How To Record Singing Guitarists

Most budding engineers will, at some time, have to tackle a performer who wants to sing and play guitar at the same time. Mic type and placement can make a big difference to the quality of the result obtained, as
Hugh Robjohns explains...

I once read a humorous definition that stated 'Art is a Science with more than seven variables!' Some people do indeed approach recording music as a scientific experiment, while others see it as as an artistic experience. I think the truth lies somewhere in between — there are definitely elements of physics involved, which are wholly predictable, but there are also those unquantifiable artistic elements which can make the difference between a good recording and a great one.

Possibly one of the most common recording challenges is the singing guitarist. I'll bet every reader has had to record this combination at one time or another — but how many have actually achieved the kind of high‑quality results that are possible even with simple equipment? I've seen it all too often: a couple of unselected mics are just plonked in front of the guitar's sound hole and the singer's mouth, both mics are faded up on the mixer, and then great fistfuls of EQ are needed to try and sort the mess out. The end result is usually a muddy and lifeless combination, often with poor control and separation. So what's going wrong and what could be done to improve the situation?

The Basic Problem

Every recording session must start with a question: what are you trying to achieve? This decision is not yours alone, of course — you must consider what the musician is trying to do too.

Consider our notional singing guitarist:

• Do we want a recording of the solo performance as it stands, or do we want to record the individual elements in isolation, so that they can be re‑balanced with other instruments which might be recorded at some other time (before or after this session)?
• If it's the performance we're after, are the acoustics of the environment appropriate, or will artificial reverberation have to be used?
• If we want isolated recordings, can we record each instrument (guitar and voice) on separate passes by way of overdubs, or will that destroy the 'performance' aspect of the piece?

The next question concerns the quality and characteristics of the instruments themselves, and this introduces a concept that may be new to some — what is the polar response of each instrument and how does this change with frequency? In other words, how does the sound of the instrument change as you move around it, and where is the optimum balance of characteristics? These questions can only be answered by carefully listening to a rehearsal of the material in the recording room.

Many engineers never bother to take the time to walk around the performer to learn about the environment's acoustics, what the performance is about, and the characteristics of the instruments. The whole essence of recording is capturing the performance — not what you think the performance should be. Of course, if there's something wrong with the performance or environment in some way, then it becomes the recordist's job to try to fix it — ideally by sorting out the problem at source. This might mean moving the performer to a different part of the room, finding a more suitable recording location, or using screens to modify the acoustics. It might even mean persuading the performer to use a different instrument, or to change playing style! It will often require you to modify the microphone technique, and as a last resort, there's always equalisation — but this really is a last resort. I know some recording engineers who feel that using EQ is an admission of failure — failure to get the right microphones in the right place, in the right venue! It might be an extreme point of view, but it is worth bearing in mind; if you don't start off with a good sound, you'll never be able to create one with the EQ — or any outboard processing, for that matter.

A fascinating insight into how musical instruments radiate sounds is given in chapters three and four of a book called Acoustics and the Performance of Music by Jrgen Meyer, ISBN 3 920 11256 3. It's quite an expensive book, so I suggest you persuade your Lending Library to track it down.

Analysis

Experienced engineers can very quickly gather all the information they need about instruments, venue and technique, but this only comes with practice. I've always found the best approach is to get the performer in the studio and rehearsing straight away, while I listen carefully to the performance, and wander around learning about the instrument's characteristics. (Many people find it easier to perceive what a microphone will capture by sticking a finger into one ear. It might look silly, but if it helps, do it — people will remember the brilliance of your recordings, not your strange antics in the studio!).

In the case of an acoustic guitar, you might typically expect the low‑frequency content of the sound to be emphasised as you move around towards the base end, and the high‑frequency harmonics and finger‑board noises to become more dominant towards the neck, but every guitar is different and the amount of variation in sound quality can be surprising. Also listen to the reflected sound within the room. There are several questions you could ask yourself:

• Are there any detrimental reflections in the room?
• Would it be better if the performer was seated somewhere else?
• What is the guitarist's technique like? Does it create a lot of finger noise, for example, or is the instrument played to deliberately emphasise the upper harmonics? Your recording technique might have to reduce or capture these attributes appropriately.

A Simple Technique

Let's assume that we've followed the suggestions above and have a good idea of how the environment, the instrument, the technique and the performance work with each other and what we're trying to achieve with the recording. The next thing is to select and position the microphone.

Let's assume that we want to capture the solo performance as it stands. The simplest solution is to use a single microphone, carefully positioned to get an appropriate balance between voice and guitar. (If the venue had particularly good acoustics, you might choose to use additional distant microphones to capture the natural reverberation. If not, you'd have to create something appropriate from a reverb unit).

One trick which often works is to put up a dummy vocal mic, which will keep the vocalist's head facing in the right direction when they sing!

In this case, because the microphone will have to be a few feet away from the performer so that it can 'see' both voice and guitar, the polar response of the microphone will probably not be particularly critical — the only sound you might want to reject is the room's own acoustic. Most engineers would use a microphone with a cardioid pickup pattern, but it would be worth experimenting with an omnidirectional mic, as this would give a different balance between the performer and room acoustic.

The electrical operating mode of the microphone will have an affect on the quality of the resulting sound. Ribbons, electrets and capacitor mics will tend to capture a lot of the upper harmonics and transient attack of plucked strings. Moving coils tend to have a duller and slower sound quality. These aspects will influence your choice of microphone, or, if your choice is limited, where you position the mic to compensate for its inherent qualities.

The optimum position of the microphone is the point where the guitar has the sound quality you want, as does the voice, and the two have the perfect balance relative to each other. Obviously, moving the microphone upwards favours the voice and down favours the guitar. Moving the microphone towards the base of the guitar reduces its upper harmonics and string noises, and often increases body and warmth. Poor vocal technique is unlikely to be a real problem in this situation, as the microphone is probably going to be two or three feet away from the mouth, but if the singer looks around or down at the finger board while singing, the balance will be affected. One trick which often works is to put up a dummy vocal mic, which will keep the vocalist's head facing in the right direction when they sing!

This single‑microphone approach will produce a mono sound source, of course, but this will sit very comfortably within a stereo room acoustic or artificial reverberation. It has the potential to sound very natural and clear, is well suited to a surprisingly wide range of recording situations, and you will undoubtedly learn a great deal about the characteristics of the instruments and microphones involved by using this setup. However, if you want to process the voice and instrument separately, you'll need to use a different technique.

Multi‑Miking

A common requirement is to record the singing guitarist's voice and guitar at the same time, but with the aim of producing isolated sounds, so that independent signal processing may be applied. This might be compression on the voice, or a chorus effect on the guitar, for example. You may also want to pan the voice and guitar to separate positions in the stereo image. Whatever, the reason, the key point here is that we need isolation between the two sound sources. Clearly, a minimum of two microphones will be needed (and possibly three, if a stereo guitar track is preferred).

So what tools might you have at your disposal for achieving separation between the recorded sounds? The best one, naturally, is physical separation, but unless the guitarist has extendable arms, you're unlikely to be able to achieve this. The ideal solution, of course, would be to record the guitar in isolation and then overdub the voice, but this often destroys the 'performance', so I'm ruling that out!

The next option out of the bag is acoustic screens, but trying to arrange one horizontally between the guitar and the singer's head is likely to be a bit problematical — so it's all going to have to be done with microphone choice and placement.

What each microphone needs is lots of the wanted sound, and very little of the unwanted sound — it is the ratio between the two which is critical. In other words, the guitar microphone should not 'hear' the voice, and vice versa. There are two ways of controlling how much unwanted 'spill' reaches a microphone.

• The first is to place the microphone as close as possible to the wanted sound and as far away as possible from the unwanted sound.
• The second technique is to use directional microphones, so that you can position the polar response null(s) (to reject the unwanted sound. This latter point is the single most important aspect of using a directional microphone: where the polar response null is pointing is almost more important than where the front of the mic is pointing!

Polar Patterns

What kind of microphone polar pattern is going to be best for each microphone in this situation? Consider a microphone positioned close to, and facing, the guitar's sound hole: this will capture lots of guitar sound, but where is the voice coming from? Almost directly above it, at pretty much 90 degrees, in fact.

If you were using a cardioid microphone in front of the guitar, on‑axis to the sound hole, the voice would be arriving at right angles, in a place where the polar response would be about 3dB less sensitive than the front. Even taking the distance between mouth and microphone into account, the voice would only be perhaps 6dB quieter than the guitar, which is certainly not enough separation to allow independent signal processing! With a similar setup for the voice, the result is the classic situation where there's almost as much vocal on the guitar microphone as there is guitar on the vocal microphone. There are two faders on the mixing desk, neither of which controls the sound you think it's controlling, and as their relative balance is altered, so the sound quality changes — sometimes rather dramatically!

The solution is to find a microphone polar pattern that rejects sound from 90 degrees, so that the on‑axis sound will be the guitar, and the rejected sound will be the voice. There's only one microphone polar pattern which fits this description, and that's the often‑ignored figure‑of‑eight. There are dedicated figure of eight microphones available (the BBC uses a lot of Coles 4038 ribbon microphones, which are fantastic in this application), but most people will probably only have access to this polar pattern as an option on a switchable studio capacitor mic, like the AKG C414, the Neumann U87, or more modern equivalents.

Remember that the rear lobes of figure‑of‑eight microphones are the same size as the front and of opposite polarity. This is unlikely to be a problem in the studio, because there will be little sound reflected off the walls which the rear lobes could capture. However, it might be significant in a live‑sounding environment, where the figure‑of‑eight mics could sound more distant than they really are.

Positioning

Positioning the guitar microphone properly requires a little thought and experimentation. The microphone has to be reasonably close to the guitar to maximise the level difference between guitar and voice, which means that it will only 'see' a small area of the instrument. The consequence is that moving the microphone a few inches one way or the other will have a pronounced effect on the instrument's sound quality, in terms of the balance between fundamental and harmonics. So the first thing is to locate a position which gives the sound quality you're after. Something to beware of is a guitarist who likes to swing the instrument about — not only is there a risk of hitting the microphone, but with such a close mic technique, changes in distance will alter the sound quality noticeably. Figure‑of‑eight microphones also suffer huge amounts of bass tip‑up (proximity effect) when used close, so you will probably not want to place the microphone towards the base of the guitar.

Once you've found the guitar's 'sweet spot', you need to angle the mic carefully to optimise the position of the side null. The easiest way I've found is to route the guitar mic through to the studio headphones, and then adjust the microphone for minimum voice pick‑up while the performer rehearses. To give you an idea of the separation possible with this technique, using Neumann TLM170s and U87s, I have consistently achieved separation of about 24dB between guitar and voice, on each microphone. This is more than enough to allow independent stereo panning and processing, and even allows one part to be overdubbed without worrying about spill off the original recording on the other track. If you want to record the guitar in stereo, this technique can be extended to use a pair of figure‑of‑eight microphones, either in a coincident arrangement or a spaced arrangement, depending on the type of stereo presentation you're after.

Using a figure‑of eight microphone on the voice is sometimes not so successful as the guitar, not for lack of separation, but because of poor microphone technique. The figure‑of‑eight is especially prone to wind‑blasting problems, and there is the proximity effect to consider too, where small changes of distance between mouth and mic have a significant effect on the bass content of the recording. If you have a good signer who can regulate the distance between their mouth and the mic, and who can avoid plosive popping, great! Unfortunately, however, this is rarely the case.

The solution is to use a well‑shielded hyper‑cardioid mic for the voice. Again, this will suffer from both popping and the proximity effect, but nothing like as badly as the figure‑of‑eight. Its polar response nulls are at about 135 degrees to the front, so careful angling should allow decent amounts of guitar rejection to be achieved.

The Bottom Line

Clearly, these techniques are not well suited to live stage recordings where the singing guitarist is part of a band. Often, in this situation, there's a greater need to reject spill from other closely‑sited musicians and the foldback monitors! However, these ideas do work extremely well in the controlled environment of a studio (professional or at home) and allow much more accurate balancing and post‑production than could ever be achieved with a two cardioids.

If you have never tried it, I thoroughly recommend giving the single‑microphone approach a go first. Most people are amazed at the quality that can be achieved, and it certainly reinforces the view that the fewer mics the better! It will also teach you more about acoustics and microphones than any book.

Remember to always consider the direction from which unwanted sounds are approaching the microphone, and then angle the null(s) to reject them. Getting things right in the studio always works better than fiddling about with EQ and outboard in the control room, and someone who understands what they're doing can always produce better results with a single mic straight to tape than an idiot with the biggest SSL, the entire Neumann catalogue of mics, and walls full of outboard gear!

Above all, enjoy the session and have the confidence to experiment. If it sounds right, it is right — and if it doesn't, go back into the studio and try something else!