Given their exceptional capabilities and ease of use, I have long puzzled over the comparative lack of attention stereo microphones get. The oversight can't be blamed on a dearth of products — no fewer than 17 microphone companies offer stereo mics, and more than two dozen different models are currently available.

Maybe one reason stereo mics get overlooked, especially among personal-studio recordists, is a general perception that they're expensive — perhaps due to the fact that a couple of well-known models cost upwards of five grand each! Then there's the issue of variety: stereo mics come in a surprising, sometimes bewildering range of shapes, sizes, and types, reflecting a number of different approaches to capturing stereo sound. As is true of all mics, some that are particularly well suited for certain duties are often not so well suited for others. Whatever the case, the basic problem is unfamiliarity — it seems a lot of folks simply aren't aware of how useful, affordable, and downright wonderful stereo microphones can be.

This article is intended to help remedy that situation. It culminates in an evaluation and comparison of six of the most affordable stereo mics out there: the Audio-Technica AT825, beyerdynamic MCE 82, Crown SASS-P MKII, MBHO MBNM 622 E PZ, Røde NT4, and Sennheiser MKE 44P. Priced below $1,000 each, these six mics cover a range of applications at prices personal-studio and location recordists can handle. Before getting into the evaluations, I'll discuss stereo microphones in general, covering the various types and how they work, the ways in which dedicated stereo mics are superior to stereo pairs of mics (and vice versa), and how, when, and where to use stereo mics. If you are considering buying a stereo microphone, or simply want to know more about them, then step through this curtain.

I'll start with a riddle: when does one plus one equal three? That might sound like a trick question, but actually it helps elucidate the common, if seemingly magical, phenomenon of stereo sound. As former NBC engineer Randy Hoffner famously put it, “Stereo does not equal mono times two.” Consider a setup for stereo recording and playback: two mono microphones send separate mono signals to separate mono loudspeakers. Yet the result for the listener is three-dimensional sound (or some semblance of it, depending on the quality of the recording, placement of speakers, and so on). Indeed, that's why it's called stereo — the Greek word stereos means “solid” in the sense of comprising three spatial dimensions: breadth, depth, and height. Stereo, then, is more than the sum of two mono signals — a kind of synergistic leap occurs between the pair of speakers; one plus one equals three dimensions.

Of course, a stereo-recording rig isn't necessary to experience the phenomenon of stereo — the apparatus sitting atop your shoulders will do fine. Just as human vision is stereoscopic (different information from each eye is required to accurately gauge depth), human hearing is stereophonic. Try this experiment: when listening to a band or orchestra (ideally with the musicians spread out across a wide stage), close your eyes, turn your head so one ear is aimed toward the music, and shut off the other ear with your hand. Listen that way for a while. You will notice that it is difficult, if not impossible, to discern the location of individual instruments; the music seems to emanate from a single point or area. Moreover, the depth of the instruments, or distance they sit back from the front of the stage, is obscured.

Essentially, by covering one ear, you've changed your head from a stereophonic to a monophonic receiver. Now uncover your ear, turn your head to face the music (with eyes still closed), and listen again. Thanks to subtly different information reaching each ear (more on this later), you can now hear the three-dimensionality of the space and mentally pinpoint the location of individual instruments on the stage.

A stereo microphone performs an analogous feat. Its two “ears,” aimed in different directions, capture two different “views” of the sound. Conveyed to two transducers (whether headphones or loudspeakers), those two views combine to re-create for the listener a sense of the original space and the music (or whatever) occurring within it. Stereo mics, then, are basically little “hearing heads,” devices that approximate, to varying degrees, what human heads do in the act of hearing. That's no small feat. To the extent that stereo hearing is the human head's peak auditory achievement, stereo mics are crowning achievements in recording technology — in these devices culminate centuries of discovery in the sciences of sound, hearing, and recording.

Of course, two separate microphones, properly positioned, can provide results that are virtually identical to what you get from a stereo mic. And certainly a pair of mics is more versatile in terms of placement. But the great thing about stereo mics is how easy they are to use — sort of like your head. The two capsules are already sonically matched and joined together in proper alignment on a single body, ensuring quick setup and problem-free stereo recording. For those with no experience setting up stereo pairs, that might not sound like a big deal. But if you regularly record in stereo with two mics, you know what a production it can be getting the mics positioned just right, especially up high over a drum kit or in other out-of-the-way places. Then all it takes is someone accidentally kicking a stand or tripping over a cable to knock the whole thing out of whack.

Contrast that with a stereo mic. Just set the thing up, aim it at the sound source, and you're pretty much ready to roll. For me, a gigging musician and recordist dedicated not only to playing but also to writing and producing music, stereo mics are a godsend. I can take my stereo mic and portable DAT recorder to gigs, for example, and with that bare-minimum setup document my playing — great for learning purposes — in high fidelity. I can also turn out pro-quality recordings culled from best takes. In the studio, a good stereo mic can serve up a complete drum-kit sound fast, which means that when inspiration strikes, I'm less apt to get bogged down by the gear and forget the groove I was meaning to lay down. Likewise, if a friend comes over with his guitar to play me a new song, I can be set up to record it in a matter of minutes. What's more, given a really good stereo mic, the result can sound like a record already, complete with a you-are-there realism that only a stereo recording can convey.

In short, a stereo mic is the quickest path to sonic realism. For those whose goal is re-creative recording — recording intended to convey actual sonic events — a stereo mic is the ideal tool for the job.

Occasions that commonly call for a re-creative approach to recording (as opposed to a creative one, in which the goal is to alter the source sound) include most on-location jobs: electronic news gathering (ENG), concert taping, nature recording, and the like. (For information on nature recording, see “Going Wild” in the April 2003 issue of EM.) In such situations, the goal is to capture the sonic event as it happened, with minimal enhancement of the original acoustics.

Many styles of music — jazz, blues, classical, and rock, to name a few — also rely heavily on re-creative recording, making stereo microphones very useful in the studio. In addition to saving valuable time, stereo mics are largely foolproof, ensuring that stereo capture is not marred by deleterious phase problems (which typically result from improper setup of two or more mics on the same source).

In the studio, stereo recording is particularly advantageous for instruments made up of multiple elements spread apart spatially: drum sets, percussion rigs, mallet instruments, and any kind of ensemble (strings, a horn section, a group of backup singers, or what have you). Of course, any sound source can be stereo-miked, and even many smaller instruments sound fuller and more natural in a mix when recorded in stereo. For that reason, engineers sometimes use stereo-miking to highlight instruments that are central to a mix — an acoustic guitar in a country song, for example. Conversely, it would make little sense to stereo-mic a shaker, especially if track count were a concern. Then again, if the percussionist were running around the room while playing the shaker, the engineer might elect to record the part in stereo to convey that movement to the listener.

That brings us to another consideration in determining whether to record something in stereo: motion. Any element that moves across the soundstage — someone opening a door and walking across a room, for example — is an excellent candidate for stereo-miking. The effect can be very dramatic and is in a different league from the simulated movement that can be created by dynamically panning a mono signal during a mix.

Another application perfectly suited for stereo mics is capturing natural reverb, as in a recording space (“room miking”), echo chamber, or reverb tank. Reverberation, after all, is a stereophonic phenomenon — in the real world it comes from all around the listener. You may have noticed that if you sum a stereo mix to mono, the reverb largely disappears. That's because with only one channel reproducing the signals, the reverberations must emanate from the same place as the direct signal — a totally unnatural situation.

A stereo field, on the other hand, allows more space around the signal, making it easier for the listener to differentiate between direct and reflected signals. It is, in part, by “calculating” the difference between the arrival of the direct signal and that of its multiple reflections as they bounce off the walls and around the room that the brain is able to determine general dimensions and sonic qualities of a space. Two channels — and both ears — are required to make that calculation. That's why classic echo chambers are typically equipped with stereo mics or pairs of mics. (For information about creating your own echo chamber and/or reverb tank, see “Recording Musician: Keepin' It Real” in the February 2003 issue of EM.)

We've established that it takes two ears (or mic capsules) to register the minutely differing cues that allow the brain to determine spatial information from a given sound field. Now let's consider the main cues: intensity and phase.

Intensity refers to amplitude, or sound energy per unit of area. Differences in intensity result from the simple fact that the closer you get to a sound source, the louder it sounds, and vice versa. That is, the brain interprets louder sounds as being closer than quieter ones.

Phase refers to the timing of the signals. Differences in phase result from the signals arriving at the two ears (or mic capsules) at different times. This results in some degree of phase distortion or incoherence. (Conversely, signals that arrive at the same time are said to be phase coherent.) As we'll see, some phase distortion in stereo recordings can actually be perceived as an enhancement — it's what creates a sense of “air” around the sound sources. Beyond a certain point, though, its effects begin to degrade the sound.

To localize sounds, the ear and brain system processes those two cues in differing amounts, depending on the frequency of the incoming sound. Low frequencies have big sound waves that go right around the head, and thus are perceived omnidirectionally, as if coming from all directions. High frequencies, because the smaller sound waves are effectively blocked by the head, are heard very directionally — a small wave coming from one side reaches mostly one ear, so the main thing registered between the ears is an intensity difference. And midrange frequencies are localized using a combination of timing and intensity cues.

As noted earlier, the variety of shapes, sizes, and capsule configurations among stereo mics reflects (no pun intended) a sizable range of techniques devised to capture stereo sound. That range can be divided into two broad categories: coincident and noncoincident. (Those designations apply equally to stereo mics and stereo pairs of mics, of course.) Basically, coincident means the two capsules are positioned close together and noncoincident means they are spaced apart.

In coincident techniques, the two capsules are positioned as close together as is physically possible. In the case of an end-address stereo pair (two small-diaphragm condensers, for example), this is typically done with one capsule directly above the other (see Fig. 1), so as to accommodate the mic bodies. A stereo mic, on the other hand, permits the option of positioning the capsules side by side. Either way, the capsules are angled symmetrically on either side of the midpoint of the sound source so that each picks up one side, or half, of the stereo image. Due to the close proximity of the capsules, incoming signals arrive nearly simultaneously at the two diaphragms, largely avoiding any phase distortion. This ensures not only a stable, mono-compatible image, but also angular fidelity to the source sounds. Angular fidelity means the sounds are positioned accurately across the reproduced soundstage, true to their positions in the original sound field.

Because coincident techniques preserve only intensity cues, they are often called intensity stereo. With intensity stereo, the character of the stereo image is determined by three things: the choice of polar pattern, the angle between the two diaphragms, and the position of the two capsules relative to the sound source. Time-of-arrival differences are not part of the equation.

Common coincident techniques include XY, middle-side (M-S), and Blumlein. XY seems to get the most use, both in terms of stereo pairs and dedicated stereo mics. Indeed, four of the six microphones evaluated in this article have XY-configured capsules. In an XY setup, two directional capsules (cardioid, supercardioid, or hypercardioid) are angled somewhere between 60 and 135 degrees apart. In most lower-priced stereo mics, the angle is fixed, typically at 90 or 110 degrees. However, some costlier stereo mics with XY capabilities provide variable control of the angle. Thanks to the directional pickup of the capsules, XY is an excellent choice when you need to minimize sound — audience noise, for example — coming from behind the caps.

The Blumlein technique, named after audio pioneer and inventor Alan Dower Blumlein, specifies two bidirectional (figure-8) microphones or capsules positioned so that their principal axes are at 90-degree angles to one another (see Fig. 2). In that configuration, the main pickup axis of each mic or capsule is precisely aligned with the null axis of the other. The result is very accurate stereo representation — the most accurate, many believe — of the original stereo sound field for sounds arriving from the front. Of course, the Blumlein array also picks up sound equally from the rear (albeit in opposite polarity). This can be an advantage or disadvantage, depending on the situation. In the case of a great-sounding performance in a great-sounding space, a Blumlein setup might be just the ticket; but if the space is acoustically undesirable — overly reflective, too noisy, or what have you — the Blumlein array would be a poor choice, given its faithful reproduction of total-room acoustics.

The middle-side technique, unlike other coincident-capsule setups, makes use of two different polar patterns. Traditionally, the “middle” mic or capsule is directional (usually cardioid) and the “side” mic or cap is bidirectional. The mid mic is aimed at the source, or 0-degree axis; the side mic is positioned vertically coincident with the mid mic such that its null plane is aligned with the principal axis of the mid mic (see Fig. 3). That way, the side mic picks up information mainly from the left and right and largely rejects sound coming from the front and rear. Because the two sides of the bidirectional mic are aimed in opposite directions, they produce opposite-polarity signals, essentially dividing the room into distinct halves — polar opposites, actually.

One advantage of M-S recording is that you can adjust the width of the stereo spread after the recording is completed — a real boon in situations where the side mic picks up more extraneous noise than you had counted on. This is done simply by changing levels of the two channels relative to one another: increasing the level of the mid signal tightens the stereo image and makes the source seem closer; increasing the level of the side signal makes the source seem more distant.

To produce a stereo image, the mid and side signals first must be “decoded” in a special sum-and-difference matrix. Basically, the mid and positive side signals are combined to create a new (directional) signal aimed in one direction (typically left), and the mid and negative side signals are combined to create a new (directional) signal aimed in the opposite direction. That is, the left channel equals mid plus side (M+S) and the right channel equals mid minus side (M-S). Together, these two new signals form the final stereo image. Engineers who specialize in M-S recordings often use dedicated sum-and-difference modules, which guarantee proper matrixing. Certain stereo M-S mics, such as the Shure VP88 and Pearl Labs MS 2, also provide built-in sum-and-difference matrixing. (For more information on M-S recording, as well as stereo recording in general, see “Double Your Pleasure” in the June 2000 issue of EM.)

Noncoincident techniques, by virtue of employing two capsules spaced apart, are able to capture both intensity and timing cues — just like the human head with its noncoincident ears. As a result, they tend to sound more open than coincident techniques, providing an increased sense of “space” or “air” around the performers. That additional sense of openness results from phase anomalies introduced by time-of-arrival differences between the capsules. That is, the phase distortion, which might be heard as “comb filtering” or “phasiness” when the signals are summed to mono, is heard as an enhancement to the stereo field. But again, there are limits on how much phase distortion the ear will accept before it begins to perceive the sound as unfavorable.

The head is the inspiration for most noncoincident recording techniques — obvious by the fact that they employ capsules spaced six and a half or so inches apart, which is the average width of human heads. Indeed, the spectrum of noncoincident techniques might be viewed as a continuum ranging from those most like the human head to those least like it. At one extreme (the one least like the head) are widely spaced pairs, also called A/B stereo, which in some cases are positioned several feet or even yards apart. (I won't delve into those here, given that the focus of this article is stereo mics.) Most noncoincident techniques, however, take their lead from the head, at least in terms of capsule spacing; indeed, they are often described as near-coincident rather than noncoincident.

If we rank the near-coincident techniques in order from most like the human head to least like it, binaural comes first, followed by quasi-binaural, and then “closely spaced pairs” (which may or may not comprise a single microphone). Most readers are familiar with binaural recording, which typically involves the use of a dummy head fitted with omnidirectional capsules inside the ear canals. An even closer-to-reality approach is to employ a headset binaural recording system, such as the Sennheiser MKE 2002 or Soundman OKM II K, which permits the user to wear the mic capsules (they generally attach at the entrance of the ear canal) and thus generate directional cues with his or her own head. (For a review of the Soundman OKM II K, see the January 2002 issue of EM.)

But though binaural comes closest to replicating the audio functioning of the human head and results in recordings with an uncanny sense of realism, the recordings translate poorly to stereo speakers due to crosstalk interference — the listener's right ear hears unwanted sound from the left speaker and the left ear hears unwanted sound from the right speaker. The crosstalk degrades both stereo imaging and frequency response. To hear binaural recordings properly reproduced, the listener must wear headphones — a somewhat cumbersome requirement that probably accounts for the relative unpopularity and general unavailability of binaural stereo recordings.

Various approaches have been devised to get around this unfortunate aspect of binaural recording. Those that employ some type of barrier between the mic capsules are described as quasi-binaural; those without a barrier are generically called spaced pairs (or closely spaced pairs). In either case, the goal is recordings that capture both intensity and timing cues, thus providing a more spacious sound, yet still translate well to loudspeaker playback both in stereo and mono.

Two of the stereo mics evaluated for this article — the Crown SASS-P MKII and the MBHO MBNM 622 E PZ — qualify as quasi-binaural. The Crown Stereo Ambient Sampling System (SASS) employs two boundary-layer or Pressure-Zone Microphones (PZMs) positioned 6.7 inches apart on either side of a large plastic housing. (The original, SASS-B mic could be retrofitted with higher-quality, omnidirectional Bræel & Kjür mics, but due to low demand that option is no longer offered.)

The MBHO MBNM 622 is a variation on the Optimal Stereo Signal (OSS) technique invented by Jærg Jecklin of the Swiss Broadcasting Corporation. Jecklin's stereo-miking system, better known as the “Jecklin disk,” specifies a pair of omni mics spaced 6.5 inches apart and separated by an acoustically opaque baffle (the disk) 11.02 inches in diameter. The MBNM 622 employs two boundary-layer caps (boundary-layer mics are naturally omnidirectional) mounted on either side of a metal plate (which serves as a boundary for the caps) and separated by a foam-rubber-covered half disk.

Two closely spaced pair techniques also warrant mention. The ORTF array, named after the French national broadcasting agency (Office de Radiodiffusion Télévision Française) that developed it, specifies two cardioid mics spaced 6.69 inches apart and angled outward at 110 degrees — sort of like two ears without the head in between. Widely used in Europe, the ORTF configuration generates just the right amount of comb filtering (predominantly in the 1 kHz region) to create a sense of air around the source, but without noticeably degrading the sound. ORTF typically produces an evenly spread stereo image with good localization, and the resulting signals sound good both in stereo and mono playback.

The NOS configuration, named after the Netherlands Broadcasting Foundation (Nederlandsche Omroep Stichting) where it was developed, employs two cardioid mics angled at 90 degrees and spaced 11.81 inches apart — considerably wider than the head. With this arrangement, comb-filter effects appear a couple of octaves down from the ORTF's, at around 250 Hz. Though this compromises mono compatibility of the signals, the spacious stereo field captured by NOS-configured mics translates quite well through stereo loudspeakers.

Both ORTF and NOS techniques are commonly configured with separate microphones. To reduce hassle, a convenient device called a microphone-array positioner can be used to hold the properly configured mics in place, allowing them to be positioned as a unit (see the sidebar “Dreaming as One”). Also notable is a tidy ORTF stereo mic from Schoeps, the MSTC 64g, which comprises a matched pair of cardioid caps mounted on either end of a handy T-bar-shaped body (see www.schoeps.de/E/mstc64.html).

In the evaluations and comparison tests of the six stereo mics, I was not so much looking for a “winner” as for a close understanding and appreciation of each microphone. A strictly apples-to-apples comparison wasn't an option anyway, given that four of the mics are XY stereo and the other two are quasi-binaural. But that disparity doesn't diminish the utility of the testing. The crux of the matter, after all, is to reveal each mic's “personality” — its sonic disposition, its strong suits and weak, its creature comforts or lack thereof. The comparison aspect simply provides context and perspective.

Speaking of perspective, I strongly felt the need for a benchmark microphone. The purpose of a benchmark mic is to establish a reliable reference so that during listening tests, you can readily switch to the benchmark tracks for a reality check. The ear, after all, is quick to acclimate to the way a given mic portrays sound sources, and it soon starts to hear that portrayal as “right” — at least until it hears a different portrayal that sounds “more right.” Having a benchmark keeps things honest by helping listeners not fall prey to the easy position of overestimating a performance heard in isolation.

After much research, I selected the highly regarded Schoeps CMXY 4V ($3,475) as the benchmark mic. Like four of the mics in the test group, the CMXY 4V is an XY stereo mic employing two small-diaphragm cardioid condenser capsules (see Fig. 4). The CMXY's capsules are arranged side by side rather than top to bottom, allowing the two capsules to be rotated equally in opposite directions by means of an ingenious gear mechanism that keeps them perfectly aligned throughout a combined 360-degree arc. A simple twist of either gear adjusts the angle of the two capsules relative to one another. In addition, the capsule bodies and gear mechanism are mounted atop a smoothly rotating cylinder with nearly 240 degrees of travel, further simplifying positioning of this elegant and remarkably unobtrusive stereo mic.

In devising tests for the six mics, I sought applications that not only would help me gauge each mic's capabilities throughout a range of performance criteria — spectral accuracy, transparency, transient response, angular fidelity, stereo realism, and so on — but also would represent typical uses for stereo mics, particularly among personal-studio recordists and/or nonprofessional concert archivists (better known as tapers). I set up three recording dates: one for a large band in a large room, another for a classical guitarist in a large room, and the third for a drum set miked from above in a small room. I also spent time with each mic individually, both in the studio recording to ADATs and outdoors recording to a Tascam DA-P1 portable DAT recorder.

I had the good fortune of getting to record the large band in the Studer Room at the wonderful Ex'pression Center for New Media in Emeryville, California. There I had the pleasure of working with two very capable volunteers whose assistance proved invaluable: Ex'pression Center Sound Arts instructor (and EM reviewer) Eli Crews, who operated the Studer D950 M2 console (see Fig. 5), and Ex'pression Center Sound Arts student André La Velle, who, like me, is something of a microphone fanatic (see Fig. 6). We used the Studer's fine mic preamps and 24-bit converters and routed all signals to Digidesign Pro Tools. The Studer Room features Dynaudio Acoustics BM15A (active) close-field monitors, which I am very fond of, and the mains are Meyer X10, a remarkable monitor that, I was told, employs built-in microphones that measure and adjust speaker response in real time according to the needs of the room.

Kind enough to perform for the tests was Garth Steel Klippert's band (www.garthsteelklippert.com; see Fig. 7), a six-piece original-music ensemble from the Bay Area featuring Klippert on guitar and vocals, Bill Noertker on contrabass, David Cooper on marimba, Carroll Ashby on trombone, Tom Griesser on baritone saxophone, and Dan Nelson on drums. Klippert's group nicely fulfilled a couple of criteria I had deemed critical for the tests. One was that the group be able to play acoustically, with no amplification. My reasoning for that was twofold. One, it would force us to create the final stereo “mix” by positioning the instruments strategically around the room at different distances from the microphone; that way, during playback I could judge how faithfully each mic represented the instrument positions (as well as the space between). And two, I would be able to hear how each mic reproduced the pure, unadulterated timbres of the instruments. The other criterion was that the band contain vocals, horns or strings (or both), acoustic guitar, an upright acoustic bass, and a drum kit with cymbals. I wanted to make sure I hit the mics with a range of timbres, including both low and high strings (upright bass, acoustic guitar) and low and high percussive sounds (bass drum, cymbals).

Ex'pression Center also kindly granted me use of the Studer Room for the classical guitar session. That time, however, I worked alone, recording directly to DAT using the DA-P1's mic preamps and, when necessary, Blue Blueberry mic cables. Oakland-based classical guitarist Suzie Metzler generously offered her musical services (see Fig. 8 and www.geocities.com/suziemetzler). She played a wonderful-sounding hand-built Takamine EP 90. I asked Metzler, who specializes in baroque, renaissance, and flamenco music, to choose selections that would utilize the full frequency content and dynamic range of the instrument, while simultaneously covering a range of playing styles and techniques. She selected three short pieces: a 16th-century work by John Dowland titled “An Heart That's Broken and Contrite,” “Rhapsody in Blue” by George Gershwin, and a flamenco-styled piece titled “Posades” by one of Metzler's former — and favorite — instructors, Juan Serrano.

I did the drum-overhead tests at my own studio, recording to ADATs through a Manley Langevin Dual Vocal Combo preamp and Blue Kiwi mic cables. I set up the drums for maximum sonic variety and a wide lateral spread (see Fig. 9). For the first test, I played to a simple folk-rock track I had recorded the week before that didn't yet have drums. For the second, I wrote a rhythm chart with a slow funk groove featuring drum breaks that utilized, in a prescribed sequence, the different sounds from one end of the kit to the other. I played the chart to a click track so all the parts lined up. That way, during the listening tests, I could jump back and forth between the mic signals and hear the same basic figures.

I took pains to ensure consistent microphone positioning during each test session. That was fairly straightforward with the four XY mics, but sometimes the two quasi-binaural units required I take a different tack. In the end, though, I chose to err on the side of what sounded best, in the interests of getting to know each mic with its best foot forward.

Signal levels and preamp gain were handled a couple of different ways. In the Studer Room we kept the juice equal from one mic to the next and compensated for level differences only later, during the listening sessions. That way, the Pro Tools tracks provided an easy visual indication of how hot (or not) each mic was. For the classical guitar and drum sessions, though, I spent a good amount of time precisely matching levels and noting gain settings — a different way of formulating the same equation.

Following the three sessions, I digitally transferred the Pro Tools and DAT tracks into MOTU Digital Performer, lined them up, and bounced them to fresh ADAT tape in the same order as the tracks I had already recorded to ADAT. Once all the tracks were side by side on tape, I used mute buttons on my analog mixer to jump back and forth between channels — an easier, if more old-fashioned, way than clicking with a mouse.

I'll describe each mic physically before getting to the evaluations. You can tell a lot about a mic's intended use just from its design, features, and packaging and get a good idea whether it's appropriate for your needs.

Known in some circles as the “Grateful Dead mic” due to its long-standing popularity among Deadhead tapers, the Audio-Technica AT825 (see Fig. 10) is also popular for ENG and other field-recording duties. Hefty without being heavy, the mic has a cylindrical metal body with a matte gray finish and is topped with a hardened-mesh grille basket. The rectangular shape of the basket makes for easy mic orientation in low-light situations, and a silk-screened L/R logo identifies left-right orientation. Beneath the logo is a small recessed switch for engaging a 150 Hz highpass filter.

The mic can be powered externally by phantom power or internally by 1.5V AA battery — the top part of the cylindrical body unscrews and slides down over the lower part to reveal the battery compartment. The battery is automatically bypassed when phantom power is present — a helpful touch. (Audio-Technica also offers the $399 AT822, a battery-operation-only stereo mic with unbalanced, ¼-inch connectors that otherwise is identical to the AT825.)

The AT825 employs two miniature “electret” condenser cardioid elements. The capsules are configured side by side in a fixed 110-degree angle.

The AT825 comes with a 16.5-foot shielded cable with Cannon connectors and color-coded sleeves on the preamp end (red indicates right) and a 5-pin XLRM-type connector on the mic end. Also included is a robust mic clip (with metal threads!), a foam windscreen, a protective vinyl zipper pouch, and even a AA battery, all neatly packaged in a sturdy cardboard box.

The least expensive mic of the bunch, the beyerdynamic MCE 82 (see Fig. 11) has a long, cylindrical body that gets progressively thicker from bottom to top. The basket's metal mesh is a bit flimsy — a hard push with the thumb will dent it. An external metal “rib” bisects the basket, both strengthening it and providing a helpful visual cue to left-right orientation. A rough-coat, matte gray finish provides a nice, nonslip feel, and the mic is light — less than half a pound — making it a good pick for lengthy handheld sessions. (Though too late for inclusion in this article, beyerdynamic recently released another affordable XY stereo mic, the $429 MCE 72, which has electret condenser elements angled at 120 degrees.)

Like the AT825, the MCE 82 can be powered externally by phantom power (8 to 52V) or internally by 1.5V AA battery. The bottom half of the mic body unscrews and can be slipped off to reveal the battery compartment. A recessed switch directly beneath the grille basket engages a 100 Hz highpass filter. A second, raised switch just below that controls battery powering via three positions: On, Off, and B.C. (battery check). When the switch is in B.C. position, a red LED between the switches lights to indicate full battery capacity — a handy reminder for fieldwork. (Careful, though — leaving the switch in B.C. position will quickly drain the battery.) Both switches are conveniently scored, and even the recessed switch can readily be moved by thumbnail. The raised switch is helpful when positioning or hand holding the mic in low- or no-light situations — you can feel the switch with your thumb, which makes for easy left-right orientation. There's also a white silk-screened L/R logo for orientation purposes.

The MCE 82 employs miniature electret condenser cardioid capsules angled 90 degrees apart in a side-by-side configuration. The elements are elastic-suspension mounted, which helps reduce handling noise.

The 8.8-foot shielded cable that comes with the MCE 82 features Neutrik connectors and has color-coded sleeves on the preamp end and a 5-pin XLRM-type connector on the mic end. No mic clip or windscreen is provided, but the mic does come packaged in a protective nylon zipper pouch.

The Crown SASS-P MKII (see Fig. 12) looks more like a vacuum cleaner attachment than a microphone. But bulky as the thing appears, its relatively light weight and clearly marked features make for easy setup and use.

The back of the SASS-P provides two standard XLR connectors, labeled A and B, and two 9V battery compartments that slide out like little drawers. Between the XLR connectors, which feature gold-plated posts, is a large knurled knob with two sets of two positions — a Flat and Cut position in Phantom-power mode, and a Flat and Cut position in Battery-power mode. Either Cut position activates a 100 Hz highpass filter. A mic-stand adapter with a large-handled tightening screw is permanently affixed to the bottom of the SASS-P.

Made of molded high-impact plastic, the body of this quasi-binaural microphone is shaped to form a barrier (baffle) between the two PZM elements and angled surfaces (boundaries) beneath them. A small mount suspends each PZM a few millimeters above its angled boundary. The two boundaries block sound from the rear, causing the omni PZMs to behave directionally for mid and high frequencies; in addition, pieces of foam rubber on either side of the baffle help to limit overlap of high frequencies between the two sides, further focusing directionality of the mic's high-frequency pickup.

Low-frequency pickup, however, remains omnidirectional. Thus, at high frequencies the SASS-P acts like a coincident pair, producing mostly intensity cues; at mid frequencies it acts like a near-coincident pair, producing both intensity and timing cues; and at low frequencies it acts like a closely spaced pair, producing primarily timing cues — pretty much how the human head's hearing works. However, the clever design of the SASS-P gets around the playback limitations of binaural capture, resulting in a well-focused, mono-compatible stereo image that translates nicely over both stereo speakers and headphones.

The SASS-P MKII comes with an enormous, high-quality, foam-lined flight case, underscoring the mic's intended professional use. The case has form-cut recesses for the mic and accessories, as well as extra spaces for cables, tapes, or what have you. Accessories include a foam-covered handgrip (which screws into the mic-stand adapter by way of an included brass adapter) and a custom windscreen that fits over the mic like a well-tailored jacket. The system is complete and clearly well thought out.

If you thought the SASS-P took the prize for most peculiar-looking stereo mic, check out the MBHO MBNM 622 E PZ (see Fig. 13), which our EM reviewer described as looking “like a cross between a sundial and a futuristic spacecraft.” (For the full review of the MBNM 622 E PZ, see the October 2001 issue of EM.)

As discussed earlier, the MBNM 622 modifies Jecklin's OSS technique by using boundary-layer rather than standard (pressure-gradient) condenser omni capsules. MBHO calls this “OSS II.” Boundary-layer mics require a boundary, so the designers ingeniously bisected the Jecklin disk in the horizontal plane with a second disk made of metal. This metal disk forms the boundary (or part of it) for the caps, which are mounted beneath bullet-shaped metal hoods 2.75 inches from the half Jecklin disk on either side. The bottom side of the disk, or base of the mic, is covered with a thin layer of felt to reduce vibration and slippage. The half Jecklin disk is a plastic semicircle sandwiched between two foam-rubber semicircles for a combined thickness of almost an inch. That puts the two caps approximately 6.5 inches apart.

The metal hoods suspend the boundary-layer caps a few millimeters above the metal boundary on one end; on the other end the hoods open out to standard XLR connectors, which feature gold-plated posts. A blue circle painted on the boundary plate passes directly beneath the two PZMs, visually delineating capsule placement with a sort of bull's-eye. Two small holes on one end of the plate allow the MBNM 622 to be mounted vertically on a wall. (I used pushpins, which worked fine.) The mic can also be positioned on the floor or any other large, flat surface (boundary). Typically, the bigger the boundary, the deeper the bass pickup. The sound is also affected by acoustic qualities of the boundary — glass and carpet boundaries, for example, yield quite different results.

The MBNM 622 comes with no accessories, but then, it doesn't really require anything more than a couple of mic cables and phantom power. Clearly, this mic is not designed for handheld use or for outdoor duties (thus no need for a windscreen). A carrying case might be useful, though — the MBNM 622 doesn't readily fit into a small pack or bag, and I'm not so sure I'd attempt carrying it by hand through airport security these days.

Of the six mics evaluated here, the Røde NT4 (see Fig. 14) is the new kid on the block; it came to market only last year. (For a full review of the Røde NT4, see the October 2002 issue of EM.)

The NT4 is a distinctive-looking mic. A cast-metal, rocker-shaped assemblage on top serves as a stereo bar, perfectly positioning the two capsules in the classic over-and-under XY configuration. The angle is fixed at 90 degrees. Unlike the other three XY mics in the test bunch, the NT4 employs “true” condenser elements (as opposed to electrets) with “full-size” ½-inch, gold-sputtered diaphragms. The capsules, which can be unscrewed from the mic for easy repair or replacement, derive their cardioid patterns by the traditional mechanical means of porting (note the two rows of narrow ports around the top of each cap).

The NT4's hefty, heavy-gauge metal body has a nice, satin nickel finish. Beneath the capsule assemblage, a rounded, internally threaded ring unscrews to provide access to the electronics inside. The bottom portion of the mic also unscrews and slips off to reveal the battery compartment — the NT4 can be powered by 9V battery as well as by 12V, 24V, or 48V phantom power.

An oval window tucks the NT4's on/off switch out of mishap's way. Just above the switch, a tiny red LED indicates battery condition. It lights briefly when the mic is switched on, indicating a good battery; if the LED stays on, the battery needs replacing.

Røde outfits the NT4 smartly with a tough, molded-plastic carrying case. Inside, cut-foam recesses fit the mic, a foam windscreen, a seriously beefy mic clip, and two shielded cables: an 11.5-inch, 5-pin XLRM to dual XLR, labeled “L” and “R” in big letters on the split end; and a 10-footer terminating in a stereo miniplug.

The smallest and lightest of the bunch, the Sennheiser MKE 44P (see Fig. 15) is both unobtrusive and robust, the kind of mic you can carry around worry-free in a baggy pocket and then deploy at a moment's notice, clandestinely if need be, and hold in your hand for long periods with only minimal fatigue. The grille basket on the 44P is virtually crushproof, and the body is made of a superhard, seemingly indestructible black composite plastic with a lustrous shine. The finish nicely offsets the white silk-screened lettering, L/R indicator logo, and other graphics (which include a pin-out diagram).

A threaded, knurled ring just beneath the basket unscrews and slips off, releasing the basket and capsule assembly, which can then be pulled off — carefully — and detached from the shaft. The bottom, interior portion of the assembly makes electrical contact with the shaft via five pins. Removing the assembly reveals the battery compartment, which takes a standard 1.5V AA. The mic can also be phantom powered, of course. Two red recessed switches on either side of the mic shaft can be moved only by pen or other pointed implement. One is the on/off switch (for battery operation; when phantom is supplied the mic is automatically on). The other is a 3-position switch offering low-cut filters at 150 Hz and 250 Hz, as well as a “flat” setting.

The 44P is nicely balanced for handheld applications; its oval-shaped shaft inclines the mic to lie “horizontally” in the hand, orienting the caps side to side rather than on the vertical plane. Still, left-right stereo orientation is not especially intuitive, due mainly to the round shape of the grille basket, which offers no visual clue. It doesn't help, either, that the on/off switch is located on the side of the mic rather than on top (top being defined as the center axis of the stereo field), or that your thumb tends to obscure the graphics — the 44P's only positive indicator of left-right orientation — when you hold the mic. Of course, none of this is an issue once you get used to it, but I do like being able to confirm mic orientation with a glance.

The 44P employs miniature electret condenser cardioid elements. The caps are configured in an over-and-under XY orientation with a fixed 90-degree angle.

The 44P comes in a sturdy, hard-plastic storage case with form-fit recesses for the mic and included windscreen and cable. The windscreen is noticeably high-quality — that kind of foam with the shiny, plush outer layer. The 3.75-inch shielded cable is also high-quality, featuring color-coded Neutrik connectors with gold-plated pins and, stamped on the cable, the Georg Neumann logo. However, the mic end (5-pin XLRM connector) splits immediately into two separate cables, which I find less convenient to work with than a single-piece cable.

As we get into the evaluations, bear in mind that all the manufacturers represented here make other, high-quality microphones, and some also make high-end stereo mics. Beyerdynamic, for example, produces the acclaimed MC 742 stereo mic ($4,999), which offers both variable-angle XY and M-S capabilities. So you should take this review as reflecting only on the specific models involved, and not on the capabilities of the manufacturers — for the most part, these mics do not represent their makers' best efforts.

The key question underlying this review — what can you get in a stereo mic for a thousand bucks or less? — brings together a fairly diverse lot. Beyond the main categories of XY and quasi-binaural, the mics also fall into other camps based on design and construction. For example, of the four XY mics, three — the Audio-Technica AT825, beyerdynamic MCE 82, and Sennheiser MKE 44P — employ miniature electret condenser elements fixed inside foam-lined grille baskets, whereas the fourth, the Røde NT4, sports ½-inch-diameter, separately detachable, “true” condenser caps. That, as well as other differences — materials, weight — predisposes the electret models to ENG, interview, and other mobile, on-the-scene — type recording duties. The NT4, on the other hand, though portable, is inclined more to studio and other indoor, mic-on-a-stand — type applications.

The other two mics, though quasi-binaural, are also distinct functionally (as well as sonically). The MBHO MBNM 622 E PZ, for example, is clearly not designed for handheld applications (though I suppose you could always mount it on a board and put a handle on the back), whereas the Crown SASS-P MKII is built to accommodate both studio and street (though you might feel a bit conspicuous packing the thing around town).

As for differences between XY and quasi-binaural mics, a big one, naturally, is stereo spread. In general, hard-panned signals from XY mics do not sound hard-panned. That's because XY patterns tend to spread elements around a semicircle stretching between approximately 9 o'clock and 3 o'clock pans. Thus, the speaker image appears more in front of rather than “around” the listener.

To get an equivalent width of soundstage from the quasi-binaural tracks, I found I had to pan them in as close as 10 and 2 o'clock. At the other extreme, when the quasi-binaural tracks were panned hard left and right, the stereo image seemed to wrap all the way around the speakers, pushing elements that had appeared in the room at 10 and 2 o'clock nearly into the “corners” at 8 and 4 o'clock. (That may not qualify as angular fidelity, but I rather liked having the extra width to work with.)

Though this is not a review of the Schoeps CMXY 4V, it was the benchmark mic for the comparisons, so some observations are in order. Not surprisingly, given its pedigree and price, the CMXY was the smoothest, most accurate, and best-sounding mic of the bunch. (“That's my guitar!” exclaimed Metzler upon first hearing the benchmark tracks.) The CMXY truly did serve to keep our perspectives in line.

But something else distinguished the Schoeps mic from the others: its exceptional ability to render depth of field. All the mics were good at creating breadth in the stereo field. Some produced a wider and others a narrower spread, and some were more focused than others; but they all placed elements in their proper positions — relatively so, at least — across the stereo soundstage. The CMXY stood head and shoulders above the rest, though, for its uncanny ability to reproduce a real sense of the depth of the soundstage. On the recordings of Klippert's band in the large room, the CMXY's stereo image was so clear and deep I felt I could throw rocks into it and hit the different performers — I knew just how far each was from the microphone. I could even sense height in the image, reflecting the fact that Klippert had stood above the mic and aimed his voice down as he sang. In contrast, the six test mics sounded more two-dimensional; they reproduced the width of the stereo field but not much sense of its depth.

The AT825 was a consistently good performer, typically producing a well-balanced and seemingly preequalized sound. The mic tends to highlight certain defining frequencies in the lows, high mids, and highs while downplaying typically troublesome low mids. This is borne out by the AT825's frequency-response plot, which shows a mostly flat line from 30 Hz to 20 kHz, with gently rising boosts centered around 50 and 80 Hz, between 5 and 6 kHz, and around 12 kHz — all common points of enhancement.

On drums, this resulted in nice tonal separation around the kit, with kick, toms, snare, and cymbals each well represented and occupying its own frequency niche. The AT825 captured a nicely balanced view of Klippert's large band, as well, with full and present bass, forward high mids, and crisp, sometimes edgy highs (“sizzly” was the word Klippert used). In both cases, the only thing conspicuously absent was low-mid warmth. Transient response was fast, if somewhat controlled sounding. Stereo spread and angular fidelity were very good — the 110-degree angle seemed to strike a good balance between the source sound and surrounding ambience.

On classical guitar, the AT825 captured a nice balance of treble and bass, but it favored the highs and lacked in fullness and low-mid warmth. Metzler pointed out that it made the nylon strings on her guitar sound “metallic” — an inexcusable sin for classical guitar.

Spectral accuracy might not be the AT825's claim to fame, but this mic seemed to always sound good — of the three XY electret models it consistently produced the most usable tracks. I came to think of it as a point-and-shoot, hard-to-go-wrong kind of mic; it tends toward a controlled, preequalized, final-mix sort of sound, making it a natural pick for fieldwork, concert taping, and general stereo applications. Though not really a studio-quality mic, the AT825 can be trusted to produce usable results in a variety of studio applications. It's reliable and easy to use, too.

Beyerdynamic makes clear right up front that the MCE 82 is not intended as a professional-quality studio microphone; according to the accompanying literature, the mic is “suitable for stereo recording in ENG/EFP [electronic film production] applications, home recording, reporting, film, video and interviewing.” Just the same, the 82 proved a decent performer — especially given its low price — typically producing a hotter, more forward sound than the AT825, but with less bass, fuller mids, and a tendency toward some harshness in the high mids. (The mic's frequency-response plot shows a relatively smooth, flat response from 100 Hz to 20 kHz, with the low end rolling off to approximately 5 dB down at 50 Hz. Two gentle boosts can be seen centered around 3 kHz and between 8 and 9 kHz.)

On the drums, the 82 produced an aggressive, somewhat over-the-top sound I rather liked; it helped make my dark and dead “drum room” sound brighter and more alive. (Of course, that same tendency might not be so welcome in a bright, reverberant space.) Dynamic response was impressive, as evidenced by strong attacks. Stereo spread is a bit tighter on this mic as compared with the AT825, and the sound is slightly drier. Frequency-wise, bass and low-mid content were under-represented (making the kick sound boxy), hats and snare sounded overly bright, and a small China cymbal translated very harshly. A touch of board EQ brought things into better balance — a cut at 12 kHz, boosts at 75 and 200 Hz.

The large-band tracks, too, sounded weak on lows and a bit harsh in the high mids, and again the sounds seemed more forward — closer to the listener — than with the other mics. Both Crews and La Velle noted that the “very high end” was subdued and that there wasn't as much capture of “airy” frequencies or room ambience with this mic.

The 82 couldn't always handle Metzler's hard strums, particularly the rasqueado (a flamenco term for a four-finger-rolling hit on the strings). Metzler, though, was relieved that the 82 didn't impart much “metallic” sound to her guitar's nylon strings. In that sense, the guitar sound was truer — less equalized sounding — than the AT825's. To my ear, though, the overall sound was ever so slightly out of focus.

Though not intended for critical studio work, the MCE 82 still performed admirably all around, proving it can pinch-hit if necessary in the studio. Fieldwork is its bent, however, and here it has the advantage of light weight, intuitive operation, and a nonreflective, nonslip surface. In addition, the MCE 82 produced the least amount of handling noise of the bunch.

Remarkably, the Crown SASS-P MKII's frequency-response plot shows a flat line from just above 50 Hz all the way to 15 kHz or so, after which it rolls off in a perfect arc to 5 dB down at 20 kHz. Given that perfection, I couldn't help but wonder whether Crown had taken a few liberties in “rounding off” the response figures. Then again, nothing sounds “off” with this mic spectrally — it does a great job of representing the whole frequency range in a balanced fashion, tending to turn out complete, if somewhat controlled-sounding, “finished mixes.” I didn't find myself reaching for EQ to fix things.

The SASS-P is not called Stereo Ambient Sampling System for nothing — this mic excels at capturing coherent snapshots of total-room sound. That makes it especially effective for recording large, spread-out groups (such as Klippert's band) but less so for recording single instruments at closer range (such as the classical guitar). On Klippert's group, the SASS-P's stereo picture, though unrealistically wide, was well balanced and very appealing, conveying a distinct sense of the space and the performers within it. Elements were spread noticeably further left and right than they had been in the room, creating a very wide, distinctly rectangular-shaped image. Klippert described this as “letterboxing” — a reference to the technique used to make motion pictures fit TV screens.

But for all its spacious width and delicious ambience, the SASS-P stereo image still felt somehow flat or two-dimensional, at least in comparison to the benchmark mic's. If I threw rocks into this image, it seemed, they'd tear through to the other side. (I've noticed a similar “flatness” with boundary-layer mics before, which leads me to wonder if this quality might be inherent in the PZM's method of pickup — rather than hit the caps directly, all signals must first reflect off the boundary. Does the flatness of the boundary impart a flatness to the sound?) Just the same, Klippert and I agreed that, of the six mics, the SASS-P was the overall best performer in this application.

As a drum overhead mic, the SASS-P produced a dramatic but again unrealistic sound with an overly wide stereo spread (easy enough to fix with panning, of course). As with the AT825, there was a sense of the individual elements having already been equalized. Transient response was good but not exciting, and overall the sound had a kind of controlled, “boxed in” quality — not rock 'n' roll by any means.

The literature that accompanies the SASS-P explains that putting the mic closer than 3 feet from a source typically results in a “hole in the middle” effect, so I listened carefully while positioning the mic on Metzler's guitar. But even well positioned, this was just not the right transducer for reproducing the warm, rich timbres of the classical guitar. Though it did nicely portray the spaciousness and character of the room, and had no problem handling transient spikes from the flamenco passages, the spectral picture was neither true nor particularly flattering to the instrument. Metzler described the sound as “flat,” “unnatural,” and “kind of metallic sounding.”

The MBHO MBNM 622 E PZ was a distinctive voice in the group, typically capturing a warmer, mellower sound than the other mics, with a strong emphasis on lows and low mids. The high end was generally subdued (which helped in rounding out some brittleness in Klippert's Kay archtop guitar) and sometimes slightly “brittle” sounding.

The 622's sonic presentation is influenced to a surprising degree by the boundary you attach it to. I experimented with several different boundaries, including a large pizza box, a piece of carpet, various sizes of plywood and hardwood boards, and several different walls and floors. The differing sounds I got indicate that the 622's response varies considerably depending on size and composition of the boundary, as well as on overall room sound.

For the large-room recording of Klippert's band, we started with the 622 positioned on the hardwood floor just in front of Klippert (guitar, vocals) and Noertker (upright bass). This produced warm, rich bass, only too much of it. We got a more balanced sound by mounting the mic on a piece of plywood held at around waist height by a sheet-music stand. This tamed the low end to manageable proportions and helped bring out the highs. The resulting tracks were still dark sounding compared with the others, and not particularly accurate — the marimba sounded like a piano in some passages. Still, I and others found the sound interesting and compelling — the 622's “take” on sounds is different from that of the other mics.

The 622 was especially flattering to lower-register instruments — drums, bass, the trombone, and bari sax. Stereo localization was generally good, but like the SASS-P the 622 distributes sounds very broadly around the stereo field, to the point that you can't quite trust their positions. In addition, high-frequency reflections in the large room sometimes “jumped over” the Jecklin disk, muddling timing cues and confounding localization. That happened during a loud passage played by the marimba, which was positioned off to one side of the room; the reflections evidently rivaled the direct sound, because the marimba sounded as if it were being played on both sides of the room.

In my studio I wanted to mount the 622 on the ceiling, directly over the drums, to take advantage of the uninterrupted surface. But that would have required drilling another hole or two in the mic base, so instead I tacked the mic to the wall behind the drum kit. My studio is acoustically treated and fairly dead, and the 622's sound “reflected” that fact: the resulting drum sound was very dark and punchy, with full low mids and fat lows. Actually, the sound was unlike anything I've gotten in my studio before — the 622 gave an incredible roundness and fullness to the kick and toms. Just for fun (on a different drum session), I put up the 622 as a supplementary mic to a pair of condensers I often use as overheads. After getting the drum balance between the condenser overheads and other mics (a dynamic each on kick and snare), I sneaked in the MBNM 622 tracks beneath the others. This worked wonders to fatten up the toms and kick drum, and to my surprise didn't introduce any phase weirdness (though maybe I just got lucky).

To record the classical guitar, I tacked the 622 to a 1-inch-thick, highly polished maple board (see Fig. 16). The mic was not very flattering in this application, however — the tracks were simply too dark and bassy to be of much use. Metzler, who proved a stickler for tone, put the 622 at the bottom of her list.

Though the 622 never produced a very balanced or natural sound in any of the tests, it usually did bring something new and different to the table. For that reason, I came to think of it more as an “effect” mic than as a documenter of reality. I can see the 622 being very useful as a supplemental, “creative” transducer in a variety of studio applications — it kept on surprising me with its unique voice.

Though not the favored performer in every case, the Røde NT4 proved the most versatile and consistently professional-sounding stereo mic throughout the range of applications — not surprising given its design, components, and higher cost. It also had the hottest output of the bunch.

The NT4's frequency-response plot shows a smooth but somewhat “bumpy flat” (plus or minus 3 dB) response from 20 Hz to 12 kHz, with gentle boosts between 100 and 200 Hz, others at 2 and 3.5 kHz, and the broadest between 5 and 8 kHz, with a slight dip between 4 and 5 kHz. A steeper peak gathers around 12 kHz, after which the high end rolls off smoothly to 8 dB down at 20 kHz. This mirrors the NT4's overall sound — bright, present, and immediately likable, but with a high-mid emphasis that can lead to harshness. In terms of frequency balance, the NT4 sounded similar to the beyerdynamic MCE 82, only warmer, fuller, and better focused.

Excepting the benchmark Schoeps mic, the NT4 was Metzler's and my favorite on her guitar. She described the sound as having a “nice balance” that was “almost there.” However, we both noted the mic's excesses in the 5 kHz region, and Metzler thought that overall the bass and treble ranges seemed not well integrated — “too separated” was how she put it.

The NT4 proved very good as a drum overhead, scoring highly in angular fidelity and realism. Transient response was excellent, resulting in punchy toms and definitive stick-to-cymbal sounds. Interestingly, I noticed a slight amount of natural compression on very hard hits. Fortunately, the natural compression sounded favorable — more an enhancement than a liability. Frequency-wise, the NT4 came up a tad short on bass and low mids (though not as short as the beyerdynamic mic), and the high mids were consistently forward and sometimes brash — the small China cymbal was still a bit much. Generally, the balance of sounds benefited from minor cuts at 5 and 12 kHz and boosts at 75, 100, and/or 200 Hz.

The NT4 produced a nice, believable stereo image of Klippert's band. Crews was impressed by the clarity of the low end; Klippert, though, felt that the bass got a bit “lost” under the horns, which were well represented. La Velle remarked that the NT4 had “more body” than the other XY mics. Also, the NT4 evidenced better SPL handling than the other XYs — it didn't flinch or falter, as some of the others did, during a passage in which Klippert's voice got extremely loud. Although Klippert preferred the SASS-P and MBNM 622 on his group, the NT4 was his primary pick of the XY mics.

The Sennheiser MKE 44P is a curious case, and I'm still trying to figure out what its designers had in mind. The main issue is the mic's low-end response, which unfortunately is very weak. Even according to the frequency-response plot, the 44P's low end starts rolling off at around 300 Hz, dropping to 12 dB down at 40 Hz — and that's with neither of the two low-cut filters engaged. My question is, what's the point of having not just one but two low-cut filters on a mic with essentially no lows?

Needless to say, I found no reason to engage either filter on the 44P. Still, I regret to report, all the tracks I recorded with the mic registered precisely the deficiency revealed by the mic's response plot. That is, they all lacked bass and sounded thin — whether compared to the other mics or not. That's the problem when there aren't enough lows: even if the high-end response is good, the overall sound will be thin. High and low exist on a continuum, after all; what we hear first and foremost from a source (or mic) is its “overall” sound. It takes trained ears to mentally break up that overall sound into distinct bands and then analyze them individually. Of course, that's just what musicians and recording engineers do.

Metzler knew her guitar sound intimately, and she rightly pointed out that the 44P was quite true to the sound of the nylon strings in the high-mid range; unlike several of the other mics, it didn't impart a metallic quality to them. In addition, the 44P's transient response was very good, and the mic never overloaded on forceful passages. It was for these reasons, I suspect, that Metzler chose the 44P as her second favorite mic of the test bunch, after the Røde NT4. She just couldn't deal with any amount of metallic sound imparted to her beloved guitar. To my ear, however, the missing lows seriously compromised the guitar's overall sound.

I detected a similar high-mid “band” of accuracy on the drums. And again, the 44P's transient response was very accurate — in my notes I scribbled “dynamic truth.” But unfortunately, none of that saved the mic's overall sound. What are drums, after all, without lows? Moreover, in contrast to those truthful high mids, the 44P boosts other, higher frequencies (10 and 15 kHz) to sometimes painful effect — I could hardly listen to the drum tracks from start to finish, even though I had played dark, hand-hammered cymbals using wood-tip sticks.

The 44P's performance in the big room was better, evidently because there was so much more going on frequency-wise that the rolled-off low end wasn't so apparent. Still, the sound was nothing to salivate over, and all it took was a few seconds listening to any of the other tracks to be reminded of what was missing. I really do wonder what the designers of this microphone were thinking. The only applications I can think of that the 44P would be suitable for are interviews, boardroom meetings, and some types of ENG. Then again, it might prove ideal for recording outdoors on windy days — the rolled-off bass response should make it much less susceptible to wind noise.

To its credit, the MKE 44P is a sleek, lightweight, stealthy little mic that nicely fits the hand and keeps a low profile. If Sennheiser were to firm up its bass response and tone down the piercing highs, it could be a contender.

The testing done for this review was not meant to be conclusive; more time spent with each mic would undoubtedly have yielded more insights into its pros and cons. But given the likely range of uses EM readers would have for stereo mics, the tests provide fair and helpful analyses.

If anything is clear from the results, it's that the pickings are good among affordable stereo mics. Whether your focus is studio recording, fieldwork, concert taping, or some combination of the three, a thousand bucks opens the door to a diverse range of stereo mics, one of which should fit your needs.

I hope the information in this article will help guide prospective stereo-mic buyers through the maze of stereo-miking considerations. I also hope it turns more people on to the simple pleasures and rewards of recording with stereo microphones. There's magic in the act of human hearing, and by mechanically embodying and extending that magic, stereo mics do more than the sum of their parts might suggest.