Of all the stages in your recording debut, successful mixing has the most to do with atmosphere. Jon Lewin explains how to keep calm and influence people.

Mixing: ah, the peaceful womb-like atmosphere of the recording studio control-room... all those coloured lights playing gently on the silent scene... those little red LEDs winking rhythmically in the effects rack... the calm serenity of that aromatic cocktail — stale cigarette smoke and strong coffee.

Then the engineer comes in, bearing a reel of multitrack tape; he laces it on to the machine, checks the Dolby tones at the front of the tape, then sits comfortably back in his swivel chair and awaits the arrival of... THE BAND (that's you).

But why aren't you there already? Because you're reading this, which is going to tell you how to go about mixing your single (or demo tape if you're unlucky). This article is guaranteed to save you time, and prevent you appearing a bozo in front of the engineer (unless you are a bozo, in which case the guarantee is void).

Let's assume you are paying for this little excursion yourselves, so time is MONEY, in a very real sense.

This means that theoretical points, such as which number to mix first, should be discussed before you arrive. You have to know which song is your best — your intended A-side. If you've recorded two songs, mix the best first; if you've done three or four, put the most important song second. The idea behind this is that should you be pinched for time, it's not the A-side that suffers; it's also sensible to mix a song before it, as this gives you time to acclimatise to the sound and the environment.

Don't mix on the same day you record, even if it means not going to the pub for a month. Recording, with its accompanying high sound levels, is a physically and mentally debilitating activity, and your brain will simply not be able to cope with mixing as well as it could if it was fresh (well, relatively fresh, anyway).

Assuming that you are mixing in the same place you recorded your master tape, you should be familiar with the effects units there. If you're not (and even if you are), have you given the idea of A PRODUCER a thought? While the engineer will be able to help you with technical details, like how to use the AMS, it can be very useful to have another person to provide aesthetic input. It doesn't have to be T. Horn or Steve Levine, just somebody with studio experience whose judgement you trust. Musicians often find it difficult to distance themselves from their own playing. Having a producer who can see the wood as well as the trees (1 muso = 2 short planks?) can prevent the mix from turning into a battle for personal volume.

The need to have one person taking overall responsibility gets more obvious as the session progresses — somebody has to settle the argument between the guitarist and the singer about the snare sound in that middle eight, otherwise you'll be there all night. A producer should keep the mix on course, and on time. (And take the blame if it all goes wrong.)

Right, you've made it into the studio. Are you sitting comfortably? You'd better be, because you're going to be there for at least another eight hours. Then we'll begin. Go through the channels on the desk, working out which instrument is where as you play through your first song. The engineer ought to have a track listing from your recording, but it's as well to check.

Once you've located all your sounds, start working on the equalisation for each track. Check your settings on all the different types of monitor speakers provided, swapping from the semi-detached size JBLs down to the grotty little Auratones and back — this will give you a much more accurate idea of the quality of your sounds. A big bassy whoomf on large speakers can easily turn into a weedy clack on a transistor radio.

If the players have a strong idea of what sound they want, try letting them set up their own EQ, providing they know how the desk works, and it doesn't annoy the engineer. But don't let too many people try all at once — you're not that pushed for time.

This is the stage at which you start thinking of outboard add-ons: effects units. We all know about putting reverb on snares; but how about putting the same setting on the whole drum kit, for that extra special lift-shaft sound? Or for a modern disco snare, try putting a noise gate on the reverbed signal. But remember, every time you find an effective (sorry) setting, you are tying up that unit — unless you record the result on to a handy spare channel.

Bouncing down to another track is also a useful trick with particularly busy parts of the song that require complex mixing/cross-fading/dropping in or out. It's easier to make two or three attempts at bouncing that section down on to one track than it is to try and get it right in the final mix. Quicker, too.

Once you have a rough set-up of each sound, it's time to start combining them into sub-groups (if the mixer has that capacity). This enables you to set a level for each individual drum (for instance), then route all those channels to one fader. This gives much simpler overall control of the various basic elements of the track.

It is usual to treat the song as a series of layers, starting with the rhythm section, and working up to the vocals. As you go through the instrumental parts, it's important to maintain an idea of their function in the context of the whole song. Do you really need that fifth harmonised octave guitar part? Does double-tracking the bass actually add anything? Bear in mind that radio producers/A&R persons/the general public are going to be listening to the song and the melodies contained therein, which theoretically renders everything else redundant. Thankfully this is not quite true, but it does help to emphasise the point that all the parts you have recorded are subordinate to the whole — which means that if the whole sounds better without a bass part, then the bassist doesn't resign in a huff.

As you build up the song from its component parts, you may find that certain sounds conflict, such as a bass guitar and a low synth line. Instruments that have the same frequency range can clash disharmoniously, and it's not always immediately obvious — often you might just perceive something indefinably 'wrong'. If you think of your mix as a way of juggling sounds in the aural-spectrum, it's evident that you have to avoid 'over-colouring' certain frequencies. Try using EQ to solve the problem.

To give a specific example of this technique: when using a synth to underpin a bass line I found two notes slightly out of tune with each other. Using the parametric EQ on the desk, we located the errant frequencies and wound one down to a point where the dischord stopped its annoying 'beating'.

The natural tendency when working on individual channels is to make them individually as rich and full-sounding as possible. But when the song is constructed, you may find that too many fat sounds make the final mix appear bland (this is the American approach). Occasionally it is necessary to use thin and weedy noises to add contrast to the fantastic opulence of the rest of the recording.

Let's take a break here. Regular ingestion of food and stimulants such as coffee is highly recommended during both recording and mixing. The use of consciousness-affecting drugs is not advised, as very few things can make you change your mind quite so quickly afterwards.

Very rich groups make copies of their multitrack recordings so they can use the copy to practise their runs-through at getting the mix right (saving wear on the original), then actually mix from the master. Smart, huh? I bet you can't afford to do that. But then you probably can't even afford enough tape for you to have recorded your master at 30in/s. (It does improve the quality.)

Right, back to work: as you progress, soloing individual instruments (checking them on their own), and dropping parts back in the mix, you may notice certain peculiarities coming to the fore among your fellow musicians. Perhaps the drummer keeps asking for more top on the bass guitar; persistent requests like this can be dealt with through application of the Producer Principle. This involves the person then at the mixer leaning over the relevant control, pretending to tweak it, then asking, "how's that?" This will invariably produce a positive response from the poor deluded drummer. This technique owes more to the power of suggestion than psycho-acoustics.

Often you will find that instruments seem to disappear in the mix. This will lead to calls for re-EQing the disparu in order to 'bring it out' by adjusting its place in the aural spectrum. Although this is a valid method of organising things, you'll often find that the result of changing the EQ is just a change in volume — don't use the EQ as a substitute for slight increases in loud.

The more channels/tracks/parts you use, the more likely it is that the mix will begin sounding muddy. Again, it is often possible to attribute this to causes other than the general grubbiness of your song. Adding echo to vocals or instruments can alter your perception of their timing. If something seems to drag, try it without the echo. Better? If it isn't, check that one of the effects units hasn't started overloading. This can happen quite easily as levels are altered and is not always obvious on quieter instruments (effects sends are normally separate from channel levels).

If you are using drum-machine and triggered synth, you might notice a slight and undesirable slackness in the triggering of the synthesiser, especially if it is being run from a click track on the master. You can avoid this by a simple trick: physically turn the multitrack tape over on the machine, so it runs backwards (making sure you don't put your sticky fingers on any bits with music on them). This means your track-by-track guide would have been inverted, too. Find a spare track, and bounce the click with a 20 millisecond delay on to it (recording just the delayed signal). This advances the click to just in front of where it should be which means the synth or drum box (when you re-record it) will be exactly in time. Sneaky.

Even if you have succeeded in demuddying everything, tricking synths into time etc, the overall mix may yet lack that edge you wanted. Try experimenting (one eye on the clock), with particular variations — drop out various instruments, or groups of instruments. Ever heard an A Cappella 12-bar? Try an inspired guess. The basic theory is Go Over The Top, Then Come Back A Bit — caution is not always a good policy.

If you think you've finally come up with a finished product, or even if you're not sure — try putting it down onto ¼in tape. Four or five different mixes of one song is not exceptional, so give your imagination full rein. When you are doing a final mix-down, don't be surprised if it's necessary to have several people clustered around the desk, as it's often easier to have four musicians with two tasks each than one engineer trying to remember eight separate Fade In/Punch Out points.

Nerves can be a problem at this, the climax of the session, but it doesn't help anyone if you berate your colleagues' lack of motor functions for ten minutes because one of them has moved a fader a centimetre too far — just relax and try again.

Remember to check your final mixes on the grotty speakers before you commit yourselves to throwing anything away; try wandering around the control room, as acoustics change even in the deadest environments. Does it still sound tolerable after several hours' repeated listening? Have another break, because there's another two songs to do yet...