Recording Techniques (Part 18)
Part 18. David Mellor concludes this 18-part mammoth series on recording techniques with a handy question and answer session.
Throughout the Recording Techniques series I have had two aims in mind: firstly, to help home recordists get more fun out of their music and their equipment; secondly, to give budding professional engineers a 'leg up' and hopefully point out a few interesting directions worthy of exploration. Now more than ever it is possible to achieve good results from very basic setups, and it is also at last possible for anyone who wants to turn pro to have a real chance of getting a shot at the big time. In my various roles as lecturer in sound engineering, composer and producer, freelance classical recording engineer (and also general odd-job person it seems), I am often asked questions on my favourite subject which range from the highly intelligent and perceptive (ie. the ones to which I don't know the answer), to the more practical ("could you use fewer microphones and give me a discount?"). So I have put a few of these questions together with one or two I that I myself have had cause to ask in the past. I hope the answers will shed just a little more light on the vast subject of Recording Techniques.
Can I achieve professional quality recordings on my home equipment?
In a word, yes! That's called positive thinking, but even being more realistic, you don't need all that much gear to produce a recording with commercial value (bearing in mind of course that every product has to be correctly marketed before it will realise that value). I would say that the bare minimum is a midprice synth, a reverb unit (if the synth doesn't have effects built in), and a second-hand Revox A77 stereo tape recorder (the high speed version, and well maintained of course). This little lot could cost well under £1000 if you shopped carefully and scanned the small ads.
You would be limited to very simple types of music unless you added more equipment, but even simple pieces can sell very well indeed and you will hear plenty in the soundtracks of TV programmes and on the radio. I have had dozens of tracks published that I created on a budget 8-track in the days when budget 8-tracks weren't anything to rave about. While I was waiting for my 16-track to be delivered I used a simple Fostex cassette multi-tracker in conjunction with an Akai sampler and a sequencer to produce twenty or so tracks that have since appeared on CD. The basic rule is to work within the limitations of your equipment, exploit its strengths and conceal its weaknesses.
Can I master on to cassette?
For your own use and demos the final stereo master can be a cassette. But the quality of even the best cassettes just doesn't compare to a decent reel-to-reel or DAT. Get your practice in with a cassette deck, and then when you think you are up to it, invest. You certainly won't regret it, and you'll be able to produce a master recording that is acceptable by publishers, record companies and record and CD manufacturers.
How long does it take to record the average song?
The clever answer is that if you are only recording an average song, then it probably isn't worth the effort. Basically, it takes as long as it takes. I heard a radio interview with Tom Petty the other day, in which he described how he and his band recorded two hundred takes of one song to get it to its finished state. It wasn't much to write home about either (although I like some of his other material), but it does demonstrate that you really have to put everything you've got into your recordings. Other people do, and you'll have to do the same to compete. It would not be unusual to spend four or five days on one song if you're starting the production process from scratch.
If I can only afford one microphone for my recordings, which one should I buy?
You have to have good equipment, good speakers and good acoustics before microphones get really exciting. After all, they don't have any knobs on them, do they? But professional engineers will talk about the characteristics of microphones all day if you let them, so they must be rather important. Every microphone has its own characteristics, and suits some situations more than others. It follows from this that there is no one mic that is suitable for everything. However, there are certain models that can almost do no wrong, with the emphasis on the 'almost', one of these being the Neumann U87 which must be just about the classic mic of all time. It's expensive of course, over a grand now, new, but second-hand U87s sometimes become available for less. They can be used on anything from tin whistles to floor toms, although they are usually considered to be at their best for vocals and orchestral section miking. The one thing they are not so good for is co-incident stereo miking, not in my opinion anyway. At the cheaper end of the spectrum the Shure SM58 is a good vocal mic, and can probably be regarded as the lowest cost fully pro mic. The only snag is that it doesn't have a wonderful top end response, so it's virtually useless for metallic percussion — tambourines etc. As a relatively inexpensive all-rounder, I can recommend the Beyer Dynamic M201, which will handle just about everything with a high degree of competence.
I have some MIDI equipment but no recording gear. Should I expand my MIDI system or buy a multitrack?
You have to buy a multitrack! The best part of using a MIDI system is when you have worked up the composition to a reasonable state and then start laying it down onto the multitrack tape. As it builds up track by track (doing it this way maintains optimum synchronisation between instruments) the whole thing really comes alive. I don't know why this is so, but it's like the difference between sitting in a hot stuffy room, and then opening the windows and letting in a breath of fresh air. And once the transfer is done, your MIDI system is freed up to add a whole new kaleidoscope of creativity.
Is there any point in going to 8-track when 16-track seems so popular?
It depends on how much money you have and what you want to achieve. 16-track is very popular at the moment because it is the entry level into professional studio operation. 8-track used to be, but musical arrangements have become very much more complex in recent years, and you really need those extra tracks if you want to be able to handle most of what's likely to be thrown at you. One day, 24-track will be thought of as entry level, and I don't think that's so far away either. It really boils down to what you can afford. If you can stretch to 16-track then do it. If you are interested in the music more than the equipment, you may get years of use out of it before before you start thinking, "it would be nice to move up to twenty-four". But if you buy 8-track, then unless you have a clearly limited purpose for it in mind, such as an 'expander' for a MIDI system, then you'll be itching to move up in no time at all, probably less than a year, and you will inevitably lose money in the trade.
Should I add EQ and effects while I'm multitracking, or leave them until the mix?
Leaving effects and EQ until the mix is playing safe. There are indeed times when it's best to play safe, and other times when there may be more reward from living dangerously. If you have invited a vocalist or instrumentalist into your studio and negotiated a special rate (tea and biscuits?), then it makes a lot of sense to play safe. Get the cleanest, most accurate recording you can and take pains to make sure that there are no faults such as misreadings in the lyrics, wrong notes etc. Then when the musician has gone home and you are left alone in your laboratory you can carry out all the fiendish experiments your imagination can devise, safe in the knowledge that you can start back from square one at any time you like. But, at the other extreme, if you have a controlled source such as a synth or sampler playing under the direction of a sequencer, then it would be a criminal waste of an opportunity not to have a go at doing something to the sound as you record it to multitrack, otherwise you are playing parrot to the factory programmer's "Who's a pretty boy then?" What you do to the track is up to you, whether it's EQ or a little bit of 'room' sound, but of course you must bear in mind that the track will have to harmonise with the rest of the mix. The great thing is that if at any later stage you don't like what you have done, the raw track can be repeated at the push of a couple of buttons. Adding a little extra creativity to the basic sounds you use is adding value to the mix and making your recording more personal.
Why can't I get a great drum sound like I hear on record and CD?
Let's redefine the question to refer to "sampled drum sound" since recording real drums well is a specialist art form in its own right, and you wouldn't really expect to compete with first rate studios and engineers in your bedroom. But it should in theory be a piece of cake to match the quality of sampled drum tracks since the quality of the drums (assuming you are using good source material) is there already. But the old saying, 'it's not what you have it's what you do with it', finds its true meaning in this situation. It's a rare studio recording where the drums have not undergone processing of some kind, even if the sounds were basically good to start with. Big mixers have sophisticated EQ, and a good monitoring environment helps enormously in getting a good drum sound because you can fine tune each component with almost limitless precision. You can't do that at home because you can't hear the sound as clearly and accurately, and you probably don't have quite as full a selection of tools.
This may sound slightly negative, but one has to do the best one can and work within the limitation that if you want the drum sound to form the basis of your production technique then you're going to have to work hard. But there is an alternative. It is becoming a rather more accepted practice than it used to be to sample drum beats from commercial recordings and loop them to back your own music. This way you get a 'studio produced' drum track, and you add the instruments on top. Of course this risks compromising your creativity and you have to ask yourself whether the music is still yours. Also, if you plan on having the recording released, then the owner of the recording you sampled has a right to prevent you doing so (in fact he or she had a right in theory to stop you sampling the drum beat in the first place). In practice, these difficulties can be overcome (and you only have to listen to the dance charts to find out how much sampling and re-sampling is going on — I'm told that one person was found sampling his own drum beat the other week). The moral is, if you can't solve a problem by tackling it head on, try and sidestep it. There's always another way.
My mix sounds great in the studio, but not so good elsewhere. What can I do?
This is a classic situation. Something similar occurs when a mix sounds good played loud, but loses an inordinate amount of impact when played at a normal level. It's very easy to get carried away by the overall sound of a mix and not listen to it sufficiently analytically. The relevant strengths of the important structural instruments have to be judged very carefully, particularly the balance of the bass drum and bass guitar between themselves and with the rest of the mix. Also, the ear (or more accurately, the brain) has a system for ignoring prominent sounds which go on too long, so that synth line which sounds just right when you let the mix wash over you may in fact be too loud, and it may jump out of the speakers and poke you in the ear when you listen the next day.
As I have said before, it's vital to try out your mix on different systems before you draw a line under it and call it finished (a work of art is never finished, so they say, only abandoned). Try it out at low level too. It's often instructive to turn the mix right down until most of the instruments are inaudible. You may find that one is more prominent than you expected and it may need to be dealt with by pulling down the fader or introducing some EQ cut in its predominant frequency band. If you have really big speakers you may find yourself tempted to put a very deep bass or bass drum on the recording. Watch out then that it doesn't disappear completely when played on small domestic speakers because it is below their frequency range. If your monitors have a particularly tight bass end, then it's also tempting to load on the LF, and then you find that the cones of ordinary speakers flap around uncontrollably because you're asking them to handle more bass than they really can.
The vocalist I'm using sounds great on stage but terrible in the studio. What can I do?
Another common complaint. It's true that poor intonation can go unnoticed in the excitement of a live gig, but it shows up only too strongly in the studio. It is very difficult to judge pitch when you're listening on headphones and it's an exceptional vocalist who has such a good innate sense of pitch that he or she can produce the right note without being able to hear what's going on properly. A partial solution is to slip one headphone slightly behind the ear so that some sound can enter naturally. A better but more complex solution is to let the vocalist monitor on loudspeakers. Of course, a certain amount of crosstalk will be recorded along with the vocal, but much of it can be eliminated by recording, on another track, just the crosstalk through the mic, but with the phase button pressed. When you add this to the vocal track a lot of the crosstalk will be cancelled. If you do this, bear in mind that the mic has to stay in exactly the same position. It's not something you can come back and do a couple of days later.
If a vocalist has a less than impressive vocal timbre, he or she can be thickened up by the traditional technique of double tracking. You don't need an expensive effects unit to do this, just a spare track on the tape and a lot of patience. The vocalist simply repeats his or her part exactly, which nearly always results in a very professional sound. In fact, this technique works so well that effects units can't even begin to approach the quality that can be achieved. Problems crop up when there are slight discrepancies in the timing of consonants and the length of words. The way to deal with the consonant problem is simply to miss out the troublesome ones on the second track. For example if a line ends with the word 'feet' (no-one's going to sue me for copyright on that example), then on the second pass the vocalist should just sing 'fee' and no-one will know the difference. Getting the length of all the words right is a matter of trial, error, practice and patience.
My recordings are basically good, but how can I give them some extra professional 'polish'?
As you may have gathered, I'm a great believer in the power of EQ and hope that manufacturers will one day give us the power of expensive units at a more reasonable cost. After your mix is finished and is as good as you can get it with your existing equipment, it may be worthwhile hiring some extra gear to take it just one stage further.
It's common practice to EQ the final stereo mix before it's transferred to record or CD, sometimes to iron out problems, sometimes to add a little more 'magic'. If you can get hold of a good equaliser and a second stereo machine then you can do this before the mix leaves your studio. One way to do it is to get out a CD which you think sounds great and play it over your monitors for 15 minutes or so. Then change over to your track and start equalising, not to copy the sound of the CD but just so that you like what you hear, your ears already attuned to a sound you like.
Work at this for a maximum of half an hour, so your ears don't get tired, then have a long break. Come back from the break suitably refreshed and play the track without EQ, then play it with the EQ. If it's better then make a recording, take note of the make and model of the EQ unit and the settings you used, and be sure to label the original master and your new version carefully.
Compression can also be used on a mix, although it's not as universally applicable. If you can hire a unit which has the so-called 'soft-knee' feature then so much the better. Experiment with EQ and compression and see what you can achieve.
My recordings are, frankly, boring. 1've heard much more interesting recordings that use far fewer instruments. Where am I going wrong?
It sounds like a case of too much precision and exactitude. There comes a point when you go beyond getting things right and start to sound like a machine. And come to think of it, with so many machines in our studios anyway, how can we stop music becoming mechanical? It boils down to artistic judgement. This is something that needs to be liberally applied at every stage of the recording process. It's very easy to record a pattern into a sequencer and then press the quantise button automatically, without thinking. But it's much better to listen first and ask yourself whether the track or pattern needs to be quantised. If it does, then quantise it, but keep the original version just in case.
Sometimes there may be just the odd note seriously out of place, in which case simply edit that one note. It can be difficult to judge when something is 'perfect enough' but not over perfect, and when you are judging against sequencer-style precision it's difficult not to think that your playing is inaccurate. But small inaccuracies give music life and vigour, and that needs to be preserved as much as possible.
Another cause of boredom in a mix is the repetition of sounds. If you use a sampler, then you may have the same snare drum beat occurring well over a hundred times in a song. Find ways of varying repetitive tracks. Instead of just using the same snare all the way through, how about making a group of three notes with the two outer ones tuned slightly, and almost inaudibly, up and down. If you use these at random points, it will go some way toward the situation where a real drummer, no matter how skilled, can never make two snare drum hits exactly the same. Finding ways of adding variations in timing and timbre will add life to your mix.
How can I make my recordings different to everyone else's?
Simple: go easy on the factory programs in your synth. They may sound great, but everyone else is using them too. I spent an enjoyable half an hour erasing all the factory programs from my multi-effects unit the other day (actually it wasn't enjoyable because I had to unlock all the presets before I could overwrite them). I knew that I would be losing some good 'uns, but I hope to make more of my own, and what's more I have made a conscious decision not to reuse effects but to create them from scratch each time. It's going to take a little longer, but I want to make music, not run a production line. The moral is to take your inspiration from other people, but use every opportunity you have to make your sound your own, by creating sounds from raw materials or by combining existing sounds in new ways.
I want to be pro engineer. What do I do?
I have to ask a personal question here: how old are you? If you are under 23 then you can choose Route A. If your teeth are rather longer and your hair a bit greyer then you'll have to go via Route B. There are now a lot of recording studios about; over two hundred commercial studios in London alone, and many more dotted around the country. These studios all need staffing and from time to time a job vacancy will crop up. The best chance you'll get of snapping up such a vacancy is by knowing someone who's in the know — recording engineering isn't the type of job where you can get in on a couple of exam passes.
Route A takes you into a commercial studio. You have to be young, 16 or 17 being the optimum age. Start by writing to all the studios you can find (in the Yellow Pages and elsewhere) and showing an interest by asking if you can pay them a visit because you are keen to get into recording. Note that you haven't actually asked for a job yet; studios get so many letters from people wanting employment that they often get filed under B for bin. But if you can wangle your way in and impress the manager, there's always the chance that he'll remember you when a vacancy does occur.
Another way of getting to meet the right people is to take a studio engineering course. If you are between 16 and 18, ask your careers office about Youth Training courses in recording engineering, and don't accept "there's no such thing" for an answer because it's not true. If you get on a course like this you could be placed in a studio for four days a week and spend the other day at college. How much of a success you make of this is entirely up to you, and you have to impress the studio that you're worth employing when your training period ends.
Route B is for those who have passed their sell by date. The most realistic thing to do is set up your own studio, preferably in partnership with someone who understands the business well. An alternative is to look for work in a related field such as PA or Audio Visual where they are not quite so age conscious about prospective new employees, but it still helps to be under 30.
However you get into recording, if you are going to take it more seriously than a hobby, be prepared for a lot of hard work and long hours. Come to think of it, you can spend a lot of time tinkering around in the spare studio room just for the fun of it. Recording engineering is particularly rewarding because for every little bit of extra effort you put in there is that little bit of extra return in the form of a better recording. Perhaps your ambitions go no further than harmless enjoyment out of the public gaze, or maybe you want to see yourself behind a mighty SSL console in a few years time. Wherever your recording takes you, I hope this series will have helped just a little, and I wish you good luck with whatever you do.
Feature by David Mellor
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!