Recording Techniques (Part 17)
The Mix (3)
Part 17: The mix (3). There's so much to know about mixing and most of it can only be learned by experience. But a little knowledge can go a long way. David Mellor offers even more insights.
I wouldn't like to pretend that I know everything there is to know about mixing; far from it. In fact I don't suppose anyone knows everything that can possibly be known about mixing 4, 8, 16, or 24 or more tracks into stereo. If you look on the bright side, this means that there is much to be discovered in the way of new tricks, techniques and methods of operation. With so many people having home studios, it stands to reason that new developments will come increasingly from people currently outside the inner circle of major professional studios. In this installment, I'm continuing with last month's theme of explaining established procedures and ways of doing things. Combine these with techniques of your own invention and you'll go a long way towards achieving an ideal combination of creativity and professionalism.
More years ago than I care to remember, when I was studying for some very important exams, I found an excellent way of relieving the tension and nerves. I took three tennis balls that happened to be lying around among the debris in my bedsit and taught myself to juggle. I practised first with one ball to get the feel of throwing it up so that it would land in my other hand without consciously having to catch it. Then I added another ball and practised swapping the two back and forth. The third ball was a little more tricky, because you have to keep them all moving and there is always one ball in mid-air. My efforts at juggling tennis balls were rather better than my exam grades, but I never did manage to add that elusive fourth ball.
Mixing a multitrack tape into stereo requires many of the skills of a juggler, but there can be as many as 16, 24 or even 48 balls in the air at the same time. What's more, there may be people gathered around you shouting 'useful' advice like, "Can you get that one a little higher", "I think you should drop it a bit there", and so on. I have great admiration for any engineer who can do a live mix without the help of an automation system and remember perhaps 10, 20 or more level changes during the song and perform them all in the right place with faultless accuracy — but this is part of the job, and with enough practice it can become almost second nature.
The biggest problem with juggling levels during the mix is that you can't sit back comfortably and listen to what's going on. So even if you can do all the level changes with a reasonable degree of ease, you are not able to put all your concentration into the music, where it should be. Automation systems which can remember all the fader movements of the mix may have a 'de-skilling' effect on the actual physical process of mixing, but the engineer now has more time to think about the sounds he or she is creating.
Automation systems are still very expensive. The top systems such as those found on SSL and Neve consoles aim to give a sound quality and degree of control as high as that obtainable in a manual mix. Lower cost systems, which are still very pricey add-ons to a console, use lower quality VCAs (voltage controlled amplifiers) which compromise sound quality slightly, and provide a finite number of fader positions so that particularly when the fader is at a low setting, you can hear the stepwise level changes as you fade up or down slowly. But despite these slight drawbacks, the musical benefits you gain can be enormous. I look forward to a day when automation is truly affordable, in the way that mixing consoles have become affordable in recent years, but unless there is a dramatic leap in technology I can't see it being this year or the next.
One way to get many of the benefits of automation without the expense is to mix in sections. The only equipment you need to be able to do this is a reel-to-reel tape recorder, although a tape editing kit will come in handy too. Songs usually fall into verse and chorus sections, probably with an introduction and middle eight also. If you're unlucky, you'll find that at the changes between the sections there are too many fader movements to be done manually. In this case, it's possible to mix the verses to your satisfaction, and record them to stereo tape, and then go back and mix the choruses separately. When all the sections have been mixed, you just cut them back together into a continuous piece of music.
Sounds too simple? Well it's not all that easy to get right. The song has to be suitable for mixing in sections for a start, with easily identifiable edit points. The way you mix the song is important too. It is virtually essential to have a solid foundation that stays the same throughout the song. Then, if it's only the superstructure that needs adjusting, you won't get any significant level changes at the joins - at least you won't on your second or third try. Mixing in sections is a technique that's well worth having a go at and getting the hang of. Once you have it in your repertoire, then you'll be able to recognise the situations where it's going to be of benefit.
When no-one but a select few had samplers, a casual phrase like "I locked it into the AMS and spun it in" turned the rest of us a medium shade of green with envy because we still had to do it the hard way. Doing it the hard way is sometimes still a valuable option because, let's face it, we don't all have access to the latest in stereo samplers with a fortnight's worth of memory at 48kHz. Fortunately this advanced sounding technique can be achieved with as little high-tech equipment as a humble cassette deck. It may take a bit longer but the result can be almost as good.
But exactly what is a spin in? One problem in almost all recording sessions is getting all the choruses in a song to sound equally good. Since it's very often the chorus that provides the hook that makes people buy the record, every repeat has to be absolutely perfect — but often we find that it just isn't possible to make them all as good as they could be. Fortunately, since choruses have the basic feature of repetition, if even just one happens to be right, we can use that throughout the song. One way to do this would be to mix the song and make several copies of the chorus, and do a cut and past job. This doesn't really work because if the chorus is exactly the same each time, then it will become boring. So what we need to do is to repeat only the elements of the chorus that cause problems — usually the lead or backing vocals.
The term 'spin in' comes from the old procedure that was used in years BS — before sampling. That procedure involved mixing the backing vocals (or whatever) on to a stereo tape, editing the tape so that it would start and end neatly, and then recording it back on to the multitrack. You don't need synchronisation to do this since it's possible to get it right in just a few tries, and it's easy to make marks on the tape to cue you in. With a cassette deck as the spin in source (not everyone uses reel-to-reel) it takes a little longer and a few more tries, but this is what recording is all about really — take that extra bit of time and make it perfect. Once you have the knack of this little trick, then you can laugh at those people who bought expensive samplers to do it, because it's costing you absolutely nothing.
Automation is expensive, but you can get much of the power of automation with a MIDI muting system. Some consoles offer this as standard, some as an extra. Studiomaster announced an add-on MIDI mute processor some time ago but they haven't said very much about it since, which is a pity because it could be a very handy mix tool.
MIDI muting systems need to work in conjunction with a sequencer synchronised to the multitrack tape. A combination of mutes can be set up, which becomes a program, and different programs can be selected via MIDI Program Change messages. Sometimes it's possible to use note-on and note-off messages to mute and un-mute channels, but the end result is the same. A mix always sounds much cleaner if non-contributing tracks are switched off, and this is very difficult to do manually. MIDI muting systems should be made compulsory on every console, in my opinion.
"When no-one but a select few had samplers, a casual phrase like "I locked it into the AMS and spun it in" turned the rest of us a medium shade of green with envy because we still had to do it the hard way."
Mastering on to DAT offers many advantages compared to reel-to-reel tape. For one thing it's extremely cheap in comparison, and that means that you can afford to take a few more chances with the mix and use a lot more tape. (In the heyday of reel-to-reel mastering, the amount of tape that went into the bin was nothing short of amazing. DAT is definitely the medium of choice for the Green engineer, and also for budget conscious studios since the initial capital cost and running costs of DAT are a lot lower than the quarter-inch alternative.) The only disadvantage of DAT is that if you want to edit your recording, then you are going to have to shell out in a big way. The cost of an editing suite (per hour or day) is astronomical compared to that of a razor blade and splicing tape.
When mastering on to tape (reel-to-reel), the procedure is to record first, and then edit the chosen mixes together with leaders and spacers to make up a continuous running programme, perhaps part of an album. There is a strong temptation to try to do the same with DAT and assemble a DAT tape of completed mixes. I would recommend however that you don't do this, and rather leave any assembly until the copying stage when you'll have to hire or borrow a second machine. It's hard enough to mix a track well without having to worry about getting it on to the tape in the right place as well. The best way to mix on to DAT is to use one DAT cassette per track. It costs more, but remember the savings you have already made. If you do it this way there is absolutely no risk of erasing another track on the same tape and thereby wasting perhaps a day's work. I would go further than this and say that if the mix is an easy one with few fader movements, then you should do an identical mix on to another tape for safety.
Chances are that you'll have several attempts at the mix, and since you have plenty of room on the tape you can record several versions and choose the best later, recording over only the ones where you made obvious mistakes. Once you have the definitive mix, it's also worthwhile making another pass with the vocal 3dB higher. If there is a risk of mixing anything too low, when judged in the clear light of the next day, then it's bound to be the all-important vocal. Doing the 'extra voice' mix saves you having to start again from scratch. When you are sure that you have the ideal mix among several attempts on the DAT tape, remember to label the inlay card with the Start ID number of the one you want. When you come to the copying stage, if you're anything like me then you'll have forgotten which one it was meant to be and have to sit down to a lengthy listening session to decide which take was the best.
Unless you're one of those insufferable people who always gets it right first time, then you'll often find that a track sounds really great when you've just finished mixing it, but next morning it sounds terrible. One reason for this is 'ear fatigue', which has both physical and psychological causes. Another reason is that you really need to be able to stand back and get a perspective view before you can mix properly; this is one of the reasons remixers have had such a success. People who do this have the luxury of being able to listen to a multitrack tape and jettison a line that took several hours and buckets of blood, sweat and tears to record. Could you do the same?
The chances are that your first mix of a song will be great, apart from one tiny but important detail — like, perhaps, the vocal being inaudible. But once you have zeroed the console and lost the settings it may be impossible to get back to what you had before and correct the mistake, unless you have an expensive mixing console that has recall facilities. A device used by theatre sound engineers to recall settings from performance to performance is the plot sheet — a diagram showing all the controls on the mixer, on to which you superimpose the settings you use and note any effects. I've included part of a plot sheet for the mixer in my home studio that I made up using DTP software on my Atari. When I've completed a particularly difficult mix, I spend time filling in this sheet, and I have to report that when I've done it, I've never regretted it. When I've been lazy and not filled it in — that's when I wish I had. It's certainly not as quick as a computerised recall system, but it works, and it's a lot less expensive.
So the mix is complete and has received the seal of approval of the artist and the producer. But so far, no-one has heard the results of your labours except you and your studio cohorts. If you are going to take this further, perhaps as far as record or CD mastering, then it's important to do things the right way — or at least in a professional rather than an amateurish manner. Whatever the project you are working on, don't think of a master tape as something that is yours to cherish and to treat with as little or as much respect as you like.
Always work on the basis that some day, sooner or later, another engineer is going to handle your tape. Is he or she going to recognise the work of a fellow professional? Even if you only work on personal projects, remember that copyright in musical works lasts until 50 years after the composer's death. This means that the recording you produce today may conceivably still be earning royalties in the 22nd century (probably as background music in the canteen of the Starship Enterprise).
Let's start with reel-to-reel tape, and consider what a mastering engineer is looking for in a recording. Well basically, he or she is looking for the easy life. If the tape is well recorded, and the spool and box labelling make your intentions perfectly clear, then they'll be happy with the transfer and so will you. If the mastering engineer receives a grotty looking box with illegible scrawl across it, then they'll look forward to yet another of life's difficult days.
I've never worked as a mastering engineer, but in the course of my sound engineering activities I have had some real shockers to work on. One tape of modern classical music (in a style known to some as 'plink plonk') had just the title written on the box. There was no leader on the tape, and it was actually impossible to tell from listening whether it was meant to run at 7.5ips or 15ips. Worse still, I couldn't tell which way round the tape was supposed to be! The moral is to make life as easy as you possibly can for your fellow professionals — that way you get better results. All you have to do is follow a few common sense guidelines. You won't go far wrong if your tape is presented like this:
"Whatever the project you are working on, don't think of a master tape as something that is yours to cherish and to treat with as little or as much respect as you like."
1. The tape itself should start with at least two metres of white leader and end with red leader. The music should start within 2cm or less of the white leader, and the red leader (trailer) at the end should follow around 10 seconds after the reverberation dies away. Items on the tape should be separated by white spacers. There should be no unrecorded tape on the reel.
Tones should be recorded at the beginning of the reel: 1 kHz, 10kHz and 100Hz all at a recording level of 200nWb/m (many recorders are lined up so that this level is at -4VU) for about 30 seconds each. The tones should be followed by a spacer.
2. The spool label should identify the tape with the tape box. At the very least it should have the artist and title, and preferably also the tape speed.
3. The box needs to be packed with information:
|Engineer's contact telephone number|
|stereo (wide or narrow guard band)|
|Noise reduction (if any)|
|Tone frequencies and level|
|Peak level in dB relative to the tone level|
|Leader and trailer colours if not white and red|
|Any unusual features|
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!