Like I said, mixing well‑recorded acoustic guitars isn't usually a strain. Badly recorded acoustic guitars, on the other hand, can be a nightmare — and this is often the case with Mix Rescue submissions!
Duff tuning can be the most insoluble problem, because the out‑of‑tune instrument won't ever really sit properly in the mix alongside other parts, no matter what you do with effects. Celemony's Melodyne DNA may be impressive, but re-tracking the part still remains the only truly satisfactory option here. A recording with rubbish‑sounding recorded room ambience is usually similarly intractable, because there's precious little you can do to get rid of it after the fact.
An unattractive basic tone can take quite a bit of head‑scratching and some very careful EQ to remedy, but to be honest this kind of thing is typically the least of the problems. The big difficulties arise when the mic has picked up too much air‑resonance boom, pick noise, or fret squeak. It's fortunate that lots of mixes don't have any room in them for acoustic‑guitar low frequencies, which means you can sometimes just get rid of booming problems by cutting away most of the low end of the guitar (just remember to do this before you compress, to avoid the lumpy LF skewing the compressor's gain reduction). However, where the instrument is more exposed in the mix, the best solution I've been able to come up with is using an EQ with very narrow digital peaking filters to pull down the fundamental frequencies of the notes that boom. The disadvantage here is that several narrow notches clustered together in the 100Hz region can really mess with the phase‑relationship between the instrument's different frequencies, resulting in a strangely 'hollowed out' sound — so if possible I'd recommend either automating each EQ band to cut only when needed, or using one of the more processor‑intensive linear‑phase EQ plug‑ins now available.
Reduction of pick noise is such a common problem that lots of ways have been devised to deal with these brief high‑frequency transients. Tom Lord‑Alge uses simple compression: "If I hear an acoustic guitar that's a little 'spikey', I put a compressor on it with a very fast attack time, to lose some of that pick spike.” In drastic cases, though, it can be difficult to take off enough edge in this way without causing unmusical gain‑pumping artifacts, in which case more frequency‑selective processes such as de‑essing, multi‑band compression, or dynamic EQ can allow you to target the high end more accurately. Another technique which has more recently gained currency is to apply some kind of transient‑shaper plug‑in in place of more traditional threshold‑based dynamics. "I use [Sony Oxford Transient Modulator] a lot on acoustic guitars,” revealed Jason Goldstein in SOS April 2007. "In cases where they're just playing rhythm and they are too 'plucky', I can slide that back a bit and take some of the attack off, without using compression.”
Transient shapers also provide the scope to sharpen picking transients in a way that most compressors can only dream of. Jack Joseph Puig made much use of Waves TransX to this effect in his mix of Fergie's 'Big Girls Don't Cry', for instance: "If you were to listen to the acoustic guitar without the plug‑in, it would sound kind of lazy and not urgent. It doesn't sound exciting or like it's really digging in. Instead I wanted to give the guitar an immediate attacking, in‑your‑face sound, as if it's really digging in and played with fingernails, as opposed to the skin of the finger. That's what [Waves TransX] provided. The brain subconsciously analyses where the front of the note is and what the feel of a record is. If you have attack like this, the recording feels exciting and vibrant, like it's moving and is fast. It's a large part of the way the record feels.”
Fret squeaks are probably the trickiest thing to scotch, because they can last a while and also excite a wider range of frequencies, therefore they don't respond as well to the kind of strategies that reduce pick noise. I usually resort to level or EQ automation on a case-by-case basis, as laborious as this can be, but if you have any specialised spectrogram‑driven restoration software (anything like Cedar's Retouch), you may find that a more elegant solution to the problem.
So far, I've been talking pretty much exclusively about how you can use mics to record acoustic guitars, but of course you also have several options for capturing a DI signal — whether via an under-saddle or magnetic pickup, or perhaps a bridge-mounted pickup — and for manipulating the signal with imaging devices such as a Fishman Aura. The topic is a vast one, and worthy of an article in its own right, but it's also worth a quick mention here.
As with everything else, opinion is divided: some producers like to use DI signals, and some don't. "I don't like the sound of DI'd acoustic guitar,” stated Ken Nelson back in SOS October 2000. If you've ever heard one, it's not hard to empathise with him, because the timbre can often seem thin, brittle, and dead compared with the fullness and spaciousness of a good mic signal. However, just because DIs might not necessarily be particularly appealing on their own, tthat doesn't mean there's no mileage in combining mic and DI signals for a bit more presence and focus. Bob Bullock, Frank Filipetti and Stephen Street have all mentioned mixing in a DI, for example. If you do choose to blend mic and DI, the most important thing to remember when doing this is that the DI signal will arrive at the recorder earlier than the mic signals will, which opens up the possibility of detrimental comb filtering, much as when combining mics set up at different distances.
As should be pretty clear by now, there is a huge number of variables when it comes to recording acoustic guitars, from the roles played by the instrument, player and room, to the choices you make regarding types and positions of microphones. If you can get this part of the process right, there's little of the battle left to fight when you reach the mix. Get it wrong, though, and you make your life at mixdown much harder, despite recent advancements in mix processing that can occasionally bail you out.
Given that the acoustic guitar only goes up so loud and incorporates a lot of important high‑frequency information, it's no shock to find that condenser mics are almost always chosen over dynamics for studio recording purposes. Nevertheless, for more aggressive rhythm parts, the higher noise floor of a dynamic close mic becomes less of a drawback, and the inertia of the heavier diaphragm can help take the edge off the instrument's transients, for a more meaty sound. Silvia Massey Shivy, for example, recommends Shure's SM57 for this kind of role: "I like recording a 57 for a very dry, woody sound. The 57 makes the guitar sound a lot less musical and more percussive.”
Small-diaphragm Condensers: When it comes to condensers, the mics that are mentioned twice as often as any others in relation to acoustic guitar are AKG's C451 and C452, small‑diaphragm models, which are sonically almost identical. "My tried and true,” enthuses Mike Clink, but these mics are also name‑checked by other high‑profile engineers such as Bob Bullock, Eddie Kramer, Al Schmitt and Jim Scott. It's probably fair to say that the use of the mic's default CK1 cardioid capsule can be taken as read here, given that no‑one ever seems to mention otherwise — and that a quick glance at its frequency response suggests a couple of good reasons for its popularity. The first is that it's bright‑sounding (around 4dB up at 12kHz), and this is compounded by a gradual sub‑200Hz roll‑off, which also takes the sting out of the inevitable proximity‑effect bass boost when close miking. "The C451 is very bright,” remarks Steve Churchyard, "but I find that it helps the acoustic guitar to cut in the track, because you really want to hear the strings and the pick and that nice presence.”
But one man's meat can be another's poison. "I generally like a warmer microphone rather than a C451,” reveals Chuck Ainlay. "I like to capture the purity of the tone on an acoustic guitar rather than just getting edge from the microphone. I'd prefer to roll out the bottom and add top with EQ rather than using a really bright microphone, where you're not going to ever get the warmth back.” Ainlay's first‑call alternative? Neumann's more understated small‑diaphragm KM84, which finds favour with big names such Daniel Lanois, Ken Nelson, and Alan Parsons too.
Large-diaphragm Condensers: Rather than simply rejecting the C451 on account of its brightness, there are other engineers, such as Eddie Kramer, John Burns and Jim Scott, who simply combine it with a fuller‑sounding large‑diaphragm condenser, a technique that allows the sound to be substantially adjusted at mixdown, provided the two mics are recorded on separate tracks. And large‑diaphragm condensers can, of course, be used on their own too, where slightly softer transients and more 'diffuse' sound are assets — for example, on acoustic ballads or gentler finger‑picking solos.
One thornier issue when using large‑diaphragm mics is that at typical close‑miking distances a lot of the sound source will be off‑axis to the mic diaphragm — which can be more of a challenge to large‑diaphragm capsules, because their off‑axis frequency response tends to be less well‑behaved. This may provide an explanation for the preference amongst some engineers (Chuck Ainlay and Ken Nelson included) for using a large‑diaphragm mic further back from the instrument to supplement a closer small‑diaphragm model, rather than the other way around.
Classic models from AKG and Neumann also dominate my survey of large‑diaphragm condenser choices, with the C414 and U87 solid‑state and vintage C12, U67, and U47 valve models all drawing numerous recommendations. After all, manufacturers like these have spent decades refining the off‑axis response of their large‑diaphragm models to cope with demanding recording tasks such as this. Designers of budget mics frequently skimp on this aspect, in my experience — so in small studios, a small‑diaphragm model is usually the better bet. Another issue with large‑diaphragm mics is that they can be unwieldy in their suspension shockmounts, especially in multi‑mic rigs.
Ribbon Mics: Although ribbon mics are regularly called upon for drums and electric guitars, their use on acoustic guitars seems much more sporadic on real‑world sessions. This is probably because clear, bright guitars are required in most productions to compete with other sounds in the final mix — in which case the warmer sound of a ribbon, nice as it can certainly be, will only find favour with the most strident instruments. To quote Steve Albini, "I'll use a ribbon microphone if it's a real bright guitar and I want to try thickening the sound a little bit.”
Recording a singing guitarist presents its own unique set of problems, because not only do you want to get a good sound for both instruments, you also want to present them in an appropriate balance. The most straightforward solution is to try to catch both sounds with the same mic, positioning it carefully for the best compromise between pleasing tone and sensible balance. For this method to work well, though, it's crucial that you choose a guitar that balances well with the singer in the room — because there's only so far you can adjust the mic position for balance reasons without affecting the vocal or guitar timbres. A mic with a well‑behaved omnidirectional polar pattern can also be a life‑saver in this setup, because it will allow you to get closer to the vocal (for an up-front sound) without detrimentally spotlighting one area of the guitar, or compromising its tonality on account of the mic's off‑axis response.
Elegant though the single‑mic method can be, in reality most engineers put up separate directional mics for guitar and vocal, either because they want to leave themselves some control over the balance at mixdown, or because they want to process the vocal independently of the guitar — perhaps with firmer compression or brighter EQ settings. There are additional difficulties with multi‑miking, though, because each mic will pick up spill from the sound source it's not ostensibly pointing at. This spill not only limits the extent to which you can adjust the balance between the sounds at the mix, but also introduces the possibility of adverse comb filtering affecting either or both sounds. Furthermore, independent processing of either mic signal can quickly impact negatively on the other. So while multi‑miking can help with some problems at the mix, it's not necessarily a much easier option.
There are a few things that can help you get better results if you decide to go the multi‑mic route. The first is to make the recording environment fairly dead‑sounding, and maybe rig some acoustic foam or duvets around the performer to dampen things further. A lot of the spill between the mics will be due to sonic reflections, rather than coming directly from the voice and guitar. Another common fall‑back is to record the guitar's DI signal alongside the mics. You can use this to bolster the guitar level if it's too low in the balance, but without increasing the level of vocal spill. Frank Filipetti put this trick to good use on James Taylor's Hourglass album, if you want to check out a real‑world example.
If you have access to figure‑of‑eight microphones, Mick Glossop suggests using those to increase separation. "If you use a [figure‑of‑eight] mic on a guitar such that the dead sides are pointing up and down, rather than side to side, then you get more rejection of the vocal than if you use a cardioid mic, which doesn't really have as good side rejection.” Likewise, you could put up a figure‑of‑eight mic on the vocal too, with the null plane carefully aligned with the guitar's body to reduce guitar spill. In a fairly dead‑sounding room, the separation using this rig can be astonishingly good. One down side, though, is that figure‑of‑eight mics usually have a stronger proximity effect than cardioids: it's important that your performer doesn't move around too much once you've got a tone you like — because otherwise you'll get fluctuations in the low‑end levels that may be difficult to deal with at mixdown.
Producer Daniel Lanois offers an alternative option to miking up the acoustic guitar itself: sending its DI output to a small guitar amp such as a Fender Deluxe and miking that instead. While this solves practical spill problems when recording singing guitarists, there's more to it than that. "I'm not such a purist that I have to record acoustic guitar with a mic. I actually prefer the amplified acoustic sound in some situations, because it can give you a little more personality, a little more harmonic distortion. When you turn it up to the edge of feedback, you get some nice surprises.”
The idea of re-amping acoustic guitars to add distortion at the mix also has its devotees. For instance, while mixing the Sara Bareilles single 'Love Song', Eric Rosse re-amped the acoustic guitars, mixing them in alongside the undistorted tracks to give them "a little more edge”. Steve Osborne mentions doing something similar as well, while he was working with Suede: "The other rhythmic sound you can hear under the acoustic guitar is the same acoustic track run through the VCS3, which makes the acoustic sound much more chunky and adds a bit of grit. I achieved that by overloading the filter on the input until it distorted. The VCS3 has got a great sound; you can put anything through it and it comes out sounding a bit harder.”
The audio examples for this article were recorded in Studio One at the renowned Livingston Studio complex in London (www.livingstonstudios.co.uk, +44 (0)20 8889 6558), connecting the resident SSL desk's preamps directly to Pro Tools without any processing. Our performer for the occasion was Cambridgeshire‑based professional guitarist Matt Harvey (www.myspace.com/harveyguitar, +44 (0)7884 010671), who played three different guitars for each of the following three main sets of examples:
- A mic shootout including the following mics: Neumann U47, U67, and U87; AKG C12, C414B ULS, and C451; Coles 4038 ribbon mic; Shure SM57; SE Electronics SE4. The vintage U47, U67, and C12 were kindly loaned to us for the session by FX Rentals (www.fxgroup.net, +44 (0)20 8746 2121) — if you fancy renting any of these mics for yourself, then the company's 24‑hour phone line is open all year round, and they can even offer single‑day rates if that's all the time you need.
- A set of examples demonstrating the effects of different mic placements and polar patterns on the quality of the captured sound, using a dozen identical SE Electronics SE4 condenser microphones kindly provided for the session by the company's UK distributor Sonic Distribution (www.sonic‑distribution.com, +44 (0)845 500 2500). These examples were also repeated to show the effects of altering the acoustic properties of the recording space.
- A set of examples illustrating different mic positions, again using the identical SE Electronics SE4 condenser mics.
These highly detailed audio examples, along with mixing demonstrations, can be found on the SOS web site at www.soundonsound.com/techniques/recording-acoustic-guitar-audio. You can also download the files and import them into your DAW to see how different mic models and placements can be combined at mixdown. If you're serious about recording acoustic guitar, you really should make the effort to listen and experiment with these.
I've mentioned a lot of producers' names in this article, and if you're wondering where you can find out more about any them — what records they've done and what their views are on other recording topics — it's easy to get hold of the original articles I've quoted. They're in one of two places: either in the huge SOS online article archive (just punch a name into the Quick Search box at www.soundonsound.com/search) or in one of Howard Massey's two excellent books of interviews, Behind The Glass Volume I (ISBN 0879306149) and Volume II (ISBN 0879309558).