Is there more to life than an SM57 and an amp?
By Bob Ross
So the Editor calls me up and asks if I want to contribute something to the guitar issue:
“Well, how about an article describing my guitar recording techniques?” I sez.
“Do you use any tricks besides the old SM57-in-front-of-an-amp?” Nick asks.
“Oh yeah, sure, you betcha,” I sez.
I lied: stick the damn SM57 in front of the amp and be done with it. Seriously. It’s a tried and true method that’s been used on every type of guitar recording imaginable, with results ranging from perfectly serviceable to absolutely perfect. What with the almost infinite variables (distance from SM57 to speaker cone, angle of SM57 to speaker cone, area of cone targeted by said SM57, etc.) it’s a wonder engineers ever bothered to come up with alternatives.
But you didn’t need me to tell you this; every textbook chapter and every magazine article ever written on the topic of recording electric guitar starts with the old SM57-in-front-of-an-amp. And then proceeds to multiple SM57s on an amp, multiple dissimilar mics on an amp, multiple amps, etc., etc., etc. (Check out Dave Martin’s primer in the Feburary 1997 issue for a couple of excellent examples.)
Guitar recording technique for commercial contemporary music has become almost canonical, due to both the success of these techniques in capturing desirable guitar sounds and the tendency of earlier successes to influence what is considered desirable by subsequent generations. (In other words, if Frank Zappa’s “Inca Roads” had been the mega-hit instead of, say, Deep Purple’s “Smoke on the Water,” we’d all be tipping Pignose amps on their backs and miking them from the ceiling…though I seem to recall that was done with an SM57 anyway.)
So what follows are some recording techniques you won’t read about elsewhere, that I’ve used when the appropriate guitar sound was not a stereotypical, commercially-acceptable norm, when the musicians were looking for something special, something unique, something little yellow different better. The sounds that result from these methods won’t (usually) resemble the average radio-friendly guitar tone, but they may be just what your tune needs to stand out on the air.
This article is far from being any sort of definitive treatise on recording eclectic electrics (as if any such treatise could exist!); the idea is not for readers to emulate the specific techniques I describe so much as to explore the curious experimental open-minded aesthetic that led to these techniques, and conventions be damned! If just one reader goes into his or her next recording session without a preconceived notion of what an electric guitar ought to sound like, I’ll feel I’ve done my part.
The aforementioned Frank Zappa/Pignose technique is a good place to start, both because it doesn’t sound that radically different from the rock guitar status quo and because it doesn’t require 120 dB SPL during your tracking session. Yes, hard to believe, but most of Zappa’s scorching, gargantuan, distorto-scream guitar solos from his early 1970’s albums were recorded with a Pignose amp, a tiny 15-watt battery operated portable with a 6″ speaker.
The trick is, the amp was laid flat on its back in the middle of the floor and miked from some distance overhead. Putting the speaker in close proximity to a large boundary (in this case, the floor) radically enhances the low frequency output; if you’ll recall your studio monitor theory, it’s called a half-space orientation. For a studio monitor, that woofy low end is undesirable; for a guitar amp, especially a dinky one with a 6″ speaker, that can be awesome, yielding a thick fat oomph that would otherwise require a 4×12 cabinet with 100 watts of British tube amp pumping into it.
This can be done with any small amp; I’ve tried it with Fender Champs and Dean Markley mini combos, both with excellent results. It also works with larger speaker cabinets, and you’d be surprised at the difference in sound between a Fender Twin in its upright position versus one lying flat on its back. I usually just park the mic de jour on a boom stand at maximum height, pointing straight down over the speaker.
In fact, tiny combo amps are an asset to any studio’s arsenal, and not just because they won’t (necessarily) wake the neighbors. Tiny speakers fit in places that large speakers won’t. Like closets. Like ventilation ducts. Like garbage cans. Like inside the piano.
For all the distant room miking we regularly use in our studios, it’s still the same room… but any acoustic space can become your “room” if it’s big enough to squeeze a small speaker and a microphone in there. I’ve recorded electric guitar with the amp inside the drawer of a metal file cabinet (miked with a 57, natch); the hard resonant surfaces of the drawer yielded a grungy “spraang” somewhere between reverb and ring modulation.
Make your speakers smaller still, and you can start putting them in really bizarre places (no gastroenterology jokes, please). I’ve duct taped Walkman headphones directly to the windscreen of a Neumann U-87 and fed the guitar track into the ’phones from an aux send. The sound of the minuscule speaker cranked almost to the point of frying combined with the proximity effect of the mic yielded a guitar sound that was thick enough to fill out the track without obscuring the bass or the piano. Curiously crispy, sort of faux-compressed. And no eq.
Consider the source
Once I started experimenting with cramming small speakers in all sorts of odd locations, it wasn’t long before they got all the way back to the source: on the guitar itself.
Take an attenuated speaker output from your guitar amp (I’ve found load boxes such as the Scholtz PowerSoak or the Marshall PowerBrake to be indispensable when recording guitars) and connect it to any reasonably heavy-duty full range driver, 5″ or smaller. Now affix that speaker directly to the body of your electric guitar! Depending on the size of the driver, this could be on the lower bout, or down below the tailpiece, or even up on the headstock.
Obviously, if you’re playing a valuable pristine ax, you’ll want to put some careful thought into exactly how you attach this hardware. I’ve used duct tape, C-clamps, and reusable weather stripping, all with adequate results; admittedly, though, the best results I got were when I decided “to hell with cosmetics” and used wood screws to bolt the speaker onto the body of a cheap Japanese Les Paul copy.
The point of this exercise is that by re-energizing the guitar body with its own amplified output, you’ve effectively harnessed infinite sustain and controllable feedback. (Astute readers may recognize this as the theory behind products such as the Sustainiac, but I must tip my hat to Steve Holland and Seth Gussow for introducing me to this technique a decade before those commercial devices were available.) If you have a spare amp head that you can dedicate to this feedback circuit, you can also experiment with an equalizer in the signal path; this offers more control over which frequencies will feed back, allowing you to even out the response over the range of the guitar.
Speaking of bolting drivers to musical instruments, the Ondes Martenot, an early electronic musical instrument that predated conventional synthesizers by nearly 50 years, came with a variety of speaker cabinets that were each designed to shape the spectral content of the output in specific and distinctive ways. One such cabinet consisted of a compression driver bolted to the center of a genuine Turkish gong. Cobble one of those together for your next guitar tracking session and you’ll be sure to wind up with either a Grammy for Engineer Of The Year or a one-way trip to Belleview.
And if you’ve got the compression driver but can’t spare the gong, pick up some rubber hose or surgical tubing at your local hardware store, cram one end over the compression driver’s throat, stick the other end down your throat, and voila: you’ve got a talk box! (Remember Peter Frampton’s “Do You Feel Like We Do?” Uh, how about Bon Jovi’s “Living on a Prayer”? Geez, I’m getting old.)
Again, you’ll want to use an attenuated output from your amp, or preferably an independent dedicated amp, to drive the talk box. Also, most compression drivers are rather unforgiving about passing frequencies below their recommended cutoff point, so a high-pass filter or active crossover is essential.
I feel it’s important to recognize that a talk box doesn’t just have to be used to make your guitar talk. (Besides, Frampton discovered that “do you feel like we do?” may well be the only English sentence that is completely intelligible through these devices.) I prefer to think of the talk box as a speaker cabinet with Dynamically Variable Multimodal Filtration. It’s just that instead of a Data Entry Knob, you’ve got your mouth…and the Unpredictability parameter can never have a value of 0%. The number of sonic elements that change while you make a subtle alteration to the shape of your oral cavity is staggering, and can’t be easily reproduced using conventional guitar processing. Besides, feeling your head literally crammed full of your own playing is an ego rush no guitarist can resist.
I’ve also recorded guitars through a talk box without the attendant human mouth feature by close-miking the end of the hose (is that what guitarists mean by “tube sound”?) and again, it’s a tone you definitely don’t hear every day. Kinda distant, kinda tinny, but for certain tunes, kinda perfect.
Well, messing around with all those funny speakers and amps is fun, and there’s a whole world of unique and usable guitar sounds available to anyone with the patience to listen to the myriad possibilities, but it’s an inevitable fact that many of our project studios must limit guitar tracking to the direct injection method. Does this mean that DI guitarists can’t access those hip crazy sounds that the mic & amp (and hose & gong) guys do? Hardly!
If you’ll recall the traditional literature/conventional wisdom on the subject, recording electric guitars direct through the board should be done through a direct box, which is a matching transformer that couples the guitar’s high impedance output to the console’s low impedance mic input. Bypassing the direct box and plugging the guitar directly into a line level input on the console usually results in a level and impedance mismatch; the sound is thin, small, two-dimensional, and generally lacking in balls.
So naturally, my first recommendation is to bypass the direct box and plug the guitar directly into a line level input on the console. I am not kidding. Well, let me qualify that: don’t do this for every tune; just the tunes where it would be musically appropriate to have thin, small, two-dimensional guitars. You’d be surprised at how many muddy mixes are the result of not leaving enough space for all the instruments to coexist. (Or maybe you wouldn’t be surprised; that statement’s practically tautological.)
Intentionally recording certain instruments to occupy a restricted part of the mix is one strategy to overcome this flaw, and I’ve found that electric guitar survives this deliberate fidelity-constraining with fewer detrimental side effects than, say, bass or drums. Be sure to listen to your sounds in context before giving them a thumbs up or down.
Reproducing plausible, convincing amp distortion when recording direct guitars is probably the bane of every engineer. The aforementioned reactive load power attenuators have proven invaluable for me. I’m also a big fan of re-amplifying direct guitar tracks and recording that amp with a mic. This can be done at a later overdub session or during the mix, in either case when the neighbors are out of town, or if you’re using a studio that has the luxury of cranking guitar amps. Conventional wisdom once again applies: if you’re looking for “accuracy,” or at least to preserve the sonic integrity of what’s already on tape, use a level/impedance matching device such as the AXI Reamp to go from the multitrack into a guitar amp. But don’t forget to investigate what a level/impedance mis-match can do for your sound. (Hint: it’ll be a lot noisier.)
Real tweak-o-philes (and I use that term with the utmost respect and endearment) will tell you that the heart of guitar amp distortion doesn’t come from the preamp tubes (which is why a lot of the hybrid tube preamp/digital fx processors on the market sound a bit lifeless). It also doesn’t come from the power tubes, they’ll tell ya. It’s the saturated output transformer that lends that unmistakable meaty quality to an overdriven tube amp.
Interestingly, one now-defunct company tried to capitalize on this concept: back in the dark ages of rack systems for guitarists, Intersound manufactured a solid state preamp that used a miniature output transformer in its Tube Voicing circuit.
Now, I’m not mentioning this because of how authentic their distortion is; in fact, it sounds completely unlike any respectable tube amp cranked to eleven. I’m mentioning this only because I’ve seen literally dozens of these preamps (it’s called the Intersound IVP) flood the used equipment market in the last year, and they’re going for peanuts! $99 for a rackmount preamp with 6-band semiparametric eq, two effect loops, balanced direct out, and this ridiculously idiosyncratic saturated-transformer Tube Voicing circuit…who cares if it doesn’t sound like a Marshall stack? Plus, the cool thing about the IVP’s distortion is that it’s post-eq, so you can shape the harmonic content of the overdrive in a limitless way. Cool box. I’ve got two of ’em.
Stepping down hard
That sort of leads me to the topic of vintage gear, especially stomp boxes. I love cheesy old guitar pedals; you’ll never catch me trying to talk a guitarist out of using his favorite Univox Super Fuzz just because our studio offers a selection of top-of-the-line Boogie, Matchless, and Dumble tube amps. (It doesn’t, I’m just talking hypothetically here.)
The problem with a lot of vintage gear is, it’s noisy. And while noise has its place in a lot of contemporary music, we’d like that noise to be an aesthetic choice, not an unfortunate inevitability. So when I’m tracking or mixing electric guitars with vintage effects, I keep a single-ended noise reduction unit close at hand. My current device of choice is the Behringer SN1000 Studio DeNoiser; I’ve also used the Symetrix 511, and found it equally versatile.
These devices combine  a gentle downward expander, which reduces the gain whenever the signal drops below a user defined threshold, with  a dynamically-sliding low-pass filter, which reduces the high frequency bandwidth in the absence of high frequency information. For salvaging noisy tracks, these things are a godsend; I’d recommend one to any studio, before you even think about your third ’verb or more compressors or gates…
Of course, if you’ve already got a decent gate, here’s a technique I like to use: patch your rhythm guitar track into the gate. Patch one of your drum tracks into the key input of the gate. Depending on the style of music (and how pronounced you want the effect), set up the gate for either subtle expansion or radical on/off gating. By rhythmically modulating the guitar track, you can tighten up a rhythm section, impart a sense of “breathing” to the feel, or simulate tempo-locked tremolo. Too tame? Okay, unplug the drums from the key input and replace them with
A. the reverb return from the lead vocal
B. the unused guitar solo from an earlier take
C. an AM radio tuned to Rush Limbaugh’s show.
You get the idea: keying the gate from a less-than-predictable source imparts surprise, mystery, and occasional serendipitous synchronicity that rivals pure magic. (I’ve also had great success using this technique on keyboard pads and background vocals instead of guitar.)
Yes we’re normal
I’m running out of space, so let me throw a few more quick ideas at you:
Ultra short delays, between 1 and 10 ms (and, if your delay can do it, between 0.1 and 1 ms) with high feedback values can create comb filtering effects that range from radically bizarre to… surprisingly normal? Huh? It’s not that dissimilar from the comb filtering that occurs when you put several mics at varying distances on a speaker cabinet. Try this instead of reaching for an equalizer next time you want to alter the frequency spectrum of a guitar track.
Use auto-panners not to zip back and forth between left and right speakers, but to zip back and forth between two completely independent signal processing chains. Send your guitar to the auto-panner input via an aux send. Bring the left and right outputs of the auto-panner into two channels of your console. Don’t assign those channels to the mix, but use other aux sends or bus assigns on those channel strips to feed the left output into, say, long delays with flanging and plate reverb, while the right output feeds, oh, how about ring modulated saturated-transformer Tube Voicing through a talk box… into a compression driver bolted to a gong?
As the auto-panner sweeps, your guitar evolves from one sound to another (potentially dissimilar) sound. Let’s see your Audio Morphing effects processor do that!
On the same theme: consider a Leslie rotating speaker cabinet as a mechanical auto-panner. Stereo mic the upper rotor, then proceed as above.
Here’s a rather brazen technique that affects both timbre and performance: when comping guitar solos, don’t put separate takes on separate tracks. Record them all to the same track. Don’t rehearse where or when you’ll punch in and out. Don’t let the guitarist hear the previous track. Now go for it. As you punch in and out randomly, you’ll be erasing chunks of earlier passes with new material while missing out on some possibly beautiful playing during the current pass.
The end result is a single guitar solo that never actually happened. A lot of the results couldn’t possibly happen: as you punch in and out during sustained notes, you may create melodic leaps and phrases that are physically impossible. Moreover, the ramping or crossfading of the record electronics shapes the attack at the punch-in point in a deliciously sinuous way.
The final composite sounds somewhere between two-hand tapping and backwards taped guitar, depending on what sort of multitrack you use. (I learned this trick on an Ampex ATR-104, which had a luciously slow ramp time, but I’ve had musically intoxicating results with ADATs and PortaStudios as well.) If you want to take advantage of this timbral effect without the aleatoric aspect of improvised punches, try recording slow sustained guitar melodies one note at a time, punching in each note after the guitarist has already struck the string.
So I hope you find this dissertation intriguing, if not downright informative. Remember, there are no laws against experimenting in the studio, and electric guitar is one of the most rewarding instruments to experiment on…just so long as you keep an SM57 close at hand.