Most engineers can cope with the normal run of instruments: guitars (amplified and not), drums, pianos, etc. They’re not easy and whole articles are written about how to record them, but the general techniques are widely known.
The scene is changing, however. As “world music” becomes more popular, exotic ethnic instruments are sneaking into bands. Bodhrans, kalimbas, and sarangis sit side by side with Telecasters and Les Pauls, not to mention what a recent review called the “inevitable didjeridoo.” And some established instruments such as synths and winds still present problems to recordists.
So I’ll go through a selection of unusual instruments I’ve worked with, ethnic and otherwise, and tell you what I’ve tried and what’s worked for me.
Obviously, your mileage may vary. You should know my prejudices before we start: I like naturalistic representations of sound rather than extreme manipulation. And I look for “body,” that elusive quality in a recording—I want the instruments to feel solid and grounded in space rather than floating in a disembodied non-space. (Am I getting too metaphysical? Sorry; I’ve been reading Harry Smith’s liner notes.)
On to the instruments, in alphabetical order.
A gentleman is someone who can play the accordion but doesn’t.
Hrrm. Excuse me. The thing that makes the accordion difficult to record is that it’s really two instruments connected by a bellows. The melody side (on the player’s right) is fairly easy, as it’s stationary; I often use a Shure SM-81. The sound is loud, so I usually use the 10 dB pad.
The chord side (on the player’s left) is tougher because it’s moving in and out as the player pumps the bellows. (Especially on button accordions, popular with Irish and Cajun musicians.) Aiming a mic inwards at the moving end of the accordion won’t work, as the volume will fluctuate radically. (You’ll also get a distinct Doppler effect on abrupt notes.)
One solution is to record the bass side with the mic facing the player, using a mic with a broad pick-up pattern (perhaps even an omni). Set the mic so that the bass section is directly in front of it when the bellows is extended halfway (assuming the player uses the bellows’ full travel). An SM81 will work well here, as will a Neumann U87 if you have one. Keep the mic a foot or so away from the instrument.
Another solution is to mount a miniature lavalier mic on the accordion. You’ll need to experiment until you find the spot that produces a balanced sound with all chords, and be sure to leave the player a lot of slack in the cable. Various Audio-Technica mics will work well in this application; I’ve also had success with the old Radio Shack mini-mic, but alas, it’s been replaced by a new model that I haven’t tried yet.
Nothing exotic about this instrument, but it’s surprising how few modern recordings get it right. What gets lost is the delicate combination of richness, woodiness, and power that makes the bass viol such a delight in skilled hands.
Too often a bass is recorded as nothing more than dull rumble, without definition, so it’s impossible to tell what note is being played. Occasionally I hear recordings with the opposite problem: the highs are boosted so much that the instrument sounds like it’s made of cardboard.
Surprisingly, truly flat microphones often produce disappointing results on the bass fiddle; a peak in the high frequencies, paradoxically, helps the instrument sound more natural while letting it cut through the mix in ensemble recordings. Many players and recordists swear that the best combination of body and brightness comes from a Neumann U47 microphone, preferably a tube model.
Great, but most of us mortals can’t afford one of these. I’ve discovered a good alternative: the little, moderately priced Electro-Voice RE200 condenser mic. It has a couple of sharp peaks in the high frequencies and a roll-off at the bottom that will compensate for proximity effect when it’s used close-in.
Where to place it? In my experience, most bass fiddles have a sweet spot on the right shoulder of the instrument (as you face it) about 18-24" above the f-hole. This gives a lovely balance between woodiness and richness, and it’s well out of the player’s way (especially useful if s/he’s using a bow).
If the sound is excessively thin, it may help to supplement the bottom with a second mic placed over the f-hole (an Electro-Voice RE15 or an SM81 with the bass set for gentle roll-off is good here). Placing the mic over the bridge gives a woodier tone, but it’s less convenient.
Many players wrap a mic in a towel and stuff it below the tailpiece or under the bridge. I’ve never heard good sound from this set-up; it’s muddy and boomy, with bad pitch differentiation.
Pick-ups are a tough issue. They’ve improved drastically in the last ten years, but they still lack the air and space of a miked recording. If you use one, supplement it with a mic, preferably on a separate track so you can combine the two at leisure.
In live performance many performers run their pick-ups into a bass amp. My reaction to this is a grimace of pain; it always sounds to me like a cheap pawnshop electric bass rather than a rich, lovely bass viol. I suspect the problem is the amps: their tonal characteristics are made to match electric bass guitars, not double basses.
This is an Irish open-backed drum, about 24" in diameter, played with a beater that looks remarkably like a small dumbbell. It’s hard to record in an ensemble without losing the beat because the powerful BOOM tends to obscure the skin sound that delineates the rhythm.
For years I used the Electro-Voice RE15 on bodhrans, but recently had the occasion to try the Electro-Voice RE200. Jackpot: the RE200’s combination of low distortion, high SPL capability, and bass roll-off prevent its being overwhelmed and “whoomped” out, while the high-end peaks bring out the skin sound beautifully. (I suspect a U87 would do well, too; use the bass roll-off switch.)
You may want to use some compression, as the RE200 gives none of the natural compression of a dynamic mic. The RE15, RE16, and RE20 also do a good job.
You might expect these instruments to behave like accordions, but they don’t. In fact, according to my friends who play them, they hardly behave at all.
In a concertina, both ends move, while the melody may jump from one end to the other. My best results have come when I miked the instrument from the front with a pair of microphones, usually Neumann KM84s or Oktava MK-012s. Ideally this will be a crossed pair a couple of feet away, but this becomes difficult if the player also sings or there’s leakage from other instruments.
The world’s leading concertina virtuoso, Alistair Anderson, favors a pair of mics spaced about 9" apart, both pointing forward. If you record the instrument this way, panning the mics hard right and left, you’ll get some spectacular stereo motion effects as the slightest movement on the player’s part makes the sound leap across the room with Dramamine suddenness. For the sake of sanity it’s probably best to pan the mics less dramatically.
This placement violates the “Rule of Threes”—that the spacing between microphones should be greater than three times the distance from mics to sound sources—but in practice the expected phase problems don’t seem acute. I’ve also tried using a single condenser mic (usually an SM81 or U87) located about 10" in front of the center of the instrument; enough sound leaks around the side to provide an acceptable recording, especially if the concertina is being used for accompaniment.
Why has this instrument suddenly become ubiquitous? Its sound is certainly exotic and intriguing, but what possesses musicians to add it to Irish bands, Cajun music, industrial and house music, contra-dance ensembles, and (for all I know) string quartets? Perhaps Australian aborigines live in one of the few Third World cultures that’s still remote enough for us to romanticize. Or maybe I’m just a curmudgeon.
In any case, didjeridoos show up a lot these days, so we need to record them. For the uninitiated, a didjeridoo is a long pipe, traditionally made from a log hollowed out by termites (although contemporary instruments are sometimes made from PVC pipe) and played by blowing into one end. They’re hard to record because when played indoors they use the entire room as a resonator.
Most players place the unblown end of the didjeridoo a few inches above the floor, and a mic placed a foot or so away picks up the main sound nicely. I’ve used RE16s and SM81s in this application with good results. The room sound may be picked up as leakage on other open mics in the room, or (in an overdubbing session) can be caught with a good omni mic (like the Oktava MK-012 with its omni capsule) or another SM81.
However, this gets only an outline of the sound. If possible, I prefer to mic a didjeridoo in classical fashion: with a pair of crossed cardioid condensers in “X-Y” configuration (nose to nose at 110 degrees with the capsules stacked directly on top of each other), placed several feet in front of the player. A pair of Neumann KM84s or Oktava MK-012s (cardioid capsules) does a great job.
Obviously, this is not the set-up to use if there’s a lot of leakage in the studio (unless you want to capture that leakage as ambience). But it gets the best rendition I’ve heard of the didjeridoo’s remarkable sound—and it’s distinctly preferable if the performer plays with the unblown end in a bucket of water, as some virtuoso players will, or whirls around in a circle like a human Leslie speaker.
First let’s sort out some confusion. There are two instruments called the dulcimer, and the only resemblance they bear is that they both have strings and are hard to keep in tune.
The mountain dulcimer is about two feet long and shaped like an hourglass. It seems to be indigenous to the southern United States, although it has relatives in northern and central Europe, and it possesses three to five strings, several of which are normally played as open drones. It’s held on the lap, strummed with a flatpick or sometimes a goose quill, and it can be fretted with the fingers or with a hollow reed.
The difficulty in recording the mountain dulcimer comes partly from its complex radiation pattern and partly from the pick. Goose quills are relatively quiet, but most players use a thin plastic flatpick. When you aim a small-diaphragm condenser mic (like a KM84) at the instrument, the recording sounds hard, harsh, and brassy (that’s the radiation pattern) with high levels of pick noise. The pick sounds like a flapping windowshade: very plasticky and unpleasant.
I’ve found two solutions. One is to use a ribbon mic, the Beyer M-260. Its softer sound helps blunt the harshness and soften the edge of the pick noise.
The other is to use a KM84 or MK-012 or a crossed pair of them, but not aim them at the instrument. Instead I aim the mic(s) parallel to the body of the dulcimer, placing the sound 90 degrees off-axis to the capsule(s) and a foot or so away, so the mic(s) will be pointing at the player’s chest. This uses the off-axis response of these mics, which is notably softer than the on-axis, as a natural noise filter, while the increased distance decreases the relative level of pick flapping.
Of course there’s a bleed-through problem if the same musician is also singing. My solution to that is to make a virtue of necessity: use the same mics to record both vocal and instrument. A carefully placed crossed (X-Y) pair will pick up the voice and dulcimer in a nice mix with good spatial characteristics, and you can control the mix by angling the array up and down.
An alternative is to place the instrument mics well to the side of the performer’s body, still pointed at their chest, and use the vocal mic close in. In either case, a pantyhose pop filter is essential.
One additional factor: in every case where I’ve made the comparison, a tube mic preamp always sounded better on mountain dulcimer than a solid-state preamp, even a very good one. The solid state preamps exaggerate pick noise to the point of unlistenability, even with the mic placement tricks described above. The tube preamp is cleaner and clearer.
The hammered dulcimer is an entirely different animal. It’s a trapezoidal box, usually about 2' x 3' and 6" thick, it sits on a stand, and has several dozen sets of strings that the player hits with small hammers. (Nancy Lippincott describes it as a “piano after taxes.”) It’s ancient, and found across Europe and the near East as the Iranian santur, the Hungarian cimbalom, and many other variations.
For some reason, although there are excellent hammered dulcimer players in Britain, most of their recordings sound awful: clangy and metallic with a sound resembling a bad music box. This comes when the instrument is miked by what would seem like the most obvious choice: small condenser mics. They pick up far too much hammer noise on the initial strike, losing the tone and richness of the instrument.
For many years I preferred to mic hammered dulcimers with a crossed pair of Beyer M-201 dynamic mics, which look like KM84s with a gland condition; they round off the sound and aren’t stimulated to ringing by the abrupt impulse of the hammers. In recent years I’ve switched to Shure SM81s, which give a very similar sound but slightly cleaner. The worst choice is a bright mic like an SM57 or an RE200; you’ll get more clang than notes.
When possible I use an X-Y crossed pair, panning the mics wide for a recording where the dulcimer is taking the leading role or more narrowly when it’s used for back-up or ensemble work. I always try to give it some width in the mix, though; this is a large instrument, and it needs to take up some space in the soundstage. (Occasionally I’ve placed the hammered dulcimer in the middle of the ensemble so its mics could pick up leakage from everyone else, doubling as ambience mics.)
I place the mics about 12" above the front lip of the instrument, aimed downwards at a point about 70% of the way in. This keeps them out of the way of the hammers and keeps the highest notes off-axis to the mics, which (in the case of SM81s) softens the impact appropriately.
In a crowded situation or with an exceptionally clattery player, it may be useful to experiment with miking the bottom of the instrument with SM81s or KM84s, although I usually find the results too muddy for my taste. A pair of miniature condenser mics located inside the instrument can also be useful, although the lovely sense of air around the body is lost.
This is a North African hand drum shaped like a rocket motor. I single it out from the general run of percussion instruments because it’s the single hardest instrument I’ve tried to record. The name comes from the two sounds it produces: a sharp “BEK” from the strike, and a deep, drawn-out “DOOUM” that radiates from the rocket nozzle. (Some players carry their affection for this sound to near-worshipful lengths—hence the phrase “Temple of Dooum.”)
What makes the dumbek a bear to record, even as percussion instruments go, is that the DOOUM sound relies on the room’s resonance; it’s mostly heard out in the room, not near the drum.
The “BEK” is fairly simple to re-cord; I’ve found the Electro-Voice RE15 does a good job, as does the slightly sharper sounding Beyer M-260. (I suspect the RE200 will do well too, but haven’t had a chance to try it yet.) The “DOOUM” is tougher; miking the bottom of the drum doesn’t work at all, yielding a muff- led roar without the sense of pitch necessary for good definition.
I’ve had my best success when other mics were open in the room, catching the drum’s room resonance by leakage. If the dumbek is on its own you can set up a few omni mics around the room (an AKG C414 or Neumann U87, if you have one, can work wonders; otherwise omni-capsuled MK-012s or an Electro-Voice RE55 are good choices).
The main problem is with phase; the different distances can give the strike and resonance an odd coloration. When the dumbek is alone in the room, another approach is to place an M-260 a foot above the drum (about 6" in front of the rim, angled toward the skin’s center). Add a crossed pair of mics (SM81s are a good choice) just above the M-260, aimed away from the instrument to catch the room sound. Again, these may also serve as ambience mics if other players are present.
All that said, I still find the dumbek almost impossible to reproduce on recordings.
Harp (the kind with strings)
These are becoming popular again after much neglect; they have a hideously complicated radiation pattern and extremely sharp transients, especially if played with fingernails.
Large floor-standing harps can be miked close-in for a “bigger-than-life” sound, good for what’s been termed “Hearts-of-Space-Celtic-Music.” A crossed pair of KM84s or MK-012s placed about 6" from the soundboard (at the bottom of the strings) can sound very good; if the bass is too boomy, try SM81s or even a large-diaphragm mic like an AKG C414 or a Neumann U87, perhaps flanked by a pair of SM81s for stereo breadth.
A more realistic sound (and one I usually prefer) comes from pulling the mics back to about 2' from the instrument, level with the player’s hands to pick up the tiny details when the fingers touch the vibrating strings. The same mics work well in both positions.
Smaller harps such as Irish table harps usually sound best when miked with a crossed pair of KM84s or MK-012s from a foot or so away.
Harp (the kind you blow)
Electric harp is simple: you mic the amp just like any other electric instrument. (Incidentally, I’ve been using the Electro-Voice RE200 as an amp mic for the last few months. It’s spectacularly good.)
Non-amplified harmonica is something else again.
The problem is, again, a tough radiation pattern and lots of spurious noise (in this case, breath blast). I’ve had good luck with the slightly bright mics often used for men’s vocals: Electro-Voice RE16s and RE20s, and the Sennheiser MD421II, the new version of the MD421.
Large-diaphragm condensers like the U87 are also good, and a too-thin harp sound can be beefed up by using a Beyer M-260 (be sure you use a foam windscreen). It’s a good idea to set up a pantyhose pop filter a few inches in front of the mic to keep the player from getting too close; if s/he’s right on top of the mic, a tiny shift of position creates gross volume differences. Wherever you place the mics, some compression is usually a good idea.
It is impossible to record a harpsichord successfully.
Back in the ’60s, arty rock bands experimenting with exotic sounds began messing around with harpsichords until they became as inappropriately overused as didgeridoos are today. (Imagine Scott Joplin rags played on a pedal harpsichord.) Art-rock died horribly—some say from an overdose of harpsichords—but as the ’90s mine every retro possibility, harpsichords have returned to pop music.
Which is actually okay with me, as I like the bold, brassy quality they impart. But they remain impossible to record, or nearly so. Their radiation pattern is murderously complex and their mechanisms clatter like a Model-T Ford. Still, we can but try.
I prefer to record the harpsichord classical-style: with a crossed pair of microphones out in the room. The basic rule, as with cigars and rattlesnakes, is “Don’t get too close!”
To find a good location, stick a finger into one of your ears, then walk around the room (bending down as needed) until you find a spot where the harpsichord sounds tonally balanced without too much mechanical noise. (Three to five feet from the edge of the instrument, facing the opened lid, is a good starting point.) The sound should be brash, but with sweetness on top.
Mark the spot and place a pair of mics there: Neumann KM84s, Oktava MK-012s, or (for a softer room sound) Shure SM81s. I like the ORTF arrangement, with the mic capsules spaced 7" apart and angled outward at 110˚ from each other. But if this picks up too much room crud try XY (mic capsules stacked one above the other, nose-to-nose, again angled at 110˚). Large-diaphragm condenser mics such as Neumann U87s may also work. In my experience a crossed pair will sound sweeter than a single mic in the same position. Why? I don’t know.
If you can’t record the harpsichord from a few feet away (because of leakage problems, for example), try using Beyer M-260 or M-160 ribbon mics closer in, again in a crossed pair. (These are hypercardioid microphones, so make the angles in XY or ORTF 90˚ instead of 110˚.) You can also try miking the bottom of the instrument (perhaps with a miniature lavalier mic) or using goboes to improve isolation.
Finally, for the experimentally minded, the harpsichord lends itself well to manipulation. In the spirit of the ’60s, try running the amplified mic signal (or recorded track) through a guitar stomp box or other finagler. Just do me a favor: don’t play Scott Joplin on the thing, okay?
Jew’s harp (AKA Jaw harp)
Jew’s harps and similar instruments are found throughout the folk cultures of the world. Balinese gamelan orchestras use large Jew’s harps as bass rhythm instruments, while the Tuva people of Siberia use them to induce religious trance states—an application also found in Southern Italy, where the sound of the Jew’s harp is considered hypnotic and seductive. The Jew’s harp in America comes in three varieties: Orthodox, Reform, and Conservative. (I’m allowed to make that joke, being of the tribe.)
Recording the instrument isn’t as hard as one might expect; it uses the cavity of the mouth as a resonator, and usually you can record it like a vocal (although one without much bass). Your favorite vocal mic will usually do fine; I like a U87 when I can get it, but am happy with an Electro-Voice RE15.
If the musician alternates between singing and playing the Jew’s harp, of course the same mic will do for both. But avoid anything too bright, as you’ll get too much buzz and not enough tone. Since sharp inhalations and exhalations are part of the technique, you’ll need a good windscreen.
On Deborah Holland’s brilliant recording “The Panic is On” (Gadfly), listen for the sound of a Jew’s harp played through a Mutron effects pedal. (Quick plug: everyone reading this should run out and buy this record for a textbook example of superb, imaginative recording technique and arranging. The instrumentation, aside from the odd Jew’s harp, is almost entirely limited to electric bass and percussion, and it sounds rich, full, and complete. Oh, and the singing is fantastic.)
Incidentally, instruments that use parts of the body as resonators (such as the Jew’s harp or the similar-sounding mouth bow) are known as idiophones. Somehow that makes sense.
“The best thing about the pipes is that they don’t smell, too.” (Robin Williamson)
All joking aside, most people never get to hear any bagpipes other than the Scottish war pipes, which were never meant to be heard indoors. I once recorded a concert of ten Scottish pipers and four bass drummers in a stone-walled chapel; it’s the only music I ever heard that was louder than the Who.
I set up an ORTF pair of SM81s in the hall, ran my cables to the lobby, and got my ears out while they still worked. Ow. I used SM81s because my KM84s went into overload, even with the pads engaged.
There is more to pipes than the deafening war monsters, however. The world’s cultures are filled with smaller instruments: the Northumbrian smallpipes, Italian zampogna, the Bulgarian gaida, and the Irish Uilleann pipes, which I’m told take 21 years to learn to play properly and which are used to lull babies to sleep. In fact, you’ll find bagpipes wherever you find sheep, a point on which it’s best not to dwell for too long.
All these pipes contain several elements: usually one to three drones, a melody chanter, and (in the case of the Irish pipes) “regulators”—drone pipes whose pitch can be shifted by manipulating valves with the edge of the hand.
There are two ways of miking small pipe-sets. The first is to record them with a crossed pair of cardioids from a couple of feet away (I use KM84s in XY position; ORTF exaggerates the player’s movements too much). This gives an excellent stereo image. An alternative method, necessary when leakage from other instruments makes close-miking necessary, is to use one mic on the drones and another on the chanter and regulators, placing each about 10" from the holes.
Again, I find KM84s (or MK-012s) ideal for this purpose; their broad cardioid pattern picks up several pipes at once. Large-diaphragm condensers, in my experience, are less successful as their off-axis colorations are too great, while most dynamic mics have inadequate transient response for the pipes’ delicate notes. When close-miking pipes, I usually pan the two mics slightly apart in the mix, just enough to provide some dimension to the instrument.
“Quills” are actually panpipes: hollow reeds lashed together into an array, end-blown, and often played in a harmonica holder. They’re common in Andean music, but were also widely used in African-American tradition, particularly in Mississippi, Louisiana, and Texas.
The pre-blues songster Henry “Ragtime Texas” Thomas recorded the classic “Fishing Blues” in 1927 (“Many fish bites if you got good bait/Here’s a little tip that I would like to relate...”—that “Fishing Blues”) with guitar and quills. With the re-release of that recording on the ‘Anthology of American Folk Music’ (Smithsonian/Folkways—another set of discs you should run out and buy), quills may make their way into the pop music sphere again, as they did in the ’60s with the Lovin’ Spoonful.
How do you record them? Like a vocal, but not too close —say 10–12" away, as the end quills will be off-axis. A large-diaphragm condenser mic is ideal.
I’m lumping together dozens (no, hundreds) of small percussion instruments because they share a common problem: they’re hell to record. I refer to claves, castanets, small bells, spoons, rattles, and hundreds of ethnic instruments whose names I don’t know—little things that you hit, scrape, or otherwise abuse.
Most percussion instruments have two components to their sound: an initial strike transient and a fundamental resonant sound. The initial strike is usually sharp and spiky, and very high-amplitude compared with the rest of the sound. Indeed, the peak-to-average ratio of small percussion instruments can be greater than 20 dB.
But we have 90 dB or so of dynamic range on a digital recording system, even one with only 16 bits. So why should a 20 dB peak-to-average ratio give us fits?
The problem is twofold. First, in the real world these instruments are quite loud—louder than they seem to your ears, because we judge loudness based on the average level of the main resonance rather than the peak level of the initial transient. Almost no one listens to recordings of these instruments at anything near their real-world loudness (at least not without the cops being called). When you listen to a sound, even one recorded perfectly, at a level much lower than its natural level, it sounds tinny and thin; the “klock” of the wood block has no body and spoons go “clink” instead of “pock” as they do in real life.
The other half of the problem is the way most people record percussion. Knowing that the initial transient is very sharp, most people assume they need a “fast” microphone (one with quick transient response) to capture it properly, and pull out a condenser microphone. Wrong answer; condenser mics usually have high frequency resonances that are excited by the fast initial transient, and they can exaggerate the transient to the point where it masks the main sound entirely. Combine this with reduced listening level, and you get thin, tinny sound every time.
So what’s the answer? It’s counter-intuitive: use a dynamic microphone, one with a flat high-frequency response. My favorite, the RE15, has miked dozens of small percussion instruments, and over and over has proven itself the most real-sounding choice. The damped high-frequency response minimizes ringing on the initial transient, while the nonlinearities of the moving-coil mechanism provide natural compression that considerably lessens the dynamic range. By using this natural compression to shrink the peak-to-average ratio, the sound remains full and rich even at low listening levels.
This is, of course, not news to many of you; dynamic mics are the norm for miking kick drum (e.g., the Electro-Voice RE20) and tom-toms (the ubiquitous Sennheiser MD-421). But I’m surprised many engineers haven’t applied the same logic to smaller instruments; dynamic mics belong there too.
Along with the RE15 (or RE16 if you want a tad more brightness) and RE20, the Beyer M-201 is an excellent microphone for small percussion, as is the late, lamented AKG D-224. The Beyer M-260 is good on rattles. The Beyer M-88 looks like a good bet on small percussion too, although my own experience with this excellent mic has been limited to vocals.
The cabasa is a special-case instrument, as its volume is overwhelmingly loud. I’ve had the best luck miking it with a Sennheiser MD-421; few other mics can handle the level (although I bet an M-88 could). If the player is jumping back and forth between several instruments, mic the cabasa separately and (if possible) give it a track of its own. Prepare to use lots of compression, and you may want to eq the treble down a few dB as you record.
Well, synthesizers are the easiest instruments of all—you just plug them into the board, right?
Not necessarily. Since the modular synthesizer emerged from Bob Moog’s lab in 1965, musicians have complained that they never sounded like real instruments. Their sound came from nowhere, ethereally, without being located in a real space. The synths had no body, no solidity, no (if you’ll pardon the expression) balls.
In my opinion, synths don’t sound like real instruments because they aren’t real instruments: devices that produce sound waves in a real space. Instead, they produce electrical waves that are recorded directly onto magnetic tape, and I hope the effects box manufacturers will forgive me, but even the best room simulators don’t sound like real space to me—yet. (They’re getting better, though.)
Luckily, the solution is simple: if you want a synthesizer to sound like a real instrument, make it a real instrument. Patch the synth into a power amplifier and run it into a loudspeaker placed in your studio. Suddenly your synth is making real sound in a real space, which you can then record.
How? In as many ways as you would any other instrument. You can set up a crossed pair of KM84s several feet away and record the sound in the room space, or mic the speaker up close with an EV RE20, or a combination of the two.
You have a wide choice of speakers too; there’s no law saying you need a high-fidelity speaker. Colorations are part of the sounds of real instruments, remember? So you can use a handy guitar amp for one type of sound, adding in some tone controls and distortion as desired (I have a blackfaced Fender Deluxe that’s done yeoman work for me on synths), or you can hook in a high-quality speaker of any sort, or a kludged array of cheap car speakers if you prefer.
Is your monitor speaker multi-way? Then it will produce a nice, complex sound-field in the room, just like other real instruments—which makes for more interesting sound. Again, feel free to experiment; once you’ve made the synth into a sound-producing device rather than an electrical-signal-producing device you open up a huge range of possibilities.
One caution: if you’re using a hi-fi or monitor speaker, go into the room occasionally and make sure the speaker’s volume isn’t excessive. Synths can produce powerful bass transients and you really don’t want to blow up your woofer. About 83 dB usually works fine (loud enough to interfere with conversation but not loud enough to require shouting). Guitar amps, of course, are designed to be overdriven—but don’t put high-level bass through a lead speaker. You knew that, of course.
NB passes on a good hint: if you’re using a MIDI wind controller, set up a microphone and record the sound of yourself blowing it, preferably on a separate track. (Or record it with the MIDI signal, if your system lets you do both at once.) Even if breath noise is part of the sound you’re triggering, this adds the sound of breathing. The same for just a touch of pick noise, if you’re using a MIDI guitar controller. A large-diaphragm condenser mic is good here, or perhaps an SM81—nothing too bright. [On EWI, I use an Audio-Technica ATM-35, which is a clip-on miniature cardioid condenser.—NB]
The tablas are a pair of tuned Indian drums with an enormous variety of sounds. They get their own spot because I’ve found the perfect mic to use on them: the AKG 409, which is also often used on kick drum. I suspect the 409 may prove useful in the Case of the Intractable Dumbek, reported last issue, and on other difficult percussion instruments.
Thanks to Brad Sarno for introducing me to this microphone a few weeks ago, when I had occasion to amplify and record tablas at a Hare Krishna rally, an interesting experience in itself.
The most important fact to know about recording woodwinds is a little-known one: most of the sound comes out of the first open hole, not from the mouthpiece or the bell (if present).
That means the folks who shove a microphone down the bell of a saxophone or in the face of a flute player are missing most of the action—and in the case of the flute, they’re also picking up a lot of annoying breath noise.
(In extreme cases, I’ve mistaken this for distortion—it can sound very like clipping—and gone frantically hunting for the elusive overload before standing in front of the player and realizing it was his wind that caused the problem. Sorry, Cathal.)
If you mic a flute from above with a KM84 or MK-012 about a foot over the flautist’s hands, you can get a rich and sweet recording without excessive breath noise. Likewise, putting a mic about a foot over the keys of a saxophone, clarinet, or oboe gives a far richer sound than miking the bell, which tends to sound squawky. Bassoons are better miked from about 18", still over the player’s hands.
For all these instruments I usually prefer KM84s or MK-012s for their wide cardioid patterns, which provide even coverage of all the holes. U47s are legendary for miking winds, but I don’t anticipate being able to afford one soon. (Sigh.) An interesting alternative for a softer sound is the AKG D-224 2-way dynamic mic, while if you have the luxury of recording from slightly farther away without leakage problems, a Beyer M-260 ribbon mic gives a nice blend of sweetness and bite.
(Well, I had to find something that began with Z.) The zither family includes the moderately common Autoharp™ and a wide variety of bizarre instruments marketed around the turn of the century that prove some people didn’t need drugs to hallucinate.
(Picture an Autoharp-sized box with one set of strings arranged in chords that you played with your thumb while simultaneously plucking a separate set melody strings with your fingers and changing their pitches with a metal slide on an articulating pantograph arm. I think you held it between your knees. A variant, the “ukelin,” had a set of chordal strings that you plucked and a set of melody strings that you bowed.)
Anyway, the only zither family instrument you’re likely to encounter is the Autoharp, which can be miked from the front with a KM84 or MK-012. If the performer sings while playing, you can try picking up both on a single microphone or attaching a miniature lavalier microphone to the player’s shirt so that it picks up the instrument’s sound from the back. This also limits the mechanical noise and gives a fullness of sound that’s closer to what the player hears than most frontal recordings.
Wrapping it up
Looking over my recommendations on these tough instruments, a few themes emerge. I note that I don’t often use large-diaphragm mics on these tough guys, and not just because I have less access to them. To my ears, the big boys aren’t for everything (neither is any microphone); I tend to use them more for voices, guitars, and pianos.
For exotic instruments, I find myself using small-diaphragm condensers like the KM84 or MK-012, medium-sized SM81s, or (surprisingly often) dynamic microphones.
I think there’s a lesson here; reading the rec.audio.pro newsgroup, there’s a sense (especially among newcomers to the field) that only condenser microphones are truly professional and that large-diaphragm condensers are the most important. Dynamics seem to be looked down on as “PA mics,” not worthy for serious recording (except of drums and guitar amps).
That’s a shame, because there are superb dynamic mics being made, and there are certain applications (like small percussion) where a good dynamic is the best possible choice.
The other theme implicit in the article is the possibility of expanding our musical palette. The craze for “world music,” although it’s produced its share of shallow and anachronistic recordings (why the fad for African percussion in Irish music?), serves to remind us that guitar-bass-keyboard-drum kit isn’t the only line-up that can produce meaningful sounds.
Humans have been musical for a long time. We’ve evolved thousands of ways to make noises: joyful, eerie, lonesome, sexy, exciting, hypnotic. The byways of the musical world are open; put on your pith helmet and explore.
See ya later, Dr. Livingstone.