It doesn't matter whether your backing track comes from a MIDI sequencer, a live rock band or a 60-piece orchestra, if the vocal track isn't recorded properly, the end result will still sound disappointing. Paul White offers advice on obtaining a great vocal sound.

Human Nature

Vocalists are human too, you know — they can suffer from nerves, stress and lack of confidence. Even if a vocalist has had loads of stage experience, the recording studio can still be a daunting place where every little mistake is exposed for all to hear — and there's no real adrenalin flow or audience feedback to keep things moving. Indeed, the recording session is often the first time a singer hears his or her own voice as other people hear it, and that can be a pretty disturbing experience, even for someone with a naturally good voice. This is where the right kind of support and encouragement, both psychological and technical, can make all the difference.

The first prerequisite is to make the atmosphere as stress-free as possible. Some singers feel uncomfortable if other people are present while they work, so ask them if they would prefer the rest of the band to go out for a beer while the vocal part is recorded. It also helps if you explain that they're in a much better position than a live performer who has to get everything right first take — in the studio you can have as many goes as it takes to get the piece right, and any small mistakes can be replaced without having to do the entire thing again from the top. An engineer with a sense of humour is a major asset, and you'll find that the best engineers and producers in the country are successful as much for the way they handle people as for their musical and technical skills.

The next psychological boost is in providing a good headphone mix which the vocalist is comfortable with; here it's worth noting that for many newcomers to recording, the first stumbling block is in getting used to working with headphones on. Semi-enclosed headphones feel more natural, as they allow in some sound from the outside world, thus reducing the feeling of isolation that fully-enclosed headphones often produce.

Few things sound less confidence-inspiring than a completely dry-sounding voice — which is exactly what the microphone picks up in an acoustically treated recording room. The trick is to add enough reverb to the headphone mix to make the vocalist feel secure. You can add quite a lot of reverb, if necessary, as this doesn't have to be recorded to tape — you can add the correct amount of reverb when you come to mix the song. It also helps if the vocalist has access to a volume control so he or she can turn the monitor level up or down.

Aside from flattering the voice and increasing vocal confidence, reverb is also a tremendous aid to pitching notes. It helps simulate the way we perceive sounds in real life and, because it causes notes to sustain or hang on, it helps us pitch notes by allowing us to compare them (quite subconsciously) with the sustain from previous notes.

Posture during performance is a subject with which classical singers are intimately familiar, but pop singers tend to neglect. Wherever possible, the singer should be standing, and there should be no tight clothing around the neck to restrict breathing. Another often-overlooked point is the question of temperature, as a room that is too hot or too cold can cause a singer to pitch slightly sharp or flat. This may be due to the effect of the air temperature on the vocal chords, but equally, it may be due to the slight difference in air density at different temperatures. Whatever causes it, most classical singers know that it happens, so maintaining a comfortable but not excessively warm working temperature is a good idea.

Some singers like to have an alcoholic drink (or several) to steady them, because, after all, alcohol is a relaxant. Bear in mind, though, that it is foolish for a vocalist to drink more than a little alcohol, as it nearly always affects performance adversely. Soft drinks are probably a safer bet, with perhaps a TCP gargle to shift any croaks.

Always record the first take, even if it is supposed to be a practice run, because very often the first take is the best. So long as the performance is generally good, you can run through again and drop in any sections that could be improved. Many professional engineers like to record several complete vocal takes on different tracks, patching together a final take from the best pieces. However, this is seldom practical in a home situation, unless you have 16 or more tracks to play with. A method less wasteful of tracks is to record a take all the way through, then do it again on another track. Keep whichever is best and replace the worst one with a new take. Once you get a take that has the right feel and is mainly correct, then drop in the bits that could be improved, giving the singer a couple of lines run-in to get the feel of things each time you replace a phrase. Furthermore, if the timing and pitching between the two takes is consistent, you have the opportunity of mixing in some of the spare take to create double-tracking effects — to strengthen choruses, for example.

We've managed so far not to mention equipment, for the simple reason that the performance is what matters most; like photography, if the subject in front of the lens isn't up to scratch, no amount of expensive equipment will save the day. Assuming that this aspect is now sorted out, let's take a look at microphones.

Taking The Mic

The vast majority of recording professionals use capacitor microphones (sometimes called condensers) as these are built around an ultra-lightweight diaphragm assembly, which enables them to capture all the subtle, high-frequency detail that less sophisticated mics tend to miss. These can be very expensive, costing up to £1000 or more, but their cheaper relatives, the electrets, can also give excellent results. One of the cheapest electret mics available is the Tandy PZM. which retails for around £30, and because it's such a useful little microphone, there will be an article dedicated specifically to it in the near future. The PZM mic is an odd-looking device which comprises a small microphone capsule suspended over a square, metal plate. If you look closely, you'll see that the microphone is actually pointing at the plate, with just a small gap between it and the surface of the plate. In fact the microphone capsule responds to the minute pressure changes occurring in the gap, and without going into too much detail at this time, it enables the microphone to pick up sounds very naturally, due to the constructive way in which it utilises reflected sound.

But you don't get anything for nothing, and the down side of this particular microphone is that it must be positioned on a fairly large, flat surface such as a wall or table top to work properly; if the surface is less than around three feet square, the bass response suffers. Furthermore, you need to work fairly close to this microphone, as it isn't particularly sensitive. Replacing the 1.5V preamp battery with a 9V battery helps a little, but you still need to be closer than two or three feet in most applications. If you want to carry out this modification to the PZM, the easiest way is to tape a 9V battery to the outside of the preamp box and solder a battery clip (by its leads) to the original battery terminals inside the box. Take care to solder the red or plus wire to the terminal that normally connects to the plus end of the battery.

Back-electret microphones are becoming more popular because they can be built relatively cheaply, yet offer many of the advantages of a true capacitor microphone. In fact some of the best microphones currently available are back-electrets. Furthermore, whereas true capacitor microphones require a 48V phantom power supply from the mixing console, the less-costly back-electrets are often designed to run on batteries. A very popular model is the AKG C1000, which is suitable for vocals and a variety of instruments. It can run on either phantom power or a 9V battery and you can buy one for under £150, which is around the same price as a decent dynamic live microphone.

Of course you could just use your regular stage mic; something like a Shure SM58 or Beta 58 dynamic works well in a recording situation but does produce a quite different sound to the capacitor microphone. The high-frequency detail is less pronounced and the mid-range often seems emphasised, giving a solid, confident sound that works well in rock situations. Such mics often have a distinct presence characteristic — a deliberate treble boost that helps them 'cut through' in live performance. Different models apply this boost in different parts of the audio spectrum, usually between 3kHz and 6kHz, which is why particular models of microphone suit some voices better than others. Dynamic microphones need no phantom power or batteries and are generally more rugged than capacitor models. However, they are less sensitive and, consequently, are of limited use in recording quiet acoustic instruments or sounds at any distance. They are also less adept at interpreting the subtleties within a sound.

Keep Your Distance

In a live situation, the singer often has to almost swallow the mic to get enough level without feedback setting in. This is less than ideal in a recording situation, for a variety of reasons, the first being that if you're very close to a mic, the output level will change quite noticeably if the mic position is changed only slightly. On the other hand, if you put the mic a foot or so away, then moving your head by an inch or two won't cause any problems. The other consideration with close-miking is the proximity effect — a significant bass boost which occurs when the sound source is closer than an inch or two from the mic. A good live vocalist can use this effect to advantage, but in the studio, it's just another source of unpredictability.

A typical studio will have the mic set up on a stand so that it is pointing directly at the singer's mouth from a distance of 6-12", as shown in Figure 1. There will invariably be a pop shield placed between the mic and the mouth, because even the best mics tend to pop on the 'plosive' B and P sounds. These pops are caused by blasts of air from the mouth slamming into the diaphragm of the microphone; the resulting noise can completely ruin a recording. A pop shield is a simple, fine-mesh shield made from plastic, fabric or metal and it breaks up the blasts of air, dissipating their energy in the form of turbulence. Normal sounds pass straight through with no effect.

You can buy pop shields, but it is far cheaper and just as effective to make your own. The classic DIY studio approach is to bend a wire coat hanger into the form of a square and then cover this with nylon stocking, but it will look much tidier if you spend a couple of pounds on a 6" wooden needlepoint hoop and fix the material in that. Alternatively, a wire-mesh, frying pan splash guard will work perfectly well. You can either fix the shield to a separate mic stand or improvise some form of fixing so that it shares the main mic stand. Don't be tempted to record vocals without a pop shield, as even those mics with spherical mesh windshields are prone to popping if used without one.

There's no reason not to use a PZM microphone to record vocals and it can be taped directly to a wall or mirror in front of the singer. A pop shield is vital, as these mics seem more prone to popping than most other types, but they are capable of producing a very natural vocal sound, so don't let their low cost put you off. A typical arrangement is shown in Figure 2.

Compression Counts

Professional engineers invariably use an electronic compressor to keep the vocal level even; pop music backing tends to have very restricted dynamics, which means that the vocals need to be made very even too, otherwise loud phrases will stand out, with quieter phrases getting lost beneath the music. Compressors are quite simple in essence, though there will be an article dedicated to explaining them in a forthcoming issue. Their main task is to turn down the level of any sounds exceeding a level threshold set by the user. This keeps loud sounds under control and prevents them from being unacceptably louder than the quieter sounds. The actual degree of control exercised in this way can be subtle or very dramatic, depending on the type of music and on the dynamic range of the vocal performance.

An experienced engineer tends to apply a little compression while recording and then a little more when mixing. This is a good way of working, so long as you don't add too much compression at the recording stage — once the sound is compressed, you can't uncompress it! However, compressing during recording does help to get a better signal level onto tape because it controls the peaks that might otherwise pause distortion. Different compressors have different numbers and types of control, but as a general rule, the so-called 'soft-knee' compressors offer a more subtle means of controlling level than the hard or strict-ratio types, which exert a far more rigid degree of control. However, for rock applications, the 'deliberately compressed' effect of hard compression can help a mix sound more exciting. Again, all these considerations will be explored when we look at compressors in more detail, but for now, I'd just like to give a few ball-park settings which will get you off to a good start.

"Back-electret microphones are becoming more popular because they can be built relatively cheaply and yet offer many of the advantages of a true capacitor microphone."

If your compressor has an 'auto' attack and release function, then use that; otherwise set the attack as fast as possible and the release time to around half a second. With ratio compressors, select a ratio of between 2:1 and 4:1 for subtle compression and up to 10:1 for a stronger, more obvious effect. If you're going to compress while recording, err on the side of a lower ratio of, say, 2:1 or 3:1. Set the threshold while watching the gain reduction meter, and aim for a reduction of around 10dB or thereabouts on peaks.

If no compressor is available, then it's up to the vocalist to move further from the mic when he or she knows that a loud phrase is coming up, and closer to the mic when a word or phrase will be quiet. With a little practice, this kind of vocal technique can be quite effective and is certainly good enough for making demos.

Vox In The Mix

For a natural vocal sound, try to get away with as little EQ as possible or, ideally, none at all. If you do need a little general EQ, see if you can achieve it with your shelving high and low controls rather than with sweep mids, because shelving controls have a smoother, more natural-sounding response. If the vocal track is at all sibilant (over-pronounced S and T sounds), then you can try to find the offending frequency range using your upper-mid sweep control and then apply a little cut. Seriously sibilant sounds must be tackled with a de-esser, (a frequency-conscious compressor); these will be explained in detail along with compressors in a future article. An easier option when working with people who do have sibilant voices is to use a dynamic microphone rather than a capacitor model; the limited frequency response may work in your favour and filter out much of the sibilance at source.

Other than some compression and maybe a touch of EQ, all that you need to make your vocal sit properly in the track is a little reverb. Usually the main vocal will be panned to the centre, with the stereo reverb outputs panned hard left and right to create the illusion of space. Listen carefully to professional records and try to hear the type of reverb they're using; is it long or short, bright or warm? Often you'll find they're using less reverb than you think and many-demos sound cluttered simply because there's too much reverb on everything. In pop music, you're after a hard-hitting sound most of the time, so be sparing in your use of reverb or the impact will be lost. This, incidentally, applies just as much to backing instruments as it does to vocals. Bright reverb sounds are popular on vocals but tend to emphasise any sibilance present in the recording so, again, use your ears and choose a warmer reverb sound if you detect any problems.

Most importantly, remember that everything I've talked about here is very general; take the points as guidelines and not as laws set in stone. Most creative records are created by people who break the occasional rule for an artistic reason. Take the nasal, almost telephone-like vocal sound used by John Lennon on some of the Beatles' tracks or the heavily-flanged sound used by Kate Bush on the album, The Dreaming — both examples work on an artistic level, yet neither approach can be considered a standard vocal treatment. The trick is to learn the ground rules first, then when you know you can get a good, basic vocal sound, you'll have more confidence when it comes to experimentation.

Finally comes the question of how to balance vocals in a mix. Different people have different ideas of what is exactly the right balance, but you'll find that there's only a narrow range between vocals that sound obviously too loud and too quiet. Even so, it is quite possible to do what you think is a perfect mix, then find that when you play it back next day, or possibly on a different music system, the mix sounds all wrong. There's one simple trick that I find helps me get the balance right first time, and that is simply to set up a mix, then step into the next room to listen to it — leaving the door open, of course. For some reason, the whole mix is much easier to judge, and as well as helping you balance the voice, it will also help you pick out any instruments that are too loud or too quiet. I've mentioned this point to several professional engineers and record producers, and those who've tried it all say it works.