Last month we looked at the corrective role of signal processors and at the use of insert points and auxiliary sends. Now let's move on and examine the mix itself, beginning with ways to create the illusion of stereo.

As intimated last month, there are no hard and fast rules when it comes to mixing. If there were, everyone would produce exactly the same result from any given master multitrack tape. Nevertheless, there are guidelines to get you off to a good start and once you have put them into practice, you will be able to decide for yourself which rules to break and which to stick with. I'm assuming that you're starting out with a 4-, 8- or 16-track master multitrack tape and that few if any of the effects have been recorded onto tape. The instrument line-up will consist typically of drums or drum machine, guitar, bass, synths and vocals, as even in these pioneering times, this still seems to be the most popular combination. If part of your instrumental line-up is driven via a MIDI sequencer synced to tape, then the same techniques apply. I'm also assuming that you intend your finished product to be in stereo rather than mono or multi-channel surround sound.

All About Stereo

A misconception exists about mixing in stereo. To many people, it simply means positioning the sounds in a mix by use of the pan pot. Though this is a valid mixing technique, there are lots of other approaches that can work just as well without being quite so obvious. Firstly then, a quick look at the mechanics of stereo in real life.

In fact, most musical sound sources are essentially mono, in that they emanate from a single point source. They may be nearer to one side of the listener than the other giving us the equivalent of the panpot effect above but this alone doesn't create the sense of depth and reality that we are looking for. Actually, the reaction of the ear and brain to off-axis sounds is a good deal more complex than the previous couple of lines might imply, but to investigate further would mean an interesting, though unacceptably large digression.

The sense of depth that we get from hearing a sound in a normal environment is not merely a function of its position but is largely due to the addition of reflected sound. You may instantly think of reverberation, which is certainly a step in the right direction, but in many everyday acoustic environments, the reverberations may be so heavily damped so as not to be perceptible as such. Take your living room or flat or whatever. It probably doesn't sound very live if you clap your hands in there, but if you were able to do a direct comparison between that sound and the same handclap in a completely dead or anechoic room, the difference would be startling due to the absence of those vital few early reflections. In recording, everything is usually kept as dry and separate as possible. Close miking techniques eliminate most of the room's natural reflections, and the room will probably be as dead as we can practically make it anyway. What we have done is robbed the sound of its spatial identity and recorded it in mono onto one track of tape. In exceptional circumstances, you might have found enough spare tape tracks to use two mics to record in stereo but the chances are that the room ambience will still be lacking. An example of this may be two mics positioned close to a piano or drum kit. An alternative approach is to record these instruments in stereo in a live room, but at home this is often impossible and, if you're using a drum machine instead of real drums, the sounds will be very dry. Likewise any synth sounds are going to be strictly lacking in natural ambience. Having set this grim scenario, we now have to look at things we can do in a mix to recreate the missing sense of space other than simple left/right panning.

False Ambience

A stereo ambience can be created to complement a mono sound using a reverb unit but a simple spring unit is likely to have too long a decay time to add ambience without adding a noticeable amount of reverb at the same time. A better approach is to use a digital unit. These generally accept a mono input and provide two outputs to simulate the different reflection patterns reaching the ears of the listener, and most will allow you to set up a short reverb time of between 0.4 and 0.8 seconds which will give an impression of roominess without much in the way of obvious reverb. Some machines even allow you to set up an early reflection pattern without the following body of reverb and this can be useful on vocals where you want to add presence in the spatial sense of the word without cluttering the sound with reverb. Of course, if you want a certain amount of obvious reverb as well, then fine, just select a longer reverb time or use a spring.

You may often face the dilemma of having one reverb unit and requiring it to do several jobs at the same time. You can plan for this at the recording stage by adding a little reverb in mono to the desired tracks as you record and then adding an overall short stereo reverb to them all when mixing. This will reintroduce a sense of space that simply adding reverb in mono will in itself not achieve. Of course you might encounter a tape that's already recorded or you may not have a digital reverb so what can you do, apart from shake your fist at the skies? Fortunately quite a lot.

Synthetic Ambience

One source of ambience that tends to get overlooked is the natural ambience of rooms themselves. Wouldn't it be nice if you could steal the ambience from your bathroom and add it to your drum machine? Well now, I wouldn't have mentioned it if you couldn't do it, would I?

Half an hour later, after evicting your sister from the shower and running some extension leads up the stairs, the set-up should look something like this. One of your hi-fi speakers now resides in the bathroom facing a hard, preferably tiled wall, and this is fed from your hi-fi amp which in turn has its aux input wired to one of the aux outputs on your desk. Turning up the drum track's aux send on the desk may now elicit howls of protest from the rest of the household as the drums come thundering forth from the bathroom. The next step is to pick up this new 'live' sound.

A little experimentation will pay dividends here but what we're trying to do is to position two mics, preferably the same type, in the bathroom such that they pick up as much reflected sound as possible and as little direct sound from the speaker as possible. A couple of Tandy PZMs taped to the walls are ideal for this job due to their excellent clarity when working at a distance. They're also very cheap. The next bit is simple, just make sure nobody flushes the loo for the next half hour and get on with your mix pretending that you have the latest mega-expensive digital room simulator on the other end of the line rather than a bathroom.

"The unsubtle use of flangers is nowadays considered to be rather passé but they do have their more creative uses if you use a little imagination."

There are dedicated stereo simulators which you can use to widen mono sounds and these use delays to simulate the sound bouncing off adjacent walls. Usually you only get one reflection per side but you often have the option of sweeping the delay in a slow chorus-like manner to add a bit of movement to the sound. I have found such devices to be most effective on synthesised sounds or the types of sound that usually benefit from a little chorus, but they don't really create ambience in the true sense of the word and they can be very expensive. The stereo simulation programs of the Alesis Midifex are somewhat better in this respect as they utilise a greater number of reflections and are altogether more convincing though there is no modulation facility. If you are limited in resources; here are a few things that you can try for yourself.

The following treatments work well on synth sounds, guitars and vocals but may not give the desired effect on drums.

DIY Ambience

First we have the old trick of panning a sound hard to one side and then placing a delayed version at the other side, the delay being in the order of tens of milliseconds. The mason that this is so effective is that it simulates to some extent what happens when we hear a sound in real life. First we hear the direct sound and a short time later we hear a reflected version bouncing off a wall or floor. True, subsequent reflections complicate this state of affairs in nature but the single delay can be surprisingly effective nevertheless. One thing to we aware of though is that this kind of trick is not always mono-compatible. As students of the art will no doubt realise, adding a signal to a delayed version of itself sets up what is known as a comb filter, which is the basis of flanging and phasing. Depending on the delay chosen, some frequencies will add to each other and some will cancel each other out. Whilst this is not apparent when the two sounds are panned to either side, it will show up as tonal colouration when the sounds are summed to mono. If you have a delay unit that offers you both in-phase and out-of-phase outputs simultaneously, then try positioning the dry sound in the centre of the mix with the two outputs from the delay unit panned to either side. This will give an impressive illusion of depth and, when you sum to mono, the two delayed outputs will cancel leaving you with the original mono sound. Neat.

And that's not all, if your DDL has a built-in modulation facility, as most of them do, then you can add a little slow, shallow pitch shift to add a sense of movement as well as width. If, however, your DDL is tied up with more important work, how about that old analogue chorus or even the flanger pedal you've got tucked away somewhere? You can do the same trick here by feeding a dry signal to one side and a chorused version to the other, keeping the levels roughly equal. The result is a very wide, three dimensional sound that works particularly well on synths, basses and clean guitar. If you use a flanger, turn the feedback control down to minimum and you should get a similar effect.

The unsubtle use of flangers is nowadays considered to be rather passé but they do have their more creative uses if you use a little imagination. For instance, a bass line generated by a synth or even a bass guitar can sometimes start to sound woolly in the context of a mix and may tend to get lost or sound boomy. Feeding this through a flanger set to a a slow speed with only shallow modulation can give a harder edge to the sound without it sounding too flangey and may just succeed where EQ fails. By adjusting the range control, which really varies the delay time, you can tune the effect to some extent which may allow you to emphasise some frequencies or cut others as needed. There's no magic formula here, just sweep through the range until the result sounds good to you. You can then either use the new flanged sound on its own or add a little of the dry sound to make the effect even less obvious, it's up to you.

As to other uses for flangers, have you considered this? Split a signal into dry and flanged version, as before, but then feed the output of the flanger via a noise gate. Trigger the gate via its key input from a rhythmic sound such as the drum track and then set the release time for half a second or so and see what happens. Every time a drum beat occurs, a flanged version of the dry sound will be added to the mix and then will fade away according to the time set by the release control of the gate. You might even try to miss out the dry sound entirely and have say a burst of flanged string synth popping up only where triggered. Use both channels of the gate with the same set-up but with different attack and release settings and you could get a flanged sound that moves from left to right as it's triggered. Has this set you thinking? I hope so because there are endless tricks based on this arrangement which all sound as original as your own application of the technique.

Pitch Changing

Until recently I wouldn't have dwelt too much on the uses of harmonisers or pitch shifters because the good ones were too expensive for widespread home use and the cheaper ones were not that much good, to be honest about it. This last year has changed all that though and we now have the pitch shifter in the Yamaha SPX90 and Roland's little RPS10 pitch shifter, both falling well into the semi-pro price range and yet still turning out an excellent performance. The creative uses of large pitch shifts are somewhat limited as the Mickey Mouse effect becomes rather obvious, but using just a fraction of a semitone of shift can fatten a sound enormously and doesn't cause 'churning' like a chorus unit. The preferred technique is once again to place a dry sound at one side of the sound stage and a shifted version at the other but if you have a dual channel device like the SPX90, you can keep the dry sound in the middle and then place an upward shifted version at one side and a downward shifted version at the other; this really fattens things up. This treatment is extremely effective and yet is still subtle enough to be used on virtually any sound source, voice or instrument without giving the game away.

Next month we conclude this short series by sitting in on the mix itself.