Stereophile: MartinLogan Aerius loudspeaker

MartinLogan Aerius loudspeaker

John Atkinson, October, 1993

Combine an electrostatic panel to reproduce music's midrange and treble with a moving-coil woofer for the bass foundation. For decades, this has seemed the ideal way of designing a loudspeaker: Each type of drive-unit is used in the frequency region for which its performance is optimized. The resulting hybrid should sing like an angel.
Yet such hybrids often turn out to be mules, in my experience, their sounds remaining resolutely earthbound. To integrate drivers with very diverse radiation patterns—the omnidirectional woofer and the bipolar panel—sets the speaker engineer a pretty problem, to say the least, while the choice of a crossover frequency that does the least damage to the music is also far from trivial.
When I read Sam Tellig's enthusiastic prose (footnote 1) about MartinLogan's new baby, the Aerius, I was intrigued. Could the Kansas company have successfully blended thoroughbred electrostat and dynamic donkey to get the best of both breeds?
Aerius
Like all MartinLogan designs, the Aerius features elegance in spades. The 55"-tall speaker occupies no more floorspace than a minimonitor on a stand. The 8" woofer is mounted in a sealed enclosure behind a downward continuation of the front stator. The enclosure is reinforced by two vertical triangular braces at the side-panel midpoints and is filled with synthetic fiber. Black-painted side panels rise from the woofer bin, tapering to the top of the speaker, while optional oak trim strips ($300) flank the panel.
The narrow, curved, electrostatic panel is sandwiched between black-painted, rigid perforated stators. The panel, coated with a very thin, transparent, conductive film to ensure even charge distribution over its surface, is divided into 14 horizontal sections, each between 2" and 3" high. Two clear plastic damping panels, each a few square inches in area and made from Lexan, are fixed to the rear stator. These are said to provide some response shaping by "pressure compensation."
Electrical connection is via a pair of 5-way binding posts on the rear of the woofer enclosure; a second pair for individual drive of the two drive-units is a $200 option. The crossover circuitry is mounted on the inside of the woofer enclosure's rear panel. High- and low-pass slopes are specified at 12dB/octave, though some electrical response shaping appears to be employed. Air-cored coils and a mixture of plastic-film and electrolytic caps, along with a handful of resistors, are used. The internal wiring appears to be Monster.
The crossover frequency is set at 500Hz—in the heart of the midrange. This is interesting; it allows nearly all musical fundamentals of bass, tenor, and alto instruments to be reproduced by the moving-coil unit, with the panel left free to handle their harmonics. Melody instruments—flutes, oboes, violins, the right-hand region of the piano, and the higher tessitura of the female voice—are reproduced exclusively by the electrostatic panel.
On the other hand, taking a relatively small panel down to 500Hz might be asking for too much excursion in its bottom octave, resulting in extra distortion products in the 1-3kHz region, where the ear is very sensitive. Note that the operative word here is "might," not "will." The proof will be in the listening.

Footnote 1: "Sam's Space," June 1993, Vol.16 No.6, pp.75-81.—JA

Like all electrostatic speakers, the Aerius needs to be energized from a wall AC outlet. An AC lead plugs into an IEC receptacle on the speaker's rear, all the power-supply circuitry being internal so that the speaker can be manufactured to be country-specific in terms of line frequency and voltage. Given that makers of laptop computers use universal power supplies that will work anywhere in the world without adaptors, this would seem to be a backward step. However, MartinLogan does this, I believe, to cope with the increasing amount of "gray-market" sales afflicting the international High End.
Listening context
Power amplifiers were either a pair of Mark Levinson No.20.6 monoblocks or a YBA 2 HC, while the preamplifier was either the Melos SHA headphone amplifier used as a line-stage or the YBA 2. A Mod Squad Phono Drive EPS amplified LP signals from a Linn Sondek/Cirkus/Trampolin/Lingo/Ekos/Arkiv setup sitting on an ArchiDee table. Digital source was variously a Theta DS Pro Generation III, a Counterpoint DA-10 with UltraAnalog DAC module, or a Meridian 263 Delta-Sigma, driven by a Mark Levinson No.31 transport via AudioQuest Pro 2 ST-optical and Digital Pro coaxial links, respectively. Interconnects included 1m lengths of Siltech 4/80 between digital processor and preamps, and 15' lengths of AudioQuest Lapis between pre- and power amplifiers. The speaker cable was a 5' bi-wired set of AudioQuest Sterling (the Aerius review samples were supplied with the bi-wiring option). All source components and preamps were plugged into a Power Wedge 116, itself plugged into one of my listening room's two specially installed AC circuits. (The Aeriuses' electrostatic power supplies were plugged into separate circuits.)
My dedicated room measures approximately 19' by 16.5', with a 9' ceiling. The wall behind the speakers is faced with books and LPs, while further book- and CD shelves occupy the positions on the sidewalls where the first reflection from each speaker would occur. The room is carpeted, and there are patches of Sonex foam on the decking'n'viga ceiling to damp the first reflections of the sound. The other wall has RPG Abffusors and ASC Tube Traps behind the listening seat to absorb and diffuse what would otherwise be early rear-wall reflections of the sound that might blur the stereo imaging precision. More Tube Traps are used in the room corners to even out the room's bass resonances, the result being a relatively uniform reverberation time of around 200ms from the upper bass to the mid-treble, falling to 150ms above 10kHz.
In the excellent and comprehensive manual provided for the Aerius, (footnote 2) MartinLogan stresses the fact that the sound of the speakers will change for at least the first 30 hours, due to the need of the woofer's butyl surround to be broken-in. The Aeriuses were run in on high-level pink noise for approximately 72 hours before any serious listening was performed, therefore. Each Logan was used with its three carpet-piercing spikes—two at the front, one at the back—making contact with the tile-on-concrete floor beneath the carpet. MartinLogan stresses that the Aerius does need some boundary reinforcement to achieve its correct tonal balance. I ended up with the center of each cabinet's rear panel 2' from the front of the record cabinets behind it. Each speaker was almost, but not quite, toed-in to the listening seat.
Sound
"Seamless." That's the word. Right from the get-go, I was struck by the unity of the Aerius's sound, the integration between woofer and panel. Unlike its rather larger brother, the Sequel II, which I always had trouble with in the crossover region, the Aerius is excellent in this respect. In fact, the transition from moving-coil to electrostatic driver is so smooth that when my son, recently embarked upon his odyssey through the "terrible twos," pulled out the woofer connection to one of the speakers, it took me most of an evening to realize that the sound had lost 3dB of its bass oomph!
With my ears 36" from the ground, pink noise sounded very smooth (though there was a touch of boost in the presence region). Moving up a few inches in my chair added excessive treble "bite" to the sound, while standing drastically cut the highs. Interestingly, with dual-mono pink noise, while only minor vertical Venetian blinding was noticeable as I moved from side to side—the MartinLogan CLS was severely afflicted by this problem—moving even a few inches forward in my seat changed the midrange character significantly, the main effect being to change a trace of "eee" coloration. This could only be detected while I was moving, however, which suggests that it is due to position-dependent interference effects—minor comb-filtering and the like—which are trivial.
Given my enthusiasm for the seamless nature of the Aerius's sound, it would seem unfair to break its performance down into the traditional frequency regions. But hey—it's hard for a reviewer to break audiophile habits.
Starting at the very top, the highest treble octave seemed shelved-down, the sound lacking top-octave air. This bothered Sam Tellig more than it did me; I only really noticed it when I listened for it, and with recordings that themselves lacked energy in this region. I suspect that whether or not it becomes a significant factor in the Aerius's balance will depend on how large a room its owner has, and how well-damped the walls and furnishings are.

Footnote 2: It includes an article on the history of electrostatic speakers (with due mention made of the pioneering work of Quad's Peter Walker), excellent hints on setup and room acoustics, and a list of LPs and CDs MartinLogan has found to offer exceptional sound quality.—JA

I suspected from my experience with pink noise that this reticent top octave is also exacerbated by some peakiness lower in frequency, in the mid-treble. Though it was not generally sibilant, the Aerius did indeed add some slight emphasis to tape and microphone hiss. Close-miked soprano voice also took on a hard edge as it got higher in level. And the trumpets in my Elgar Dream of Gerontius recording on the second Stereophile Test CD had a little too much brassy blattiness. Changing from the YBA to the Mark Levinson No.20.6 monoblocks softened the mid-treble. This change also added an octave to the speaker's subjective bass response, though the lack of air in the Aerius's top octave was somewhat exacerbated.
Even with the Levinson, however, the Aerius 2 was unkind to poor recordings: the ridiculously bright live Queen concert from 1986 (Live at Wembley, Hollywood HR-61104-2, which I recently bought to remind myself of the late Freddie Mercury's dominance of the 1985 "Live Aid" concert) literally screamed at me at anything other than quiet levels.
But such recordings were the exception. The sheer amount of unforced, musically natural detail I could hear through these speakers was astonishingly addictive. Most of my time with the MartinLogans was spent pulling LP after LP and CD after CD from the rack, following a particular musical thread. Chrissie Hynde's superbly ornamented vocals on "Talk of the Town" from the Pretenders' The Singles LP (Sire 25664-1), for example, led me to Barbara Bonney's vanilla'n'cream soprano in the Nikolaus Harnoncourt CD performance of Mozart's Exsultate, jubilate (Teldec/Das Alte Werke 2292-44180-2), which in turn led to that old favorite, the Klemperer Brahms Ein deutsches Requiem (EMI Classics CDC 7 47238 2), jumping to the lushly mysterious and moving Symphony 3 by Henryk Görecki (Elektra Nonesuch 79282-2), and finishing up with Stevie Winwood's eponymous 1977 album (Island ILPS 9494).
The Aerius's midrange was sweet and uncolored, but more importantly, voices and instruments sounded correct. I kept returning to vocal recordings: James Taylor's New Moon Shine (Columbia 46038) and Eric Clapton's Unplugged—the German LP pressing that Guy Lemcoe was kind enough to search out for me.
When he wrote about the MartinLogans, Mr. T kept returning to the phrase "truth of timbre." This they have, coupled with a surprising degree of "realness" to both sound and image—the much-overused "palpable presence."
Changing to analog source raised this aspect of the speakers' sound by more than the expected degree. Whether this was due to the second sample of the Linn Arkiv cartridge that had just been installed, to the new Cirkus bearing upgrade, or to the fact that the speakers' intrinsic performance is to such a high standard that the front-end had been the limiting factor, I don't know. (Sometimes you can't limit the variables in your system to just the subject of the review.)
The Aerius offers surprising bass weight from a speaker with a single 8" woofer. From my brief exposure to the Quest Z, I agree with Dick Olsher that its intrinsic balance tends toward leanness. Yet via the smaller Aeriuses, the warble tones on the Stereophile Test CDs reproduced with full measure down to the 32Hz band. At the beginning of Klemperer's Brahms Requiem, the downward octave jump in the double basses is easily heard, not obscured by any doubling. And again on James Taylor's 1991 New Moon Shine LP, when, at the end of each line in the chorus of "Down in the Hole," bass guitarist Jimmy Johnson drops down an octave to a dominant D to bring the harmony full circle, back to G for the start of the next line, the low note—36.7Hz—is given full measure (footnote 3).
Yet MartinLogan has not obtained this powerful bass performance by tuning the Aerius's woofer to be some sort of boom machine. Throughout the mid- and upper bass the response was exceptionally even, no single note sticking out any more than another (within the vagaries of my room acoustics, which exaggerate the 60-65Hz region no matter what I do). The Aerius's woofer is "fast," by which I mean that it does not hang over. It refrains from adding extra mud where it is not required. The bass guitar and kick drum in "Hey Nineteen," from Steely Dan's pinnacle Gaucho album, were as tight as a nut, two instruments of similar pitch acting as one—yet the speaker's low-frequency clarity allowed them to maintain their individual identities. (This track's lyric rings a responsive chord in this 45-year-old's consciousness—"Hey Nineteen, that's 'Retha Franklin, ([but] she don't remember the Queen of Soul)."
Within strict loudness limitations, the Aerius offers excellent dynamic contrast. It does reach a point where turning up the volume doesn't seem to produce much extra loudness. However, the speaker didn't sound strained until the spl was above 100dB in my room. It also has an excellent sense of pace. Willie Weeks's loping Fender-bass octaves drove Stevie Winwood's effortless vocal and strutting guitar solo in "Hold On" from his seminal 1977 album just fine, while the bass guitar and drums on Michael Ruff's new Speaking in Melodies CD (Sheffield Lab CD-35) smoked. (This album features probably the best-recorded rock drums I have ever heard, coupled with inspired use of, yes, a trombone (footnote 4). Robert Harley's drum recording on our Test CD 2 also fared well, the image of the kit being nicely set back within the subtle ambience of David Manley's studio.

Footnote 3: The addition of low-C capability to the orchestral double bass and a fifth, low B-string to the electric bass guitar are the most important developments in subterranean music making, at least as far as this bass player sees it.—JA
Footnote 4: Musician Humor: A frog and a trombone player pass each other in the street one Saturday night. At least the frog is probably on her way to a gig.—JA

Which brings me to the Aeriuses' soundstaging. Recorded ambience, indeed, seemed more obvious, with longer decays, than I had expected from listening to the same tracks on conventional speakers. The differences between the various artificial reverberation programs Corey Greenberg had used on his "Eden" track on Test CD 2 seemed larger than I had thought. And instruments didn't move forward or backward in the stage as they were played louder or softer.
Soundstaging was generally well-focused, with no tendency for images to pull to the center. But I didn't get much of a sense of the stage extending beyond the image positions when appropriate. In track 10 of Test CD 2, for example, where I had recorded Larry Archibald mapping out the soundstage in the Santa Barbara church where we recorded our Brahms Intermezzo album, the image of Larry's handclaps should start well to the left of the left speaker and finish well to the right of the right speaker. Instead, the image of Stereophile's main man moved along a U-shaped path behind and between the loudspeaker positions. But again on this track, there was excellent depth and a feeling for the church's acoustic as Larry walked from the back of the nave up to the microphone.
In the ring with the Thiel
The natural comparison for the Aerius is the Thiel CS2 2, which I reviewed at the beginning of the year (footnote 5) and now suffers a slight price disadvantage since its recent price hike to $2750/pair. The beautifully engineered 2 2 is ST's other favorite speaker (when he can be torn away from his Advent Cheapskate Specials, that is).
Listening to the two pairs head to head, the Thiel's midbass is a bit fatter than that of the Aerius, which adds a pleasing bloom to the sound of bass guitar. It sounds rather "boxy" beside the Logan, however, with a more congested lower midrange (footnote 6), and the sound doesn't float quite so clear of the speakers. Though the 2 2's mid-treble is slightly more sibilant, its top octaves sound more airy and spacious. Its imaging is even better focused, and the soundstage thrown by a pair of Thiels is as deep and wide as that produced by the MartinLogans. Both are champions at ambience retrieval; both excel at lunging for ultimate transparency.
The Aerius's midrange character is more ethereal, the Thiel's more robust. Neither sounds lean, though the 2 2's balance is richer overall. Both make demands on the upstream equipment they'll be partnered with; neither is a rock-the-house-down party speaker. Asked to choose between them...I wouldn't.
Conclusion
Other than the humongous Statement system, I've heard all of the MartinLogan range under more or less familiar circumstances. It may surprise some of you that, notwithstanding Dick Olsher's enthusiasm for the Quest Z elsewhere in this issue, I feel the Aerius to be the most successful speaker design yet to come from this Kansas manufacturer. Within the limitations to be expected from its cost and size—fundamentally a relatively limited loudness capability—it is a superbly musical speaker with a surprisingly extended bass, an uncolored, detailed midrange, a transparent treble, excellent dynamic contrast, and superb soundstaging. It does lack HF extension, and the lower mids are a little thickened in absolute terms. However, the Aerius has the capability—rare in affordable speakers—of being able to grow as your system grows. It certainly wasn't embarrassed by being used with the mondo-expensive Levinson monoblocks, for example. And the improvement offered by Linn's Cirkus upgrade for the LP12 was easily perceived.
Along with the Thiel CS2 2 (and, at a slightly lower performance level, the Vandersteen 2Ce), the MartinLogan Aerius makes it harder for high-ticket loudspeakers to justify their cost of admission. Enthusiastically recommended. This mule can fly. Or at least sing!

Footnote 5: Stereophile, Vol.16 No.1, p.238.—JA
Footnote 6: This is at least partly due to a strong enclosure resonance affecting the speaker's front baffle. In a forthcoming "Follow-Up" I will investigate the effect Combak Harmonix dots have on this aspect of the Thiel's performance.—JA

Sidebar 1: Specifications
Description: Two-way, floorstanding loudspeaker. Drive-units: 37.5" H by 7.5" W electrostatic panel, 8" (203mm) pulp-cone woofer in sealed enclosure. Crossover frequency: 500Hz. Crossover slopes: second-order, 12dB/octave. Frequency response: 40Hz-20kHz !X3dB. Dispersion: 30 degrees horizontal, 38" line source vertical. Sensitivity: 89dB/W/m (2.83V). Nominal impedance: 4 ohms. Minimum impedance: greater than 2 ohms at 20kHz. Amplifier requirements: 60-200W.
Dimensions: 55.5" (1410mm) H by 10.5" (267mm) W by 13.5" (343mm) D. Weight: 55 lbs (25kg) each.
Serial numbers of units tested: AEEA008/9.
Price: $1995/pair in black finish; bi-wiring option adds $200/pair; light oak trim adds $300/pair. Approximate number of dealers: 65.
Manufacturer: MartinLogan Ltd., P.O. Box 707, Lawrence, KS 66044. Tel: (913) 749-0133. Fax: (913) 749-5320. Website: www.martinlogan.com .

Sidebar 2: If You Can't Wait...
To sum up the MartinLogan's pluses: looks the business; sets the soul of music free; works well in small rooms near the rear wall; superb integration between the disparate drive-units; seamless overall balance; transparent, timbre-true midrange; well-extended, well-defined lows; excellent depth and retrieval of recorded ambience; costs less than Wilson WATTs/Puppies; not for Red Hot Chili Peppers fans.
Minuses: mid-treble has tendency to glare at high levels; high treble shelved down; tape hiss accentuated; won't go very loud without strain; unkind to overcooked recordings; image focus good rather than excellent; costs more than a pair of PSB Alphas; might lead you to upgrade your front-end and amplification to the detriment of your disposable income; not for Red Hot Chili Peppers fans.—John Atkinson

Sidebar 3: Measurements
As can be seen from its impedance plot (fig.1), the Aerius is considerably easier to drive than its larger sibling, the Quest Z (reviewed by Dick Olsher elsewhere in this issue). Yes, it does drop to 1.7 ohms above the audio band—the cursor position—but the phase angle in this region is close to 0 degrees; if you don't use an MC cartridge with a vicious tip-mass resonance or you only listen to CD, the amplifier won't be asked to do any work in this region. And lower in frequency, where the bulk of musical energy lies, the Aerius's impedance magnitude stays above 5 ohms all the way to 4kHz. The sealed-box woofer tuning is revealed by the 25 ohm peak at 46Hz. The wrinkles in the solid magnitude plot at 230Hz (major) and 300Hz (minor) indicate some sort of cabinet resonant problems at these frequencies.

Fig.1 MartinLogan Aerius, electrical impedance (solid) and phase (dashed). (2 ohms/vertical div.)

The Aerius is reasonably sensitive for a panel speaker, at an estimated 85dB/W/m (B-weighted), though this is below specification. However, this is the same sensitivity as I measured for the Quest Z, despite the more expensive speaker's significantly larger panel area: 668 in� vs 281 in�.
The Aerius's impulse response on an axis 36" from the ground, measured with a B&K 4006 microphone calibrated to be flat on-axis and the DRA Labs MLSSA system, is shown in fig.2. Its shape implies good time coherence, as confirmed by the step response (fig.3), but the tail of the impulse is disturbed by high-frequency ringing. The large hump centered at 4.8ms in the step response is due to the panel's lower-frequency ringing being reinforced by the woofer output, though this is lower in level. Both drive-units are connected with the same acoustic polarity; reversing the woofer's electrical polarity eliminates the hump in fig.3, implying that the outputs of the two drive-units cancel in the crossover region when connected in this manner. (Inverting the woofer polarity does indeed result in a large measured suckout covering the entire 200Hz-1kHz region, showing that MartinLogan has correctly implemented the Aerius's crossover.)

Fig.2 MartinLogan Aerius, on-axis impulse response at 45" (5ms time window, 30kHz bandwidth).

Fig.3 MartinLogan Aerius, on-axis step response at 45" (5ms time window, 30kHz bandwidth).

Turning to the frequency domain, fig.4 shows the individual responses of the woofer and panel, the trace for each a composite made up from the upper-frequency quasi-anechoic response at 45" spliced to the lower-frequency nearfield response. The woofer can be seen to roll off smoothly above 500Hz, the 12dB/octave rollout slope being broken up by a couple of cone modes. These are well down in level, however. The woofer's output is flat down to 55Hz, rolling off to -6dB at 37Hz. The gentle 12dB/octave slope, the result of the sealed-box alignment, means that the Aerius should give good bass extension in a typical room.

Fig.4 MartinLogan Aerius, acoustic crossover on tweeter axis at 45", corrected for microphone response, with the nearfield woofer response plotted below 300Hz.
The panel's output is less easy to interpret. The top octave is shelved down, while the mid- and low-treble are broken up by peaks and notches which may be due to resonances, or to interference effects specific to the particular microphone position chosen. The peak at 420Hz and the suckout just above that frequency might be at least partly due to the nearfield microphone technique. The panel can be seen to roll out very quickly below 300Hz, however.
Looking at how these two responses integrate at a 45" distance, fig.5 shows the quasi-anechoic response averaged across a 30-degree horizontal window to minimize the effect of mike-position-dependent interference effects. It gently and smoothly slopes down by some 4dB through the midrange and low treble. The mid-treble still looks messy, however, and the top octave is shelved down. The latter, however, will be at least partly due to proximity effect, as explained in my Quest Z measurements elsewhere in this issue. Nevertheless, the response in my listening room (fig.6) was still a little lacking above 10kHz, which correlates with my feeling that the speaker's sound lacked top-octave air. Overall, however, the in-room response of the Aeriuses is very smooth, with a very slight prominence of low-treble energy which might explain the way the speaker slightly emphasized tape hiss. (Ignore the peaks and dips in fig.6, which are room effects that have not been minimized by the spatial averaging used to derive this graph.)

Fig.5 MartinLogan Aerius, anechoic response on-axis at 45", averaged across 30 degrees horizontal window and corrected for microphone response, with the nearfield woofer response plotted below 300Hz.

Fig.6 MartinLogan Aerius, 1/3-octave, spatially averaged in-room response.

Moving on to how the speaker's balance changes as a listener moves to its side (fig.7, only differences shown), the Aerius is basically omnidirectional within the woofer's passband. The bipolar panel output, however, increasingly shelves down with off-axis angle, though not without some peakiness appearing in the high treble. The flattest high-frequency balance is to be obtained with the speakers toed-in to the listening seat. Vertically (not shown), the speaker's output is relatively uniform as long as your ears are between 33" and 40" from the floor. The treble gets very peaky between 40" and the top of the panel, however, while moving above or below the panel results in severely depressed highs—don't listen to the Aerius standing up!

Fig.7 MartinLogan Aerius, lateral response family at 45", normalized to response on tweeter axis, from back to front: differences in response 90 degrees-15 degrees off-axis, reference response, differences in response 15 degrees-90 degrees off-axis.

Returning to the time domain, fig.8 is the cumulative spectral-decay, or "waterfall," plot calculated from the Aerius's impulse response (fig.2). This appears much cleaner in the midrange than that of the Quest Z, presumably because the well-behaved dynamic woofer is taken two octaves higher in frequency. The high treble is similarly hashy, however, while the low- and mid-trebles feature a number of resonant ridges.

Fig.8 MartinLogan Aerius, cumulative spectral-decay plot at 45" (0.15ms risetime).

As I explain in the Quest Z review, this kind of waterfall behavior is typical of panel speakers in that a large, uniformly driven, low-mass diaphragm appears to exhibit chaotic behavior. While the diaphragm's average position responds uniformly to the driving force, there are small areas which move more than the average and others that move less. I believe that such behavior is often mistaken by audiophiles for "clarity," "fast transients," and "transparency." I also believe that it tends to accentuate/exaggerate detail by surrounding each transient edge of the music with a little halo of hash. However, whether its presence outweighs the virtues of any particular panel speaker is an individual decision. In the case of the Aerius, I felt that the sound developed glare or hardness only at very high playback levels. At more practical levels, I reveled in the wealth of unforced detail presented. I also failed to hear any significant distortion or overload problems in the bottom two octaves of the panel's passband, unless the playback level was ridiculously loud.
Harking back to the Aerius's impedance plot (fig.1), a wrinkle at approximately 230Hz implied some sort of cabinet problem. Using a simple PVDF accelerometer (footnote 7), I investigated the speaker's structural resonances. The woofer bin does indeed have a strong resonant mode near this frequency, as can be seen from the waterfall plot calculated from the impulse response of the accelerometer fastened to the center of the back panel (fig.9). This mode could also be found on the side panels and, to a lesser degree, on the curved surface of the front stator. The latter also featured a mild mode at 305Hz, again correlating with the mild wrinkle in the impedance trace. Given the high level of the 227Hz mode, I'm surprised that I didn't hear much lower-midrange congestion. Turning the speakers around so that their backs faced the listening seat made this problem very audible. But with the speakers set up correctly, only occasionally was I reminded that there was something untoward going on in this region.—John Atkinson

Fig.9 MartinLogan Aerius, cumulative spectral-decay plot calculated from the output of an accelerometer fastened to the woofer cabinet back panel. (MLS driving voltage to speaker, 7.55V; measurement bandwidth, 2kHz.)

Footnote 7: See Stereophile, June 1992, p.205; and September 1992, p.162.—JA

Sidebar 4: Sam Tellig on the Aerius i
(From July 1997, Vo.20 No.7)
"Can we have these?"
Such was my wife Marina's reaction to the MartinLogan Aerius i loudspeakers as soon as I'd set them up in the living room. I had to agree: they do look elegant, especially with black side panels—almost like Japanese screens. The speakers stand nearly 5' high (55½"), and 10½" wide by 12½" deep. Small footprint.
"Of course," Marina hastened to add, "that's if you think the speakers sound good."
"What do you think?"
"I think they sound as good as they look."
No question about it. MartinLogan speakers appeal powerfully to the Russian soul—as witnessed by the fact that so many are now sold in the Motherland.
That night, Slava and Elena visited for dinner.
"Vat do zhees speakers cost?" Slava inquired in his thick Muscovite accent.
"Roughly two thousand a pair."
I played the Shostakovich Symphony 5 with Slava Rostropovich conducting the National Symphony Orchestra (Teldec 112226).
"Verrry gud. Vere do I buy zhees speakers?"
Slava doesn't act quite so quickly, and Elena is not sure where to place the speakers in their living room, but I think they'll buy a pair. So might Lev and Sopha, and Mark and Natasha. And perhaps even Radislav, Marina's ex-husband.
On changing wives
"I had a drrream," Marina said to me the other morning.
"So what else is new?"
"Radislav came over to veezit and saw your seestem. He said to you, 'I must have zhees speakers. I will geeve you whatever you want.' "
"I already have his former wife," I said.
Marina ignored my comment. "He vould die if he saw and heard those MartinLogan speakers. He is leestening to a mini-seestem. Radislav always has to have the best."
"Not any more. He's got a new wife now: Myshka." (Myshka means little mouse.)
I paused.
"Changing wives is like changing interconnects or speaker cables. You don't necessarily upgrade."
"Vhat's that supposed to mean?"
Marina was now wide awake. I was feverishly recording the conversation—and the dream.
"You're not going to put this stuff in your column—are you?
I laughed my evil laugh.
"Radislav readsStereophile!" Marina cried in panic. "Plochoi!" Bad boy!
I told Gayle Sanders of MartinLogan about Marina's reaction to the Aerius i...and Radislav's, too.
"We're selling a lot of speakers in Russia," Gayle told me.

The speakers are popular in France, too, from what I see by the ads in French hi-fi magazines. I should point out that the Aerius i, which retails for $2000 in the US, goes for around $4000 in France—thanks, in part, to France's killer 21% VAT. And in Britain, a pair of Aerius i speakers will set you back $l2300 (about $3750).
In the US, you get a bargain—$1995 for the basic speaker with black oak rails, single-wire. Light oak trim, please fork over $100 more. Bi-wire, gotcha for another $100. I'm not sure the oak trim is worth a C-note—the black is fine. But bi-wire is probably worth the money because you'll have the option to bi-amp, if you like. You may want to.
See-through sound
On top, you have the electrostatic panels—those see-through screens. Talk about transparency—you can see through the sound. Underneath, there's an 8" sealed-enclosure woofer. If you know the original Aerius, the Aerius i has the same dimensions but the cabinet has been redesigned, streamlined as all get-out. This is one of the most visually elegant loudspeakers in the world.
The original Aerius sounded so transparent that in the June 1993 Stereophile (footnote 1), I proclaimed it something of a steal. The new Aerius i is even more of a steal. MartinLogan has substantially improved the speaker, but has not raised the price. The bass driver is new. Designer Gayle Sanders says it offers better power handling and achieves a better blend with the electrostatic panel. Yes, it does. The crossover has also been revised.
This is a neat trick: combining an inherently fast electrostatic panel with an inherently slow—well, almost inherently—cone bass driver. It's the reason why most subwoofers work so poorly in most high-definition systems: The bass just muddies things up. Not here—unless you position the speaker so as to over-excite room boundaries. (Not too close to corners, please!)
There are some other changes, too, from the original Aerius. The crossover frequency is now set at 450Hz instead of 500Hz. Now, virtually all of the midrange is handled by the electrostatic panel. However, sensitivity is reduced—down to 88dB/W/m from 89dB. Impedance is down a tad, too—now nominally given as 5 ohms, dropping to no more than 1.7 ohms. The impedance of the original Aerius—Aerius "1" as opposed to Aerius i, I suppose—didn't dip below 2 ohms.
No big deal, you say?
Well, the new Aerius, although improved, is more difficult to drive. I can attest to that. An amp that might have marginally driven it before—like the Conrad-Johnson MV55—might not be suitable now. The superbly written instruction manual recommends 80-200W per channel. That's more or less right, but a good 60Wpc integrated like the Bryston B60 might do the job just fine.
When I auditioned the original Aerius, I thought that MartinLogan couldn't have done better for the price. I was wrong. They have done better—considerably better—with the Aerius i.
The bass is deeper, tighter, even more tuneful—now extended down to 40Hz or so. No need for a subwoofer, probably. It would probably add more boom than bloom to the sound. Moreover, the Aerius woofer is even better integrated with the electrostatic panel than before—that is, if your amp can get it up.
I found that the Conrad-Johnson MV55 had a tough time driving the new Aerius. Two MV55 amplifiers would have probably done the trick—one amp for the panels, the other for the bass. With only one MV55, the bass performance was compromised—it turned mainly sluggish, and some of the superb extension was lost (footnote 2).
So much for the bottom—it's the top where the improvements are most notable. The original Aerius could sound a little rolled on top—almost dull. The new Aerius has more top-end extension, more sparkle. That's not all. Off-axis listening is better too. Of course, it's better to deposit one's derri%#232;re in the sweet spot, but that's true of almost all loudspeakers.

Footnote 1: Vol.16 No.6, p.75.—ST
Footnote 2: Some folks suggest a combination of heaven and hell: heaven (tubes) on top, hell (solid-state) down below, where it belongs. I'm not so sure.—ST

In the past, I've criticized speaker manufacturers for not making their models easier to drive. But an electrostatic/cone speaker hybrid is a special case. You can't expect that such a speaker will present an easy amplifier load. All in all, I think MartinLogan is to be commended for making the Aerius i as easy to drive as it is. This speaker, while fairly demanding, is not a current hog.
While a 25Wpc Musical Fidelity A2 was not quite up to the job (hardly surprising), a 50Wpc Musical Fidelity A220 was. So was a Bryston B60 integrated, which is driving the pair of Aerius i's in my living room now. Does the Bryston run out of power? Sure it does. But the Aerius i isn't a speaker that's meant to be played very loud. There's only so much volume—as Wes Phillips uses the term, meaning to fill a room with sound—that you can get out of this relatively small speaker, especially in larger rooms.
Keep in mind that a 60Wpc tube amp will probably put about the same 60Wpc into 4 ohms that it does into 8 ohms—but the 60Wpc Bryston solid-state integrated is said to put out 100Wpc into 4 ohms. This is likely the reason why the Bryston B-60 could drive the Aerius i's with a reasonable sense of ease and why the tubed Conrad-Johnson MV55 could not.
Those who listen in small rooms—particularly those who live in apartments—may find that the MartinLogan Aerius i is a godsend. Big speakers tend to overload and overwhelm small rooms, both sonically and visually. The Aerius i won't.
A person with a relatively small listening room could put together a truly high-end system on the cheap: two grand and change for the Aerius i, $1500 for the basic Bryston B-60 integrated without remote control, $1550 for a Micromega Stage 6 CD player. Well, five grand is not so cheap, but when you hear the sound of these components—exactly the system that's playing in my living room right now—you'll know you've got a bargain.
Plus, the Aerius is elegant. Your nonaudiophile friends will ooh and aah over it, as will your audiophile cronies. And they, too, will immediately hear that the sound is as transparent as the look: see-through, hear-through.
Those who have big rooms and who want big sound will best be served by other, larger, more expensive MartinLogan hybrids, such as the SL3 or the reQuest. My own living room measures 17' by 27', with low 71/2' ceilings. This is just about the Aerius i's room-size limit, I'd say. If my room were any larger, the speakers would be too small.
Placement of speakers in the room is relatively fussy, I found. You want to bring them at least 3' out from the rear wall to get a spacious image—I settled on 5'. If I positioned the speakers more or less equidistant from the rear and side walls, I got boomy bass—no surprise. I ended up with the speakers not quite 3' in from the side walls.
You need to play with toe-in. I was able to get a superb soundstage. But I was also able to compromise the soundstaging with poor speaker placement. Be prepared to experiment. Thoughtfully, MartinLogan has supplied rounded floor supports that you can thread into the bottom of each speaker in lieu of spikes—until the speakers are exactly where you want them.
The sound has the virtues traditionally associated with electrostatics: openness, speed, freedom from coloration. Remarkably, these qualities are preserved below 450Hz, where the sealed-box bass enclosure takes over. The downside is that the Aerius i is now slightly more difficult to drive than before. And the speakers will probably not fill a large room with big sound.
Sorry. In hi-fi, you can't have everything.
Also, speaker cables. I find that with electrostatics—Quads as well as MartinLogans—I prefer solid-core speaker cables. They sound...well, less wiry. Kimber power cords also helped smooth out the sound (footnote 3).
"Zhees speekers, zhey are Amerrrican?" Slava wanted to know.
"Da," I replied. "About as American as you can get—made in Lawrence, Kansas."
I'll have these speakers around for a while. Before I could buy them for the living room, Marina did.—Sam Tellig

Footnote 3: The speakers must be plugged into the electrical mains to energize the electrostatic panels.—ST

Sidebar 5: Aerius i Measurements
As coincidence would have it, Stereophile Guide to Home Theater received a pair of MartinLogan Aerius i's for review in late 1998. The planned review never actually appeared in print, but I did perform a full set of measurements at that time.
Fig.1 shows the later version's plot of impedance magnitude and electrical phase angle against frequency. The broad-brush picture is similar to the earlier speaker's plot. The impedance drops to a very low value above the audioband and there is a single, sealed-box peak in the bass, but the details are different. The lowest value is now 1.5 ohms around 25kHz instead of 1.73 ohms, and the woofer tuning is a couple of hertz lower in frequency. But instead of the original's single impedance peak at 1kHz, due to the crossover, the i has a more complicated appearance in this region, perhaps due to some equalization. There is also a dip to below 4 ohms in the midrange, which will, as ST surmises, prove a little more demanding on the partnering amplifier.

Fig.1 MartinLogan Aerius, electrical impedance (solid) and phase (dashed). (2 ohms/vertical div.)

The original Aerius had a vicious cabinet resonance at 230Hz, which was associated with a wrinkle in the impedance trace at the same frequency. The i still had a slight wrinkle at a slightly lower frequency—200Hz—and investigating the cabinet's vibrational behavior with a tape accelerometer revealed that this mode was reduced in amplitude.
In the time domain, the 1998 Aerius i's step response (fig.2) is virtually identical to that of the 1993 speaker's, as was the waterfall plot (not shown). A point that has become apparent in the seven years since I first measured the original Aerius is that the hashy-looking appearance of panel speaker waterfall plots is not only due to the chaotic nature of the diaphragm, but also to the presence of early reflections. A cumulative spectral-decay plot cannot easily distinguish between the effect of the multiple arrivals and delayed energy due to resonances.

Fig.2 MartinLogan Aerius, on-axis step response at 50" (5ms time window, 30kHz bandwidth).

In the frequency domain, the i's amplitude response (fig.3) is broadly similar to the original's, with the top three octaves of the panel's output shelved-down. But the new speaker has less presence-region apparent, which may be the result of the revised crossover design. That the speaker doesn't sound recessed is probably due to the fact that this frequency region is not suppressed off-axis to the sides of the panel (fig.4), which in smaller rooms will tend to balance the lack of on-axis energy in the same region.—John Atkinson

Fig.3 MartinLogan Aerius, anechoic response on-axis at 50", averaged across 30 degrees horizontal window and corrected for microphone response, with the nearfield woofer response plotted below 300Hz.

Fig.4 MartinLogan Aerius, lateral response family at 50", normalized to response on tweeter axis, from back to front: differences in response 90 degrees-5 degrees off-axis, reference response, differences in response 5 degrees-90 degrees off-axis.