Recording Techniques (Part 4)
PART 4: David Mellor takes a first step into the wide world of multitrack recording with an explanation of cassette-based 4-track techniques.
There are essentially three ways to learn about multitrack recording techniques. The first is to be lucky enough to land a job as a trainee - they used to call them 'tape ops' - in a recording studio. With a desire to learn, contact with helpful and knowledgeable engineers, and an ability to work all hours for minimal pay, there is a direct road to potential success for the fortunate few. To avoid the possibility of the Sound On Sound offices being flooded with enquiries about how to find such a job, the recommended ways are to make contact with as many recording studios as you can and hope that your letter falls on the manager's desk at the very moment a vacancy arises. Alternatively, you can go to your local Careers Officer who, you may be surprised to find, will be very open-minded on the subject. Obviously your chances are better if you live in the big city; the bigger the better.
The second way to learn about recording is to find yourself a training course. There are quite a few about these days - look out for their adverts. Be warned that some courses are very expensive.
The third way to get into multitrack recording, and perhaps the most accessible route, is to buy yourself a cassette-based multitrack recorder. If you already have an instrument or two, or your mates have, then you could start producing demo recordings for a not unreasonable sum of money. There are some multitrackers (as I shall refer to cassette-based 4-track recorders, for want of a better name other than 'portastudio' - Tascam's brand) right at the entry level price which give excellent sound quality - but you'll have to pick them out from those that are not quite so hot.
In this part of my 'Recording Techniques' series, I hope to explain exactly what these multitrackers are, what they do, and the basic techniques you need to master to achieve a good result. These basic techniques will still prove useful as you progress up the scale to 8-track, 16-track and, hopefully, 24-track recording.
Let's look for a moment at the design of the cassette tape itself. It certainly wasn't designed for multitrack recording, although it has ultimately proved itself in this application.
The original idea behind the cassette was to be able to record fairly low quality stereo sound on both sides of the cassette. As you know, after filling the first side you can turn it over and start recording again. This means that there are four separate tracks of audio on a normal cassette, but the tracks on side two run in the opposite direction to those on side one.
The multitracker uses a tape head with four record/playback elements stacked up so that all four tracks can be recorded in the same direction, spread across the width of the tape. Since the full width is used on each multitrack recording you make, it is not possible to turn the cassette over. Well you can, but you will hear all four tracks running backwards.
Although with modern design and noise reduction techniques normal stereo cassettes can give a fairly good sound, they are working right at the limits of the magnetic tape medium. Any slight wear or misalignment results in that well known muffled and uneven cassette sound. Most multitrackers get around this with a simple dodge - they run the tape at twice the normal speed. This really does give a great improvement in the stability and consistency of the sound. It means that the tape is used up twice as quickly though, and what started out as a C90 cassette on a normal stereo machine ends up as a cassette lasting only 22½ minutes - enough for about four pop songs. It's a good job that cassettes are reasonably cheap to buy these days.
And while I'm talking about cheapness, it is important not to go too cheap but to buy the correct type of cassette for your machine. This usually means a Type II variety, such as TDK SA, but your machine's manual will tell you what's recommended. The sort of things that you will be doing with your multitracker will demand a lot from the tape. If you weigh up the amount of time you put into your recording, and the cost of the tape, then it makes sense to buy the very best.
Besides running at twice normal speed, the other quality enhancer used on multitrackers is noise reduction. For stereo cassettes, Dolby B is the de facto standard. It works very well with a properly aligned machine in a good state of repair. With multitrackers, Dolby B doesn't reduce the noise level nearly enough, only by about 10 decibels. Dolby C and dbx are the systems generally in use. Dolby C, found on Fostex multitrackers, gives a very clean sound; dbx, used by Tascam, Yamaha and others, gives less noise during quiet sections but is not as transparent while the music is playing. As always, the best advice is to listen for yourself and decide what's best for your sort of music.
The evolution of the multitracker has split into two lines, one of which more or less follows conventional recording studio practice, and the other is dedicated purely to budget 4-track operation. The former is, obviously, a better education into multitrack ways and will provide a springboard of basic knowledge from which to leap to more advanced systems. The other system is less expensive to implement, and hence is found on less expensive machines. It can still produce excellent results.
Figure 1 shows a typical control panel from the first type of multitracker - the distinguishing feature being that it has one set of equalisation controls per input channel. It's like a recording studio in miniature, with all the features you would expect arranged in a similar way. This isn't a diagram of any particular model, but it has the typical features you would like to find on a high end machine. As you come down the various manufacturers' ranges you will find that some features are dropped, and some are combined with others, making the machine less versatile.
Model 1, as I shall call it, combines a 4-channel mixer with a 4-track cassette tape recorder, all in one box. Starting with the input channels, at the top of each channel is a GAIN control. This boosts the signal from whatever it happens to be from mic or instrument up to a level suitable for further processing. Not all multitrackers have gain controls, some have a choice of LINE input (low gain for a strong signal) or MIC input (high gain for a weak signal). Above the gain control is a switch that takes the input to the channel from the input connector or from the output of the cassette. Switch to INPUT when you are recording an instrument on this channel, switch to TAPE when you want to hear the tape track whose number is the same as that of the channel.
The EQ (equalisation) controls tailor the incoming sound to your particular requirements. Here I have shown only two fixed frequency controls per channel (HF and LF), each with variable cut and boost. Sophisticated machines can have variable frequency, too. Notice that my Model 1 has four EQ sections, one for each channel. The alternative type of multitracker has just two EQ sections configured in a rather different way.
The AUX (auxiliary) SEND is an extra mixed output from the mixer section. The auxiliaries of all four channels are mixed together and emerge from a connector on the rear panel. The most frequent use for the aux send is to send signal to a reverb unit, as mentioned later.
The CUE control is used for listening to the signal in the channel, independent of the level being recorded on tape. You can either monitor the input signal (INPUT) or whatever is recorded on the corresponding track of the tape (TAPE).
Our hypothetical unit has four MIX BUSES and a BUS SELECT switch. Each mix bus is simply a piece of wire, or copper track on the printed circuit board, that picks up signal from any or all of the channels and takes it along to a single tape track. Four buses obviously correspond to four tracks. Here the switch sets the channel to buses 1 and 2, or 3 and 4. This means that if you select 1 and 2, the PAN control pans the signal between track 1 and track 2 of the tape. If the pan control is set all the way to the left, the signal goes only to track 1. If the pan is all the way to the right, it goes only to track 2. With the pan control centred, the signal is sent in equal amounts to both tape tracks.
Some multitrackers only have two buses. You can still record on all four tracks (it would be pointless if you couldn't!), but you are restricted to recording a maximum of two tracks at a time. This is not so good when recording bands playing live, although individual types of two bus multitracker may have ways of getting around this limitation.
At the bottom of each channel is the level fader. I'll let you guess what this is for!
The other, non-channel, items include the MASTER fader (which is actually a stereo fader) and the RECORD READY buttons and indicators. The transport controls. Fast Forward/Play etc, are as you would expect on any tape machine but are deliberately not shown in the diagram. A couple of items missed out in this section will be mentioned as I progress.
Figure 2 shows all the bits you need for a basic setup, and how they fit together. Notice that there are two main audio paths out of the unit: the MASTER and the MONITOR paths, both of which come in left and right channels for stereo. The master outputs are connected to the stereo tape recorder, which may be a reel-to-reel machine (best) or conventional stereo cassette (while you are saving up).
The amount of signal (level) that comes out of the master outputs is vitally important for a noise-free and distortion-free recording, and is not related to the level that comes out of the monitor speakers. The monitor outputs are provided for you to listen to what's happening on the master, but you can set the level of the speakers to suit yourself rather than to suit what the tape in the track laying or mixing process requires. You can also turn the monitor level up and down without affecting the signal going onto the tape in any way.
The other signal path out of the multitracker is the auxiliary send, or AUX SEND for short. As you can see, this goes to a reverb unit in our example, the two outputs from the reverb returning a stereo signal to the AUX RETURNS. There are two aux returns on any self-respecting multitracker, preferably with a level control - probably a single stereo control rather than two mono ones. Some multitrackers don't have level controls on their aux returns. An uncontrolled aux return is sometimes called a 'bus in' input. The bus in, which has nothing to do with a Number 48 or a 60p fare, mixes directly with the master output to the stereo recorder. Therefore, to add reverb to your sound mix in the correct quantity, you need a reverb unit with an output level control. This is a point that needs to be checked when you are buying suitable equipment, otherwise you will have to accept excessively noisy mixes with an uncontrolled reverb level output.
The great advantage of the multitracker is that it is so handy. Notice how few cables there are in this system, and it can produce amazingly good results. (Be prepared for a shock, cable-wise, when you eventually upgrade to a larger number of tracks!)
Am you a live or a MIDI musician? There's no prejudice here, because you can make good music either way on a multitracker. But let's concentrate on the guitar-picking instrumentalist with a bit of vocal ability. We'll get him, or her, started by including a bass - guitar or synth - and a basic drum machine in the package.
Perhaps the most important factor in obtaining a good 4-track recording is the planning. Sometimes, you have to make an initial personal demo of a song in order to develop the arrangement. Then, with the instrumentation sorted out, you start again to make the definitive, properly planned, 'master' demo recording.
Following last month's article on simple arranging techniques, I shall assume that the arrangement will consist of drums, bass, chord backing, lead instrumental line, and vocal. I'll add, just for fun, two vocal harmony lines. This adds up to seven different musical parts. It sounds like three too many for our poor little multitracker to cope with - but with a bit of a squeeze, it can.
The first track to record has to be the drum machine, simply because a musician can synchronise to a drum machine but a drum machine can't sync to a musician. Besides the musical material in the drum track, you will need something extra - a count-in. Without a count-in, of course, you won't know when to start playing. I find that the best count-in is a quarter note (crotchet) hi-hat pulse lasting two bars, but missing out the last beat before the music comes in. We'll need that gap when we mix later.
Following the instructions provided in your multitracker's manual, the drum machine should be routed to track 1, a recording level set with the fader, and the track recorded. You could equally well start from track 4 and work backwards. Don't start in the middle or you are more likely to run into feedback problems when we start bouncing tracks.
The second track will be our first overdub. This will probably be the bass line, but could be the chord backing. The channel to which the bass is connected must be routed to track 2. Since you want only the bass to be recorded on track 2, only channel 2's fader and the master fader must be raised. You can hear the drums on track 1 by setting the monitor to CUE and raising the cue level on that channel. Set the cue level on channel 2 for a good working balance. The cue levels that you set do not affect the recording.
This procedure is repeated for the third track. Now you have only one tape track left, but four more parts to record. The thing to do now is to bounce those three tracks onto track 4. 'Bouncing' simply means mixing a number of tracks onto another track on the same tape. To achieve this, channels 1, 2 and 3 are all routed to record onto track 4 and are mixed together the way you will want to hear them in the final mix. Although you can have as many goes as you like at the bounce, once you are satisfied and proceed with the next stage, you are committed to what you have done. You can't change the levels later. The next step is to erase tracks 1, 2 and 3!
(Sometimes when you are bouncing tracks, you get a high pitched whistle as soon as you enter Record. This is due to feedback in the record head itself. It can be cured by simply lowering the recording level.)
With the first three tracks now freed, you are ready for the next step. Since you have two harmony vocals to add, I would suggest recording these next onto tracks 1 and 2, then bouncing them both to track 3 and erasing tracks 1 and 2 again. The final stages would be recording the main vocal onto track 1 and the lead instrumental line onto track 2. Finished!
This is a fairly simple example. It is quite possible to record more than seven parts, in fact up to 10, without bouncing any track more than once. To do this, you need to add a live part each time you bounce. So when the drums, bass, and chords were bounced to track 4, there might have been another part included at the same time. Also a third harmony vocal could have been added. It is slightly more difficult to add these extra parts, because when you are recording parts individually it is possible to punch in, or drop in as it is sometimes called. All multitrackers these days have a socket for a punch-in footswitch, with which you can play the tape and punch in and out of record mode at any time, re-recording over any wrong notes. Punching in will be explained in more detail in a later part of this series, but for the moment I'll just say that you have to do it in the right place. When you are recording just one instrument, it's easy. If several are being mixed together it is unlikely to work.
Working with MIDI expands the capabilities of the multitracker. Most sequencers will synchronise to tape. A timing pulse is recorded on track 4 and thereafter the sequencer will sync to that. This leaves you with only three audio tracks to record music on, but you still have as many input channels as you had before, and sequenced instruments can supply other parts. More on this when I discuss multitrack and MIDI later in the series.
Mixing is the part of the recording process where your raw tracks are turned into a finished master. It is a process that demands close attention to the sounds coming from the speakers and painstaking adjustment and re-adjustment, even with just four tracks.
The first stage of mixing consists of listening to the individual tracks, some of which will contain more than one instrument. What I like to listen for in the individual tracks is the character of the track, to find out what its essential components are, and to see what qualities it has that are not really required. Even with just four tracks, it is easy to create a mix that sounds like all the instruments are fighting with each other for their share of the space between the stereo speakers. What we really want is to get them all working together, each making a valuable contribution to the whole sound.
To optimise each track, I experiment with the EQ. I consider whether a sound benefits by having a particular frequency range boosted. If it does, I do just that, and just enough to make a difference. If it doesn't, then I try cutting that frequency range. For example, if the instrument doesn't make any useful contribution in the high frequency range, then turn down the high EQ and reduce any HF clutter it is producing. Another example: the bass instrument is producing the bass frequencies. Other bass-heavy instruments are probably congesting the bottom end, so lighten them up.
After experimenting with EQ on the individual tracks and having decided where to pan the tracks in the stereo picture, see how much reverb each track needs just by itself. At this stage, add no more than makes a difference. This individual approach will help tune in the ear to the capabilities of each track and what they may be able to do for the mix when you raise all the faders together.
At this stage, it's up to you. You are the only person who can tell what adjustments improve the mix, and what makes it less good. How long will it take to arrive at the perfect mix, and be ready to transfer to stereo tape? Even with just four channels and a reverb unit, don't be surprised if it takes a couple of hours. When you move up to more tracks, it may take longer still.
Assuming the mix is right, you have all your fader levels set and have planned any level changes you want to make (not forgetting to set the record level on your stereo recorder), take out a Chinagraph (wax) pencil and make a mark next to the centre of the fader knob on each channel. Remember that hi-hat count-in? Play the tape up to the seventh beat of the count-in and pause the tape, so that when you start, you miss out any of the extraneous sounds that accumulate before the beginning of the music (if you have a reel-to-reel, then you will be able to edit them out rather than doing it this way). Set the stereo recorder to record, and hit the play button on the multitracker.
As the final notes die away, pull down the channel faders first, and the master fader as the reverb dies away. This ensures that your master ends in silence rather than noise and hiss. The Chinagraph marks are for when you make a hash of it and want to do the mix again. Just reset the faders to the marks and have another go.
Hopefully, this article will have explained just a few of the techniques of multitrack recording in a simple way. But there's a lot more to it than can be covered in a couple of pages. Look out for further installments in the near future which will explain how multitracking takes place on a grander scale.
Feature by David Mellor