Recording Techniques (Part 7)
Part 7: David Mellor puts a typical band into a typical studio and looks at the various stages of recording, from initial planning up to instrumental overdubs.
Recording a band in a studio can be a lot of fun, but it will test your abilities over a pretty wide range. Do you, for instance, have the diplomatic skills of a Secretary General of the United Nations; the sleight of hand of a member of the Magic Circle; the persuasiveness of a door-to-door life insurance salesman; the mind-reading capability of an electro-encephalograph? Well, if you can boast all these skills and more, don't forget that you have to be a recording engineer too! Recording a band means dealing with people, and people can be a lot more awkward and demanding than machines such as samplers and sequencers, but the end result can amply justify the effort put in.
As I said at the very beginning of this series of articles, I am going to consider the typical situation where the engineer works directly with the musicians, without the assistance of a producer. Taking the whole spectrum of recording studios into account, from garage level up, this probably applies to the majority. Also, in situations where one of the musicians considers himself to be the producer of the recording, the fact is that the engineer is probably doing most of the real production work, even if he doesn't claim his fair share of the credit! I'm going to look at the recording process mainly from the engineer's point of view, but if you are principally a musician working on the other side of the control room window, then perhaps you'll get an idea of what it's like for the guy, or gal, in the hot seat, and hopefully end up with better studio recordings.
This is where the bit about mind-reading comes in. From my own experience, a request like "I just want a straight cassette copy" actually means "I'd like you to cut out eight bars from the middle, get rid of some of the hiss, speed it up slightly - and can you make another couple of copies while you're at it?" If someone says "It's just a basic track with a vocal and a couple of guitar overdubs," what they really mean is "It's a basic track with a lead vocal, three harmony vocals, several guitar overdubs through a stereo effects unit, and I wouldn't mind a few tracks of that synth you have over there."
It's not as though things always escalate completely out of proportion, but when you budget for a certain amount of time or a certain number of tracks, you can count on having to add on a fair safety margin. If that margin isn't there, then it's the engineer that's going to get the blame! The solution is to talk the project through thoroughly, before even opening the instrument cases. Unless the band is experienced in studio work, you will really have to pump out the information you need, otherwise you are going to have problems later on. Here are a few sample questions that will set the ball rolling:
Have you played this song live very often?
If the song has been part of the band's set for months, then they'll know how it goes (no sarcasm intended) and will probably have a good idea of the studio embellishments they would like to add. If it's a new song, then you are virtually bound to find sections where the musicians are not sufficiently confident, or indeed sections which just don't sound right under analytical studio conditions. The advantage of recording a song which hasn't been played live too much is that the musicians will be more open to suggestions on how the arrangement could be changed for the benefit of the recording. Performing arrangements and recording arrangements should ideally be worked out separately, according to their respective needs.
What is the instrumentation of the song?
Rather than just counting the instruments, you need to know whether the guitarist (and possibly the bassist) use any effects pedals, or if there are any surprises in store that ought to come out now rather than later.
What does the vocal arrangement consist of?
You need to know how many vocal parts to expect, and also establish as soon as possible who the lead vocalist is, so that you can start building up the friendly relationship that you'll need later on.
What do you intend adding to the arrangement that you don't play live?
Make sure you have received a full answer from the keyboard player before you leave this topic - they are notorious for wanting to add endless embellishments. It's not a problem if you know about them now.
Is there anything else you need to tell me?
Don't expect a response to this question, but at least the ball is in the band's court and they can't come back later and say "Well you didn't ask about..."
Perhaps this is starting to sound like a battle of wits between the band and the engineer, when really it ought to be a total collaboration towards achieving a good recording of a good song. Such a total collaboration, however, only comes when you have a good communicative relationship with the members of the band individually - where they respect the engineer's advice and ideas, and where the engineer understands the musicians' aims and needs. But all this takes time (and is rarely properly achieved even within the band itself). The question and answer routine, properly applied, is an invaluable shortcut.
At this stage, you should have a good idea of how the recording will proceed, and you will know whether you have enough tracks for all the instrumental and vocal parts or whether you will have to bounce some tracks together. You will also be able to plan the sequence of overdubs, and think about when those difficult - but important - vocals should be recorded.
Now it's time for the band to unpack their instruments and for the engineer to decide how he is going to tackle them. Note that he is not going to help the band unpack - he will be directing them to where he wants them positioned in the studio, and taking mental notes of the equipment coming out of the flightcases.
In the studio, a good layout for the band is important and fairly easily achieved if you have enough space. The two essential points are that the members of the band must face each other and that microphones, as far as practicable, must only pick up the instrument or amp at which they are pointed. Figure 1 shows a typical layout, which should be fairly self-explanatory. Notice that the bassist stands next to the drummer; this is always the case and helps towards greater precision in the rhythm section. Although the layout of the band is the engineer's responsibility, he should of course consult the band just in case they have some special requirement.
When all the instruments and amps are set up, adjusted, and ready to go, send everyone except the drummer to the pub for an hour (the singer especially will want to lubricate his/her vocal cords). Unless you have worked with the band before, it will take a good while to mike up the kit and obtain a basic drum sound. If you have the other band members around during this time, they will clutter up the studio and control room, make noises with their instruments, and generally get in your hair at a time when you want to concentrate on the most difficult instrument recording engineers ever have to cope with. If there are no problems with the drum kit itself, an hour to 90 minutes should be enough to achieve a good basic drum sound, which fits in with the time scale allowed for demo recordings. Of course, it is quite possible to spend much more time than this getting as close to perfection as you can. And exactly how do you get close to perfection in drum recording? Well, you'll have to wait until next month for that!
Let's assume that you have your drum mics set up and you're ready to start on the other instruments - guitar, bass, and keyboards. The guitarist will almost certainly want to play through an amplifier - hopefully a small one that doesn't have to be turned all the way up to a hundred watts' worth of neighbour-blasting power to sound good.
In Figure 1, the guitarist's amp is shown with a microphone pointing at the edge of the speaker (I shall use the convention of representing microphones by appropriately directed arrows). It may come as a surprise, the first time you experiment with mic positioning on a guitar amp, that a wide variety of different sound qualities can be obtained by mic placement and pointing. The centre of any loudspeaker normally produces the brightest, cleanest, and punchiest sound. This may be the sound you are after, but I usually find that the edge of the speaker gives a sound more akin to actually hearing the amp with the naked ear - from a rather greater distance, of course. Figure 2 shows what happens at the speaker's edge to cause the difference. Microphone choice can be any dynamic or capacitor model that you think sounds good. Bear in mind that it may have to cope with high levels without distortion. Also, if you use a dynamic mic, remember that it will pick up hum from the amp's transformer. Look round the back to see whether the mains cable goes in on the left or right side of the amp, and position the mic on the opposite side.
The guitar amp should be angled so that the sound it produces doesn't fire directly at any other mic in the studio. Remember that the less unwanted 'spill' each mic picks up, the cleaner each track will sound on the multitrack tape - which makes life easier for you. Acoustic screens are a great help in reducing sound spill. They are usually not very thick, as sound insulators go, but they do reduce the annoyance frequencies significantly. Screens come in a variety of types and it's a simple job to make your own with chipboard, foam, and covering material. Large screens with double-glazed windows are better at sound insulation, but they tend to make the musicians feel isolated from each other. Half-height screens are less obtrusive, and still keep their effect because most of the sound is being produced near to the studio floor.
Many bass guitarists do without their own amp these days and direct inject (DI) the signal from the bass straight into the mixing console. If you have very flexible EQ controls on your console, direct injection will work well; better still with a little compression added. If your console only has average EQ, and unfortunately even today most consoles are lacking in this department compared to what is possible, then an amp will make the bass sound a lot more butch and beefy. Combining the miked-up signal from the amp with a DI feed can give the best of both worlds.
With both the guitar and bass amplifiers, it is often useful to employ two mics; one very close and one more distant. The close mic picks up a clean sound with very little spill from other instruments; the more distant mic picks up a fuller, more rounded, tone. Mix them together according to personal taste.
Keyboard instruments are the easiest to deal with because they are usually DI'd into the mixing console. You can use an amplifier if you wish (which can help the synthesized sounds blend in with 'real' instruments better), it's up to you.
The one last important member of the band's line up is the vocalist. Although the vocals will be recorded for real later on (involving much time, blood, sweat, and tears), it is usually necessary to record a 'guide' vocal on tape right from the start. This not only helps the band know where they are in the song while they are laying down the basic track, it helps you judge what the finished result will be like and allows you to think ahead to what problems may occur later on.
In the studio, the vocalist can go anywhere he or she feels comfortable, as long as there is no spill into any of the instrument mics. If the guide vocals spill over onto other tracks at too high a level, then they may be audible in the final mix. You may getaway with it, according to how well the guide is performed, but it's not worth the risk. The singer may decide to change the vocal line later and any guide vocal spill will be ruthlessly exposed.
Once the band are happy with their instruments and amplifiers, and you are happy with the mics (as well as miking the instruments, it's a good idea to set up an extra microphone in the centre of the room, through which the band will talk to you when they are not playing), you can bring out the headphones and give each band member a pair. A good studio should have at least two circuits for headphone foldback, giving the possibility of alternative foldback mixes (usually for the drummer's benefit). Before you disappear into the control room, tell the band - above the noise they are probably making by now - that you'll be a few minutes setting up the headphone balance and suggest that they give the song a run through while they wait.
While the band are busy running through the song and having last minute arguments over minor details, you'll be checking each channel with the solo button and setting the mic input gain. Write the name of each instrument onto the mixing console's scribble strip with a wax pencil (Chinagraph) before going on to the next channel. (You will have done this for the drums already while you were working with the drummer alone). With this done, but no faders up as yet, you can turn up the foldback on all the instruments so that the band can hear what they are playing through the headphones. Listen to the foldback mix yourself on headphones and set up what you think is a good balance. When you think all is well, ask the band - over the headphones - and they'll tell you what they think of your efforts. A good foldback balance is one of the keys to obtaining a good performance. It's a very unnatural thing to play while wearing headphones, but it has to be done for the sake of the guide vocal, which is usually essential.
The musicians will have to negotiate a compromise between what they all would like to hear, since it's unlikely there will be enough separate foldback circuits for everyone to have an individual headphone mix. It's up to you to watch that the level in the cans doesn't get louder and louder, but as you add level to some instruments, you bring down others. If you don't do this, you'll end up with distortion in the headphones and deaf ears all round.
How many tracks do you have? Eight, 16 and 24 are all good numbers, but let's stick to 8/16-track operation for the moment. On 8-track, you need to limit the number of tracks for the basic recording to just six (plus the guide), perhaps like this:
|Track 1||Snare or bass drum|
|Track 2||Drums left|
|Track 3||Drums right|
|Track 5||Guide vocal|
Whether you isolate the snare or bass drum on track 1 depends on which you think is the most important to the song. All the other drums and cymbals will have to be mixed into stereo at this stage. With 16 tracks to play with, the options are more open:
|Track 2||Bass drum|
|Track 3||Snare drum|
|Track 4||Tom 1|
|Track 5||Tom 2|
|Track 6||Tom 3|
|Track 7||Overhead left|
|Track 8||Overhead right|
|Track 13||Guide vocal|
This only leaves four tracks for vocals and overdubs, so you might want to condense it a little (you'll know which way to go from your earlier discussions with the band):
|Track 2||Bass drum|
|Track 3||Snare drum|
|Track 4||Drums left|
|Track 5||Drums right|
|Track 9||Guide vocal|
This track layout leaves much more room for plenty of extra musical interest, and the result can sound technically just as good as the first layout - but you have to get the drums right now. There's no margin for error here. One nicety that you might notice is that track 1 hasn't been used yet in either of the 16-track track lists. This is because the edge tracks of any multitrack tape are prone to damage. It's unlikely with modern machinery, but it can happen (it happened to me a couple of weeks ago, in fact!), and if it does, it had better happen to a minor instrument where it can go unnoticed or be mixed down in level. The instruments being recorded at this stage are all important and need the protection of the inner tracks. You could, however, allocate the guide vocal to an edge track if you wanted, because it won't be used in the final mix.
As you allocate each instrument to a tape track, you need to write it down on a 'track sheet'. This is usually just a sheet of paper or card containing as many boxes as tape tracks. I prefer to use something like Figure 3. This gives me space to change things: for example, if one instrument is replaced by another, or if several are combined onto one track to make room for more work. Make no mistake about trusting this sort of information to memory. It's so easy to erase a track by accident, and musicians can be very unforgiving about this.
When you and the band are ready for a take and have set up levels and a monitor balance, ask the guitarist and bassist to tune to the keyboard (or if they have an electronic tuner, so much the better). In the old days, before synths were digitally controlled and stable, I used to record a tuning note before anything else (D2 is a good choice of note), then whenever there was any doubt about tuning, there was always an absolute reference available. A tuning note on tape also comes in handy if you are going to mess about with varispeed - more on this in a future installment. Once you are tuned up, it's time to begin. Get the band to count in audibly onto the tape for the benefit of future overdubs, and then you are into Take 1...
Achieving a good solid backing track is very important in this type of recording. Sequenced MIDI recording tends to progress gradually, track by track, but here you are covering several steps with one giant bound. If the backing track is not perfect, then there is no hope of success. Perfection here means no wrong notes, no drum sticks clashing together, consistent tempo, and a good 'feel'. Even if you record over bad takes, you may end up with three or four good ones from which, in consultation with the band, you can choose the best. In theory, it is possible to edit the multitrack tape so that you can combine the best bits of two or more takes. This relies on the tempo being consistent between the takes, so if you do feel adventurous, check the tempi of the bits you would like to edit together before you physically cut the tape.
Back in 8-track land, once you have a good take it's time for another giant leap. Since you have filled all but two tracks (you will probably have to lose the guide vocal very soon, but it will have served a good purpose while you had it) you will need to mix the six instrumental tracks onto tracks 5 and 6 (best not to use edge tracks, remember), then record your overdubs onto the tracks you have freed up. Mixing the backing tracks isn't easy, because you don't have all the instrumental parts available to make your judgement. But you can listen to the guide vocal right up to the point that you commit yourself to the record button. This will help achieve a good balance. And once you have recorded the backing track onto tracks 5 and 6 and cleaned up the others, there's no going back.
From the engineer's point of view, instrumental overdubs can be quite a chore. Especially with keyboards, you can spend a lot of time doing little more than nursing a fader, making sure that the levels on tape are correct. But on the other hand, if the band isn't experienced in recording, they will not know as well as you how precise things must be to ensure a good result. In live work, obviously you don't want to make a mistake, but if you do then you shrug it off in a moment and it's on to the next bit hoping that no-one noticed. In a recording, however, even the smallest mistake will shine through like a beacon, telling everyone that there has been a cock-up. Perhaps on the first playback it will sound very minor - maybe not all the band will notice it. On the second playback everyone will hear it; on the third and subsequent plays it will grow in magnitude until it's all you can think about. The answer is not to let any blemish get past a first listening. Everything you record should be auditioned immediately to make sure it is good enough. This means listening in complete silence and in a very concentrated manner. This is very important - to be able to listen to a take once only and spot even a small mistake is an art that is only acquired with practice. Casual listening, coupled with chit-chat between band members, is not nearly good enough.
Spotting mistakes is one thing. On a higher level of skill is the ability to judge whether a take is good enough. When a musician finds his part difficult to get just right, you are faced with having to decide when he has played it to a sufficient standard. Don't forget that it's nearly always the engineer who has the most experience in close listening and in assembling a song from a collection of separately recorded tracks. It's not good practice just to accept a musician's say so, but be as tactful as possible when making comments on the quality of peoples' playing. Make sure that you are satisfied with each overdub before moving on, then you'll produce a good end product, the band will be satisfied even on repeated listenings back home, and they will come back to you with more business.
Next month's installment looks at the twin challenges of recording drums and recording vocals. Later in the series, I'll be devoting a complete article to the important topic of mixing. Now there's an adventure to look forward to!
Feature by David Mellor
Previous article in this issue:
Next article in this issue:
mu:zines is the result of thousands of hours of effort, and will require many thousands more going forward to reach our goals of getting all this content online.
If you value this resource, you can support this project - it really helps!