It doesn't matter how well you've recorded your music; you'll blow it if the mix is a mess. Patience is the key.

AND SO WE COME to mixing. The main presumption of this article will be that you have some music that is actually worth mixing, so let's try to think of a way of not messing it up, shall we?

The first point about mixing is this: if there's somebody in your setup who can't make it along to the studio for the mixing, somebody who has some prior interest or a love life hassle or a dental appointment, or a religious objection to mixing on a Sunday, then let them go. Say you can get along fine without them; it'll be difficult of course, you will point out, but somehow you will manage. The reason I say this is that mixing causes fights. Big fights. Whereas recording music is ideally a group activity, mixing is for one. It is for one silent, well-balanced person who is good at taking notes and can get through a lot of black coffee.

Mixing is nothing more than a constant series of decisions that need to be made. If you have, say, two pieces of music down on 16-track, then it's probably going to take not less than five hours of studio time to mix them. So in a group, arguments are inevitable, and unless you are abnormal, you will be tempted to hit one another.

It's not just that everybody wants their part turned up. A more fundamental problem is that mixing comes down to a matter of personal taste. There are no rules about mixing. Even the process by which it is done is not fixed. Some people think it best to get one sound right at a time, whereas others want to keep all the sounds in play all the time, and judge how one sound should be in the context of all the others. I prefer the latter method, as do most people I know who are not utterly stupid.

What I suppose I'm saying is that you should appoint somebody to be the final arbiter, the umpire, whose decision will be accepted by all. It may be the engineer - if the guy can give a toss about your music. He's probably the best person and he will be flattered if you constantly defer to him. He might even stay awake.

Right. Now if you're new to this lark, you'll probably think that a mixing desk is just one of those impossibly complex things, like Maths 'A' Level, a thing understood only by weirdos. This view is, I'm afraid, in many ways correct. The technology is pretty heavy, but look at it like this.

The mixing desk receives a signal from your master multitrack tape and transmits it to the tape on which you will mix-down. It adds to this signal another signal which is received from the various effects which you will wish to add.

The mixing desk itself contains no effects, strictly speaking, but it does contain some rotary controls which are called EQ, and which govern the tone of the sound by boosting and cutting different frequencies within it. These are like the tone controls on your record player, and EQ can be regarded as a sort of quasi-effect.

Otherwise, what you need to bother about on the mixing desk are the pans, which position the various sounds in the stereo field (more on this later), the manual faders (the sliding controls with which a sound can be brought in and out), and the mutes, which are buttons which instantly bring a sound in or take it out. These controls are set in a column, and there should be one column for every track.

I've already said that there are no rules about mixing, but before we get to the nitty-gritty, let's look at what should be your aim.

Basically, you should try to get everything heard. After all, that's the reason you recorded the sound. You should also get things working together in a way which is dramatically correct for your music. Here, I would like to insert a personal note once again.

An A&R man once said that a song of mine lacked structure. At first, to my serious consternation, I thought he was saying that I couldn't write a song - a fairly fundamental and quite possibly justified criticism. It turned out, however, as I probed deeper into his pathetic, confused thoughts, that he thought my songs lacked "build". This was the key word. What he was talking about was not the tune, so much as the arrangement and the mix. The tune was there, but not the depth of sound, the little aural gimmicks, which made it progress.

You should be aware that any "professional" listening to your tape will be listening to the overall effect, and not the one element that you think is your strong point, like your songwriting or your slap-bass playing. Achieving the right overall effect is a matter of mixing.

If there is one individual sound that an A&R man is going to listen out for it is, as I mentioned last month, the vocals.

You may have added some compression to your vocals when they were recorded. This effect, which helps to maintain the consistency of a sound, is exceptional in that it often should be added at the time of recording, though you can always add it later.

You may also have double-tracked your vocals, in which case the main task in the mix will be to make these tracks blend. This is just a technical exercise, and it's quite obvious when it isn't working because the tracks will not sound like one person singing, but two.

If you haven't double-tracked, or even if you have, then you are likely to want to add some delay. This is probably the main vocal effect used, and gives your voice more presence and drama by the addition of a slight echo. It's applied for a similar reason as reverb, but is sort of, well, classier.

Reverb (also a delay effect) can be used on vocals, but too much reverb is one of the classic hallmarks of the bad demo tape, as opposed to the good hit record. The voice just sounds as if it is swathed in technology, with no human being in sight.

The addition of too much chorus - which thickens a sound - also sounds amateurish. On the other hand, an effect known rather thrillingly as "aural excitement" (after the machine that started it, the Aphex Aural Exciter) can be very useful. This enhances the treble in your voice and raises it out of the murk, without adding too much in the way of spurious noise.

Incidentally, always mix vocals up. Low-mixed vocals are for megastars and can cause even the best of them to sound bad, as the recent work of Bryan Ferry demonstrates.

As for drums, one of the latest gimmicks dragged into vogue by such successful businessmen as Phil Collins and Robert Palmer, is gating. This is the addition of an effect which cuts off the sound very sharply, giving a penetrating sound that's crucial for drums. It is often used in conjunction with reverb, hence the term "gated reverb".

Gating is often used on the snare, but you should also consider gating the sound of the bass drum if you've used an acoustic kit. Gating helps the critical but easily muffled bass drum to cut through the overspill from the other drums which the bass drum's mike will inevitably have picked up.

If you've used a drum machine, you will still need to make the sound cut through, and so gating may appeal. You may also, like me, hate the sound a drum machine makes whatever effects are added. If this is your problem, you might consider asking the producer to use your drum track to trigger the sampled sounds of real drums. This will also have the beneficial effect of teaching you about sampling.

With drums, panning is particularly important. This is the technique of positioning a sound, of making it appear - as it comes through the speakers - to be to your left, your right, or directly in front of you. A typical panning arrangement for a drum kit is to have the bass and snare fairly central and the sound from the toms pushed to either side. With luck, your producer will have a vague idea how to do this.

Bass guitar is particularly hard to make audible: the distant throb of a half-heard bass is another of the trademarks of a bad demo. You should, as with the vocals, consider compression (for consistency of sound) on the bass. Another useful bass effect is chorus. A bass sound can easily become fuzzy. To reduce fuzz (and this goes for other instruments too), you might use EQ to reduce certain frequencies within the sound.

As long as they are playing chords, keyboards and guitars are meant to be "background" and you can concentrate on making them sound what you consider to be "nice". Riffs have to be relatively louder, just as a lead vocal has to be louder than a backing vocal. (This may sound elementary, but once made, backing vocals should often be lower in the mix than you think. Nothing sounds worse than some joker who's been given a backing vocal specifically because he can't sing, warbling at the forefront of your tape.)

Once you've decided on a mix, if it involves much manual fading, or general fiddling about with the mutes or other controls, you should rehearse it a good number of times. Then commit it to tape in three or four slightly different versions. Once you've done this, play it back (obviously) and then make copies on cassettes. You may want to take copies of the two-track master away in order to make better quality duplicates than you can with a cassette. Always ask the engineer to retain the master tape so that you can go back and remix if you don't like any of the versions you've got.

Well, that's essentially it. I'm not going to tell you not to add too many effects to your mix. I'd like to, honestly I would. And if your taste in music veers towards such things as rap, house and rare groove, you're actually better off not putting too many effects on things like bass and drums. But the style of records currently in the charts - records which, presumably, A&R men favour - is one in which effects play a major role. We live in the age of Kylie Minogue, whose every other meaningless word has so much delay on it that you can hear it four or five times, though je ne sais pas pourquoi.

On the other hand, if you really want to be great, get out there, ignore all the technical advice, and just do what comes naturally. Because rules are made to be broken.