You are here

The Mix

How To Become A Record Producer, Part 7 By David Mellor
Published July 1996

As the final mix looms, the end of the producer's job draws near. But which approach to mixing is best? David Mellor fires up the Moulinex...

Much has already been said in these pages about mixing, but from an engineer's point of view, rather than a producer's. Many years ago, before the modern era of multitrack tape, recording and mixing were part of the same process. You had to get a good balance between the voices and instruments at the session, because there was no way of making adjustments later. In fact, much orchestral music is recorded in the same way today, directly to stereo. Mixing directly to stereo is also very common for today's live and broadcast popular music. It's possible to do without a multitrack stage in all of these situations because the engineer knows before even one fader has been lifted what the final recording should sound like — as close to the original as possible. Since skilled engineers can achieve such remarkable results, you might wonder why we need a multitrack stage at all. The answer is that it allows much more attention to be paid to each element of the recording, in terms of sound quality, musicality, and creativity. Giving these three aspects the attention they deserve is why recording takes so long and demands so much effort.

Get It Right First Time

Ideally, the producer should start the mixing process as soon as the project gets off the ground. He should have an idea of the sound being aimed for, allowing a certain amount of scope for creativity, depending on the nature of the project. If it's a dance record, for instance, the producer must understand the style well enough to know the elements of the music his audience demand, and should add to those elements new and different sounds and textures to push the style further into the future. If the producer knows how every step of the preparation and recording process is going to contribute to the final mix, the mixing stage should be straightforward and successful. This means, among other things, getting the arrangement right and selecting the right sounds, making sure the musicians are playing in time and in tune, and obtaining a good performance from the singer by whatever means necessary. If there is a problem in any of these areas, you can only turn a deaf ear to it for so long — until the mix, in fact. Any problems present on the tape at the mixing stage will have to be disguised or covered up. Those problems should have been corrected as soon as they occurred.

Monitor Mixing

As more tracks are being added to the recording during the overdubbing stage, the engineer and producer will be working with a mix that may bear a passing resemblance to the finished product — the monitor mix. The monitor mix is what you listen to during the recording process, and is usually thought of as a rough guide to what is already on the tape — good enough so that the musicians can get a proper feel for the music, and good enough to tell the producer how the recording is shaping up. The exact monitor mix depends on the type of mixing console you are using: some have only very basic monitor facilities — perhaps just level, pan and a couple of auxiliary sends. This means that you can't do anything in the monitor mix apart from set how loud each instrument is, where it appears in the stereo image, and how much reverb it has (the other aux will be used to send foldback to the musicians' headphones). This is not a bad way of working, however, because you will hear exactly what is on the tape as you progress through the recording. If the monitor mix sounds good, you can be sure the final mix will sound great.

On average, people probably listen to each record they buy about six times; the producer of the record has to listen to it more like six hundred times during the recording and mixing process.

This simple style of monitor mixing has its merits, but large‑scale consoles offer vastly more sophisticated monitoring facilities. You can create a mix on the monitors using EQ, compression, gating, and everything else that is part of modern studio technology. If you regard the monitor mix as something temporary, but you — and the engineer — then proceed to use all of these facilities, you may find yourself in big trouble by the time you flip the multitrack onto the big faders and start to mix from flat, because the sound will be totally different. By the time you graduate to SSL‑ or Neve‑class studios, you should have learned the first rule of recording: nothing less than 100% effort is good enough. You should regard everything you do as being part of the finished product and make it as perfect as possible. This includes the monitor mix.

With a console that only has rudimentary monitoring facilities, you'll probably want to perfect everything on tape. With a console that has sophisticated monitoring facilities, you'll record a good clean sound on tape, and then anything you do to the monitor mix will become part of the final mix. The console will allow you to do this so you don't have to start from scratch when the overdubs are finished. In fact, you can do this with any console that has enough channels. It's very common, once a few tracks have been recorded, to route the multitrack to the channel faders and start to mix as overdubs progress. Working in this way, you never have to say, "Right, let's clear the desk and start to mix". Instead, you come to a realisation that everything is done and all that is needed is a little polishing. Self‑produced artist George Michael works in this way and his status at the top of the singles chart as I write confirms the value of this philosophy.

From Another Point Of View

There are many ways to make a record, and I can imagine some producers reading the above and disagreeing with it. Another style of album recording involves putting down all the basic tracks, then overdubbing other instruments and vocals, then taking a break of a few days before starting to mix the whole lot. The disadvantage of working a song at a time, all the way from basic tracks to mix, is that you can easily lose perspective. On average, people probably listen to each record they buy about six times; the producer of the record has to listen to it more like six hundred times during the recording and mixing process, and over‑familiarity with a song and the recording of that song means that you can't judge it in the same way as a punter would. Taking a break between recording and mixing means that you can come back to the song with fresh ears and hear very clearly which bits need to be brought out, and which elements play an important but subservient role. If this is the philosophy of mixing that appeals to you more strongly, it would probably be a waste of time working on an elaborate monitor mix. You could store it on a console with recall facilities, but that would negate the advantages of taking a break before mixing. A simple monitor mix is probably the best idea.

Still on the subject of monitor mixes, another common practice is to swap between songs during overdubbing, according to which one you and the band have most enthusiasm for at the time. Or you may have booked a session player who you want on more than one song, so he might as well do them all in the same session. This means locating to each song on the multitrack and resetting the monitor mix on the console. If you confine yourself to level/pan/reverb monitor mixes, it won't take too long to set up. Sometimes, however, during the later stages of overdubbing, you may feel that the mix you're hearing sounds really great, just by chance, and you would like to keep it as a reference for when you start mixing proper. In this case it's a simple matter to copy the monitor mix onto a DAT so you can check it later. With so few variable elements, it's pretty easy for a skilled engineer to reconstruct the mix almost exactly, and then you can go ahead and improve it still further.

How To Get A Good Mix

Simple. Use a good engineer and go away! I mentioned earlier in the series that engineers acquire a vast amount of experience of working with music and sound, and they are the people who should be operating the faders — not the producer, unless the producer comes from an engineering background, of course. If the producer sits in the studio from the moment the first fader is raised all the way through to the finish, he will be nothing but an inhibition for the engineer, who would really like to get on and tinker with the sounds and try out lots of ideas, many of which might not work. So this is a good time for you to take a walk in the fresh air and clear your mind ready to make an objective judgment on how the mix is progressing, two or three hours after you left the engineer alone with it. You may leave behind a few ideas or guidelines, or you may even encourage the engineer to go wild and try out some crazy things. When you return, you will hear your production in all its glory and you will be able to advise on what you want more of and what you want less of. You could even say that it is entirely wrong and you want to start again. An experienced engineer accepts that the producer is in charge and won't take offence (he just won't work with you again!).

A trickier question is what makes a good mix. It's especially tricky for the engineer who has to learn every detail of how to get a good mix. A producer, on the other hand, doesn't need to know the details but has to be able to recognise when something is right, and offer meaningful comments when it isn't. You need to keep in mind the purpose of the mix. Is it a dancefloor mix that should sound great on a club PA? Or is it intended for CD listening at home? A radio mix should emphasise the 'buy me' factor, whatever it is that will attract the listener to the record store.

The engineer will always sit in the optimum listening position, directly between the speakers, while mixing, but you will probably wander around the room. This is so you have the opportunity to hear the mix in less than perfect conditions, which is exactly the way the record buyer will hear it — in a club with the bass turned up to stomach‑pounding volume, or on an average home hi‑fi, or on a car radio in heavy traffic, with a hole in the exhaust. Your mix has to sell the song in each of these situations, so while the engineer considers the finer points which will only be appreciated by those with good quality home stereo systems or a decent pair of headphones, you will be looking for the overall impact. If the mix sounds good from any listening position in the control room, it probably is good.

All studios have two or more pairs of monitors, so you can check the mix on very high‑quality speakers or on the console‑mounted nearfields. You can also have a cassette copy made so you can check the mix on a cheap stereo system, on a Walkman, or in the car. The more ways you can listen to the mix, the better.

Stereo Format

Most multitrack recordings are mixed to DAT these days, but at a professional level, DAT isn't always considered to be entirely satisfactory. For one thing it's only 16‑bit, which means that its sound quality isn't any better than the CDs people listen to at home. The engineer, therefore, has absolutely no headroom to play with, and inevitably there is a margin of unused capability that makes the recording not quite as good, technically, as it ought to be. It won't be too long before we see 20‑ or even 24‑bit stereo formats in the studio on a regular basis, although it might be some years before any one is accepted as a standard. In the meantime, many producers are opting for the 'old fashioned' alternative of analogue reel‑to‑reel tape. They don't use a battered old Revox, however. Top studios have slightly worn but well maintained Ampex or Studer stereo machines that run at a speed of 30 inches per second (twice the 15ips long considered the professional norm) and take half‑inch rather than quarter‑inch tape. Such a machine isn't totally transparent but has a definite sound of its own, and it's a sound that producers like, particularly if a recording has been made on a digital rather than analogue multitrack. The frequency response is, in fact, better than DAT or CD, which can only manage around 20kHz at the top end. Half‑inch analogue at 30ips can go up to 25kHz and beyond, and quiet signals can still be clearly heard below the already low noise floor — even without Dolby SR noise reduction. There are many who will say that half‑inch is better than digital for these reasons, and so many successful records have been mixed to half‑inch that it is difficult to disagree.

When the stereo master is finished, the producer's work continues into the CD mastering studio. This is the final stage, where the stereo master is committed to a U‑Matic video tape or Exabyte data cartridge. After this, no further alterations can be made to the sound. CD mastering isn't quite such a creative opportunity as vinyl mastering used to be (mostly because of the technical limitations of the medium). It is, however, a chance to make sure that all the tracks have the right relative levels, EQing and compressing where necessary. You'll also set the length of the gaps between tracks, and perform any crossfades between tracks that you think are appropriate (and to hell with radio plays!).

When you leave the CD mastering studio, your work as producer is complete and you can look forward to the financial rewards for your labours. Actually, you may also have to look forward to your recording being handed over to specialist remixers — a fact of life that you will accept as gracefully as a true professional would! (See box on the Rapino Brothers.) Next month, in the final part of this series, we'll look at the business aspects of enjoying a long and fruitful career as a record producer.

The Remixer's View — Marco Sabiu Of The Rapino Brothers

The Rapino Brothers are Marco Sabiu and Charlie Mallozzi. (They are not real brothers, obviously — 'rapino' is Italian for robber). They have built up a significant and lucrative following among record companies as writers, remixers and producers, making new versions of original recordings by Take That, Roxette, Sleeper, Suggs, Rozalla, and Dubstar, among many others.

    "I'm not the one who said remixing was necessary. It's the record companies who want remixes, because they want to have their songs played in the clubs. If people like the song in the clubs, they will go and buy the record."
    "Sometimes they tell us what kind of style they want. Sometimes they don't tell us anything, they just give us the vocal and we do what we want. Usually we only keep the vocals and redo everything else, unless it's a band with guitars — then maybe we'll keep the riff. It depends. It can be almost like doing a new song."
    "I listen to the original producer's mix to try to get the kind of vibe that they used, so I can avoid doing the same thing. I don't know whether producers like what we do. We always do a radio version of a remix and sometimes the record company will use it as a single. If that version goes into the charts I think the producer should be quite happy because the album will sell better."
    "I can't see any difference between a remix and a production, because at the end of the day when you do a remix you start from scratch. The only thing you have is the vocal. To me that is a production."
    "On that side we are quite lucky, because when I was in Italy I worked as a sound engineer, so for me now it's helpful to have that kind of knowledge. If I want a sound I know how to do it. I don't have to ask someone else, which would be just wasting time. I think it's quicker and better if you have a good knowledge of the machines you have. We have a Pro Tools 3 system with Logic Audio, so we do everything with that."
    "We quite like to use live drums and real guitar and bass. When we do that, we usually go into a bigger studio because we don't have the space. Then we come back to our home studio [pictured here], transfer everything into Pro Tools and start editing."

The Engineer's View — Paul Gomersall

Paul Gomersall has recently enjoyed the privilege of engineering George Michael's new CD, Older, where George, as usual, is his own producer (for the full story, see the article starting on page 40 this month). Paul's other credits include work with producers Trevor Horn, Phil Collins (as a producer), Stephen Hague, Stephen Lipson, Thomas Dolby, Chris Porter, Laurie Latham, and others.

    "Producers like to distance themselves from the desk, as long as they have an engineer they can trust. It's one of the joys of production. If they have a problem, they might dive in and try and sort it out themselves, but usually there won't be a problem. You are the interface with all the technical stuff so the producer doesn't have to think about that. If the producer comes up with an idea you make it work for them."

What should an engineer do if a producer appears not to hear a problem?

"Point it out. One of the good phrases is 'I think we should listen back to that'. He will be listening by then. There are ways of getting your point across. Diplomacy is a big part of the job."

    Monitoring: "I give him the mix output of the desk in his headphones. The headphone mix is very important to anyone. That's what they are listening to in the studio so that's what they want to hear in the headphones."

Reverb: "The way he sings the consonants, it gives a bit of excitement to his vocal sound. It's being used as another instrument. I generally set up two, three or four reverbs that I think will suit the track as it's coming together and take it from there, and George will pick what he wants."

Mixing: "The mix is an evolutionary thing. Virtually every song we do we work through to the mix. You never get a stage where everything is finished and you start mixing. It evolves, and somewhere along the line he'll want to do a vocal ride so we'll switch the mix computer on, then we'll switch it off again and add some more things and continue like that. I know we're finished when he says 'let's put one down'. He'll take a mix home and come back the next day and probably continue working on it. The programming diminishes as the work carries on and we get more involved in riding levels of things, EQ and other bits and pieces. And then suddenly it's there. It's finished. The rough mixes are really important because sometimes George will ask to refer to something we've done before and then carry on from there. So we keep all the mixes we've done in case one of them had something that was really good.