what is known as 'gapless' punch in and out. In practice, this works very well, but it is still advisable to punch in and out in a convenient gap rather than in the middle of a word or phrase, as any slight discontinuity in the performance will show up and give the game away. It also helps to time your punch in and out points to coincide with a drum beat, to help hide any small discontinuities.
Open-reel recorders produce significantly better audio quality than cassette models, though the recent introduction of the new Dolby S noise reduction system into cassette models has narrowed the gap considerably. Standard cassette models are available in 4- and 8-track versions while semi-pro, open-reel machines come in all sizes, from 4 to 24 tracks, running on tape from a quarter of an inch wide to two inches wide. Open reel machines have to be used in conjunction with a separate mixing console, while most cassette multitrackers are part of an all-in-one studio package which combines a multitrack recorder and a small mixer. These are often referred to as portastudios, though the name Portastudio is a trademark of the TEAC corporation, one of the leading manufacturers of such machines.
Whether you choose open reel or an all-in-one cassette multitrack system, you will need a stereo tape machine on which to record the final stereo mix, and you'll also need a stereo amplifier and loudspeaker system to monitor your recording as it is being built up and to play back the finished master. It is possible to work entirely on headphones but this is rather restricting, especially when several people are working on the same project.
Furthermore, for any multitrack other than an integrated cassette multitracker, you'll need a mixer — and the more tracks you work with, the larger the mixer needs to be. Next month we look at the fundamentals of mixing.