It's easy enough to make a guitar and amp sound loud and gutsy in the live room, but how do you capture the same sense of volume and distortion on your recording? Phase alignment, that's how...
The quest for ever-bigger distorted guitar sounds is a common source of headaches in the studio, and not just because it can involve you having to stick your head directly in front of a blaring 4x12 cab! Putting a Shure SM57 or another trusty dynamic mic in front of a speaker and pressing 'record' is certainly a tried and tested approach — and sometimes it's just what the doctor ordered. But often it can deliver a recorded sound that suffers from a lack of 'thickness', for want of a better word. So how do you get around this? In a mix where the guitars really need to dominate, making the drums and vocals fight to be heard over a wall of distortion, a phase‑aligned multiple mic technique can be your friend.
What we're talking about here is setting up a speaker cabinet, putting a different mic on two, three, or even four of the speaker cones in that cabinet, and then blending those signals together to create a more saturated sound that takes advantage of the different characteristics provided by each of the mics.
It all sounds simple enough so far — it's just a case of layering sounds — but why do we need to phase‑align the mics? Well, what we're looking for is the most direct, solid and 'ballsy' sound that we can achieve. The best way to do this is to ensure that every peak and trough in those recorded signals is perfectly aligned. In other words, we want each and every diaphragm to be exactly the same distance from its respective speaker cone. If you don't manage this, you'll get some filtering creeping in, which will manifest itself as 'phasiness': a slightly hollow sound, a lack of bottom end, or perhaps a nasal quality. I'm not the best person to write an essay about the science of phase (see Mike Senior's article in SOS April 2008 if you want to know more about the theory), but I can offer a simple analogy... Imagine a rowing four: you want every oar going into the water at exactly the same time on every stroke, to move the boat forward as fast as it can go. If one oarsman is a bit out of sync with the others, it will upset the balance of the boat and reduce the speed.
Before I move on, a quick word of clarification: engineers tend to use the word 'phase' rather loosely. If you hear them referring to 'flipping phase', they're talking about reversing polarity on mic preamps — so that's something to watch out for.
With the concept of phase alignment in the bag, let's plan what equipment we're going to need to do this sort of recording.
Obviously, we need an amp, and the sort of sounds that we're talking about tend to come from high‑gain heads, coupled with big, heavy, closed‑back cabs — often 4x12s. The phase‑aligning technique can be applied to smaller cabs and combos, but if we want to use more mics than there are speaker cones, we need to double up on one of the cones. That's not necessarily a problem, but it can sometimes be impractical, so for the purposes of this article we'll be using one mic per cone on a 4x12. Life is far happier when the cab is in another room. Isolate it if you possibly can and, wherever you put it, try to get the cab up off the floor slightly, which can help to avoid problems with reflections off the floor — we're looking for close‑miked sounds here. It's also a very good idea to keep the amp head with you in the control room, and run a long speaker cable to the cab: that way, you can tweak the controls of the head while monitoring in the control room in the context of the track, which is clearly an advantage. Furthermore, for reasons that I'll explain later, we don't really want people going near the cab once we're miked up, which they would have to do if the head was sitting on top of it. One note of caution here, though... you should avoid sending amp signals to cabs via a patchbay, because the amp needs to 'see' a load all the time.
We're also going to be using a 're-amp' device, which is a very useful thing to have around. The function is simple: it allows you to output signals from your DAW at a level that a guitar amp expects to see. Not only is this useful for running white noise out to the amp (I'll come on to this), but it will also enable us later to layer multiple sounds from a single DI signal. The 're-amp' also allows us to drastically reduce the overall time spent with the amp up very loud, as we can track at low level and then re-amp at the optimum setting for the amp and speaker. There are various commercial products available, including the original Reamp and boxes from Radial and Little Labs, amongst others.
Next, let's think about mic preamps. Almost all of these will have polarity‑reverse switches on them, but if yours doesn't, you can easily use a DAW plug‑in. Some people suggest using polarity‑reversed cables or adaptors where such a switch isn't present, but I wouldn't advise that you rely on these, simply because you need to be able to 'flip phase' easily. In terms of what mic preamps you should use, there's plenty of choice. If they're available, the typical Neve devices are always great, and I personally love APIs on guitars. That said, I wouldn't lose too much sleep about it: I've made records using far less expensive options with no problems at all, so just use whatever you have to hand. In fact, a good amp and setting the correct gain structure will make a far greater difference to the final result.
This recording technique benefits from the differences between our microphones, and we want to look for mics that complement each other well. One might be brighter, one bassier, one smoother, and so on. The tried-and-tested dynamics and condensers are certainly worth a go. I regularly use Shure SM57s, AKG C414s, Electrovoice RE20s and Sennheiser MD441s, but there are plenty of other models used by different engineers (for a detailed discussion of this, have a read of Mike Senior's article on guitar-amp recording in SOS August 2007).
Of course, we're talking about recording loud sounds here, so you're going to need mics that can handle a good bit of level (I definitely wouldn't recommend that you stick your delicate vintage ribbon mic a millimetre from the grille and hope for the best!). When using mics close like this, you'll most likely find that if you use a condenser mic, you'll need to switch in a pad so that it can cope with the level. Even then, some will find it harder than others; I'll talk more about this later.
Now I need to have a quick word about mic stands. This technique requires very precise positioning of microphones: if your stands slip an inch or so over a couple of hours, or even overnight, all your hard work will come undone. Equally, you need to be able to easily loosen a stand to sweep a mic around. So the long and the short of it is that it's well worth making sure you have good‑quality, well‑maintained stands.
Obviously we need some sort of recording medium, but the only requirement here is that you have the ability to meter the sum of your mics. I strongly advise that, before having a crack at the recording itself, you check your entire signal path to make sure that nothing is reversing the polarity of the signal at any point. Some mics, speakers, outboard gear and cables are wired 'backwards' — and if that's the case, you need to know, or the whole process will become rather confusing. The easiest thing to do is to get a mic up and record a drum hit sent out through the amp. Record each speaker in turn and have a look at the resulting waveforms to make sure that the initial transients are all going the same way. If some are going up and some down, then you have speakers in the cab that are wired with different polarities. Assuming they're all OK, just use one speaker and swap out the mic cables and then the mic amps to make sure they're all the same. Then try each of the mics you intend to use. The last thing you want is to be unable to work out why you can't get the mics to play nicely at the same distances from the cab because one of the signals is coming in inverted! If you find that everything is going the same way, but that it's opposite to the original hit, it's likely to be the amp that's flipping it. Don't worry about that though, because the important thing is that they're all the same.
The final thing we're going to need is a guitar signal. You can use a guitarist, or you might have a pre‑recorded DI guitar part — which you can use, provided it's nice and clean. The guitar itself will have a hugely significant bearing on the sound, and generally speaking it'll be the Les Pauls and other guitars with humbuckers that give the big, chunky sounds. But having said that I often use Telecasters — so it's horses for courses, as usual.
So here we go! Start off with one microphone, and think of this one as your 'core' mic. In my experience, this one is probably going to be a dynamic mic, and it should be the one that you feel has the most balanced individual sound. A Shure SM57 should provide a decent starting point. Get that mic up on one speaker cone, and sounding good on its own. You can run a pre‑recorded DI signal out to the amp via your re-amp box, or just have the guitarist play. It doesn't matter which, but the DI is often better because it won't get bored and start soloing when you need to hear chunky rhythm parts...
Now send the relevant mic-amp signal to a record‑enabled channel in your DAW. Route that channel to a headphone mix that you can monitor when you're in the live room. If you have your other mics connected at this point, make sure that they're not feeding the headphones.
Now get into the live room and put on your headphones. You want to get that mic close to the grille, but leave a couple of inches of space between the mic and the grille, for reasons that I'll explain below. Loosen the stand and sweep the mic around in front of the speaker. In any cab, you'll most likely find a speaker or two that you prefer to the others, so pick one of those. Try angling the mic and moving it around the speaker cone, left and right, up and down, until you find a 'big' sound with plenty of mid‑range bite and solid low end. Then tighten that stand — and don't touch it again!
Now a quick word about amplifier level. Often, the amp itself will dictate level by having a sweet spot. I generally find that you need enough level to get the speaker cone really working, but that if you crank things up too much, your condensers won't handle it well. Typically, it's a sort of loud, gigging level — one that's uncomfortable to stand next to in the room, but not registering too high on the Richter scale. If you're hearing any condensers starting to crackle, back the level off a bit.
At this point it's time to stop using the guitar sound for a moment, and move over to white noise. Set up a signal generator in your DAW and use it to send the noise to the re-amp. Another word of warning, though: white noise is used to torture people; it makes your brain hurt. So get the cans on your head before you go into the live room with the amp, because if you expose yourself to high‑level white noise for any length of time you won't be mixing anything that day!
Now get your second mic up on another speaker cone, and position it so that it looks like it's in the same place (relative to the cone) as the first mic. Then go back into the control room and create a DAW channel for the second mic. Stick that one into record along with the first and adjust the mic-amp gains so that the two levels are as closely matched as possible. I usually get the signal coming in at about ‑15 to ‑10 dBFS or so. Send that channel to the headphone mix, so that you have two channels of white noise at the same level going to the cans.
If this all sounds a little mysterious so far, let me spell out what we're doing. The idea is to go back into the live room and sweep the second mic around until you find the point at which the signals are perfectly aligned with each other; in other words, the position at which the sound waves travelling from 'speaker 2' to 'mic 2' are exactly aligned with those travelling from 'speaker 1' to 'mic 1'. So make sure that both mic amps have their polarity switches set the same way, and then go back in and sweep the mic — and you'll hear all sorts of phasey, sweeping noises, as the mic is moved around. You're looking for the place where the noise is really roaring. Pay particular attention to the bottom end and the mid-range. And don't worry if one mic looks a bit closer to the grille than the other, because their diaphragms may be in different places relative to the body of the mic. This is the reason why we left a couple of inches of space between the first mic and the grille: we don't want, after the first mic has been nicely positioned, to find that we can't get the second mic close enough. Similarly, you might find that where the speakers are close to the edge of the cab, the body of the mic can be obstructed by the cabinet itself if the microphone is placed right on the grille.
Although I've said you're looking for the point where the noise is roaring, I actually find that it's easier to listen for cancellation than reinforcement — so when I do this I flip phase on the second mic amp and listen for the deepest null that I can find, trying tiny movements to see what makes it better or worse, but you'll have to see which approach you prefer. Hopefully it's now clear why you need to know that your signal path is free of reversed cables. It's very confusing when you can't find a null where there should be one!
Remember, though, that we're not using identical mics, and when you're looking for reinforcements and cancellations, sometimes two very different mics won't cancel each other as much as you might hope. For example, a really bass‑light mic will not cancel a lot of low frequency out of a condenser.
Get yourself back in the control room and tweak the mic amps to match the gain, in case the new position has changed the relative levels a little. Set up a meter in your DAW or on your mixer that shows the summed level of your two channels. The simplest way would just be to route both to an aux or master fader. Now flip the polarity back and forth on one of the mic amps: what you're looking for is a big drop in level on that meter. Depending on the combination of mics you're using, you might get 10 to 14dB of cancellation. See how it sounds in the monitors. You should go from a roar to a thin, hollow, hissy sound. If you're not happy, go back and try again with the mics — and don't worry if you don't hit it first time, as it can take a bit of practice. If you want to try for a third, or even a fourth mic, you can mute the second and align a third with the first in exactly the same way — but remember that, as a general rule, the more mics you add, the harder it is to maintain focus in the sound. Typically, you'll find that the main body of the sound is made up of two mics, with perhaps a bit of a third. It's not easy to get four to play nicely, but that doesn't mean you shouldn't put them up to try. It might be that you use two dynamics and two condensers, and for different guitar parts you combine any two of the mics.
It should become obvious pretty quickly that small movements of the mics make a huge difference to the sound. This is why we can't use stands that are likely to sag. It also brings us back to the reason we don't want to go near the cab once it's been miked up. We only need to knock one stand a tiny bit and we're back to the horrors of white noise. It's worth impressing upon everyone present that they must tell you if they accidentally knock a stand.
So you now have your mics up on your cab and phase‑aligned. You have amp heads in the control room that you can switch between at will. You can have your guitarist playing, or you can run DIs out through the amp. You can adjust your mic‑amp gains while the part is playing back, to give the optimum gain structure, and/or the blend you desire sonically. You're also able to re-balance the levels of the different mics later on by just playing with the tracks in your DAW.
Alternatively, if you have a mixer you might choose to blend your signals at tracking, perhaps applying some further outboard before printing them to a single track in the DAW. This has the advantage of saving space — not to mention making a simpler session file — but it obviously reduces flexibility later on, too. My advice is to keep your options open at first. You might decide that certain parts are too thick in the mix and revert to just one of your mics (don't be scared to take decisions like this during the mix, just because you captured the 'perfect' sound in isolation), or a different blend. And if things really aren't working, there's still no need to panic — because if you've captured DIs as well as the mic signals, you can just re-amp again whenever you need to.
Some examples of this technique are on the SOS web site at /sos/jan09/articles/loudelectricguitaraudio.htm. Steve Flash is playing a Gibson Les Paul into a 50W Soldano Head, and an Orange 4x12 cab with Celestion Vintage 30 speakers. There are four mics on the cab: a Sennheiser MD441, a Beyer MC740, a Beyer M160 and an Electrovoice RE20. Mics one and four are going to API 512C preamps, while mics two and three are going to a Neve 1073 DPA. The signals are recorded to Pro Tools HD via the A-D converters of a Cranesong Spider, connected digitally to the AES input of a Lynx Aurora 16 audio interface with HD card. I used the ubiquitous K&M boom stands and some SE Electronic tripod stands, as well as two Enhanced Audio M600 mounts.
I've included two examples of sweeping a mic with white noise: one illustrates reinforcement, and the other cancellation. There are samples of each of the individual microphones at unity gain, and a double‑tracked blend of all four, plus a three‑mic blend using the MC740, RE20 and MD441. You may find it useful to import the four individual files into your own system and experiment with the balance. In practice, using all four mic signals at the same time would be quite unusual.
Jack Ruston is a recording engineer, producer and guitarist. He works primarily with veteran producer Nick Tauber, making records for rock bands.