Floor Loudspeaker Reviews

Currently, Stereophile's Recommended Components list includes 19 loudspeakers in Class A (Full Range). As a group, these 19 speakers are, according to the Editor's Note we use to introduce Class A, "sufficiently idiosyncratic and differ enough from one another that prospective customers should read Stereophile's original reviews in their entirety for descriptions of the sounds." True enough, and good advice—yet, in my opinion, the speakers in that class have a few things in common, maybe more than those in the other classes: They're all full-range; they're all capable of more or less realistic SPLs; and for each one, there's at least one reviewer who believes, "as a result of his or her experience," that it "approaches the current state of the art in loudspeaker design."

This last bit is the most important: They all approach this thing we call "the current state of the art," even if the avenues of approach are diverse. To put it differently, all are perfectionist designs, which means that they all aspire to the same, or at least quite similar, sound: the sound of real music in real space.

If ever you're wondering why your favorite (full-range) speaker isn't on the list, that may be why. Some excellent speakers simply have different ambitions.

Magico's M2 (footnote 1)—which I am reviewing here 13 months after John Atkinson's original review—fits the perfectionist mold. Compared to, say, the Bang & Olufsen BeoLab 90 or the mbl Radialstrahler 101 E Mk.II, both of which are also in Class A (Full-Range), the M2 is a fairly conventional design. True, it's a sealed-box, acoustic-suspension design, which these days is pretty rare. And yet it's a dynamic, passive loudspeaker with four drivers, all firing forward, mounted in a box.

What sets the Magico apart (other than its acoustic-suspension design) is its execution. Except for the front baffle and the top and bottom caps, that box is made from carbon fiber, which has a very high ratio of stiffness to weight (although, at 165lb each, the speaker's no lightweight). The top and bottom caps and the front baffle are made from aluminum, itself a fine material to make a loudspeaker from, with a high ratio of stiffness to weight. The M2's enclosure has an adjustable truss-rod system, not unlike an acoustic guitar's, if for a different reason: Over time, and especially when shipped over long distances, the internal bracing can loosen; with a supplied wrench, you can retighten it. The front baffle is shaped to optimize driver dispersion, with a touch of horn-loading applied to each driver.

Magico's constrained-layer–damped footers—the M-Pods—which optimize the coupling of the loudspeaker to the floor, are optional, and at $7600/set they cost as much as many high-quality loudspeakers do. And yet, I suspect that most M2 buyers, who already are paying $56,000/pair for the M2s, spring for the footers, too. They do make a substantial improvement (footnote 2).

"Perfectionist" is how speakers like the M2s should be judged. Speakers not in this category can provide intense, involving, satisfying listening experiences; some people may even prefer a different, nonperfectionist sound: Nothing wrong with that. It's when you focus on details that you notice the shortcomings of these other designs. In other speakers, low-frequency extension may be great, but definition isn't optimal. Or, the sound in the sweet spot is great, but move off-center a little and the soundstage collapses. (But, hey, if you usually listen alone, who really cares, right?) Or, in time, you start to notice that the timbre of certain instruments is a little unnatural: That well-recorded viola is recognizably a viola, but it has taken on some cello character, or it sounds too much like a violin.

In general, I prefer a holistic approach to reviewing—I want to give readers a clear impression of what it's like to hear a pair of speakers (or more broadly, the component under review, whatever it is). Yet, evaluating perfectionist speakers requires a certain amount of list-checking: Getting everything, or at least most things, right is what sets them apart. For discriminating listeners who live with a loudspeaker over time, details matter.

For a variety of reasons, having mostly to do with the pandemic, I have lived with the M2s for a long time. They were installed on December 3, 2019. Today is December 7, 2020, and Magico's Peter Mackay is due here in 45 minutes to help me pack up the M2s so that a shipper can haul them away.

It's rare for a reviewer to live with a component for a full year. As a result, I've gotten to know these speakers very well. I've checked those details. So, here's my checklist.

The Highs Are Present But Unobtrusive.
I've listed this first, because this is important to me. It wasn't long after I got seriously interested in hi-fi that I started to realize that in general there's a problem with highs. Not the low treble but higher up, in the region above the fundamental frequencies of voices and instruments, in the range of harmonics and "air"—plus a lot of annoying crap.

I'm convinced that this is at least part of the reason some people prefer vintage speakers: They tend to roll off the high treble, so there's less up there to annoy. That makes it easier to focus your attention on all the stuff happening further down, in the actual music. PRaT, I think, is often the avoidance of high-frequency distractions (although, inversely, it can also result from the accentuation, by whatever means, of parts of music that provide "pace, rhythm, and timing"). High-frequency noise can give even pure acoustic music an electronic-sounding overlay. High-frequency noise can be repellent. It can make it harder to hear deep into—indeed, to listen into—the soundstage.

But, rolling off the highs is a serious compromise; depending on where (in frequency) it starts, it can alter the timbre of instruments. And rolling off the highs inhibits the experience of music in space.

It's true that no highs are better than bad highs, but there are undesirable consequences to giving them up. You're giving up information, if not about the main part of the music then at least about the musical experience as a whole. It's best to keep the highs but in an unobtrusive way.

With the M2 and other accomplished, perfectionist speakers, the highs are all there—the measurements show it and when you listen you can hear it—but they don't draw attention to themselves, and—most importantly—they don't repel or annoy. (The M2's tweeter is a metal dome, by the way—beryllium, with a diamond coating—and some people say that metal tweeters sound hard, or harsh. Not these.)

What's the explanation, the formula for successful highs? I don't know. Details of the frequency response matter—you need the right amount of energy up there—as does proper dispersion: You don't want high frequencies beaming at you. I've also come to believe that high-frequency distortion is important.

The Lows Are Well-Resolved.
I mentioned above that some old-style loudspeakers roll off the highs. Well, the majority of loudspeakers roll off the lows—all but full-range speakers, and even some of them may reduce the amplitude of the lowest audible tones. It's the same thing, but different: No bass is better than bad bass, but good bass is better still.

The room, of course, is important in the bass, but even in a good room, some full-range loudspeakers, even good ones, do a poor job of resolving, or defining, sounds at the lowest frequencies. Higher up in the bass, they're comfortable and pleasing, but at certain lower frequencies—so low it's not easy to hear—the music can become a smeary mess. What should be distinct instruments and notes are more or less, well, noise. You may not even notice the problem unless you make a direct comparison, but once you hear it, you can't unhear it, and you start to appreciate what better speakers can do.

Definition can be tested—with test tones: Try the warble tones and half-step tonebursts from the Stereophile Editor's Choice CD. (Step back from the speakers first, then move up close, to distinguish between room effects and what's actually coming out of the loudspeaker.)

That kind of muddiness is, of course, far more important with music. John Atkinson addressed this in his original review of the M2, listening to a moment in the third movement of Sibelius's Symphony No.5 with Simon Rattle and the Berlin Philharmonic (16/44.1 Tidal FLAC stream). "At 4:30 in the third and final movement, to echo the ambiguous tonality of the symphony, Sibelius has the double basses, normally used by composers to provide a solid foundation to the harmony taking place above, playing divisi two notes that 'fight,' G-flat and F natural." I listened to this several times; the M2s resolve the notes. They're even better resolved on two more recent recordings of this work conducted by Osmo Vanska, with the Lahti Symphony (16/44.1) and Minnesota (24/96) orchestras, both on Quobuz. Vanska draws out and emphasizes this moment.

Right up to the end of my time with the M2s, even after a year of listening, something in the music would force me to look up from whatever I was doing and take note. Usually it was of some detail in the bass so well-resolved that it seemed real and corporeal beyond my unconscious expectations. It startled.

They're Not Hyperarticulate.
You've read reviews that emphasize the amount of detail a loudspeaker uncovers: the rustling of sheet music or programs, squeaky piano pedals, a musician's intake of breath. Such detail can add a sense of realism to music, but sometimes it's fake detail, a result of a deviation from neutrality in the frequency response, maybe a rise in the upper midrange and treble. At first, it can be amusing, like looking at the world through a magnifying glass, but over time it can literally start to hurt—as though the energy focused by that sonic magnifying glass is slowly setting your ears on fire.

The M2s are appropriately resolved. I feel like I'm hearing what's on the recording—a good proxy for what I'd hear in a live performance from (depending on the recording) good seats. But music from the M2s sounds natural. There's nothing hyper about the M2's articulation.

The Soundstage Is Robust.
As mentioned above, even some very good speakers have a soundstage that's wide and deep only when you listen from within a few inches (left–right) of the prime listening position. Move to the next chair over, and you find yourself leaning toward the center to get the full sonic picture.

There is always some loss of soundstage in a two-channel system when you move off-center, but the amount varies greatly from one pair of speakers to another. Even some very good speakers are poor at this. (There's probably more than one reason, but optimal dispersion—off-axis frequency response that matches the on-axis frequency response—is partly responsible, or so I've read, and so it seems to me.)

With the Magico M2s, I can sit aligned with the inside edge of either speaker with only a little loss of 'stage.

I just tested this. I am sitting just as described above, aligned with the inside edge of the right speaker. I used my Bosch GLM 30 laser measure to determine my distance from the speakers. I am 8' 2" from the right speaker, 11' 6" from the left. I'm listening to Peter McGrath's recording of Vadim Repin playing the Shostakovich Violin Concerto No.1 with James Judd and the Florida Philharmonic Orchestra (unreleased, 24/44.1 MQA, unfolding to 44.1kHz). When I sit in the sweet spot, the soloist is slightly left of center. Now, from my far-right position, Repin and his violin have lost a little bit of focus—just a little—and the soundstage has lost a little depth, but he and the orchestra are still far to my left; indeed, the sound of Repin's violin is coming from a point that's almost exactly halfway between the speakers. Also, there is no noticeable change in the instrument's timbre.

Images Are Fleshy.
I'm very fond of MacArthur Award winner Cécile McLorin Salvant's recordings on Mack Avenue, engineered by Todd Whitelock and mastered by Mark Wilder. (I also like seeing her live and have done so several times.) Choose any track on any album, LP, CD, streaming, or high-rez download, and Salvant's voice is profoundly present, spookily real. In my experience, the best examples of this—the sense of an actual, flesh-and-blood human floating in space—come off best on certain old-school systems, often midrangy high-sensitivity speakers with tubes. Maybe it's the high-frequency rolloff thing: less distracting noise.

Right now, I'm listening to a live recording of Salvant singing Bernstein and Sondheim's "Somewhere" from West Side Story. (This is on The Window; the CD is Mack Avenue 1132, the LP 1132LP, but I'm listening to the 24/96 download, and downloads don't have catalog numbers.) This song was recorded at the Village Vanguard—regrettably I missed it, but my son was at one of the sets—with the brilliant Sullivan Fortner on piano. The microphone alters the timbre of Salvant's voice a little—and yet there she is, microphone in hand. (Actually, what is tangible is that point of contact between Salvant's voice and the mike, floating there in space. Magic.)

The Scale Varies With The Recording.
I listen to musicians and ensembles in a wide range of scales: orchestras, big bands, solo piano, grand opera, chamber music, small jazz ensembles. Some of it is recorded in a big space with lots of room sound. Some of it is recorded in a smaller space and closely miked. Every recording has its own, appropriate scale.

Some big speakers turn pianos into battleships. I admit that this can be a pleasant experience, but it's not realistic; plus, it's unnerving to encounter a 15' soprano. Matthias Goerne is appropriately sized when he sings Schubert's "Litanei auf das Fest Aller Seelen," D.343 (MQA unfolded to 24/96), standing near a realistically sized piano. (This is an MQA test track; the only place I know to find it in any commercial form—it's not MQA—is in the 12-CD set, Schubert Lieder, Harmonia Mundi HMX 2908750.61.)

The perceived scale of the performance depends very much on the recording of course—specifically on mike placement—so scary-big sopranos are not always the speakers' fault. It also depends on how your speakers are set up. Over time, though, listening to lots of music, you get a sense for how a speaker scales music. Some big speakers—I'd call these medium-sized—make all music sound big, and many smaller speakers are incapable of significant scale.

The M2s' scale appropriately miked music well. Well-recorded string quartets (and other chamber groups) are string-quartet sized. A well-recorded orchestra isn't quite full-size—not in my space, with the speakers placed about 9' apart—but it's big enough to be satisfying, and I suspect the 'stage would be bigger in a bigger, acoustically competent room. At an appropriate volume, recordings played back by the M2s make small, intimate spaces sound small and intimate, big spaces seem big.

Timbre Is Consistent Across Frequencies.
As on a piano, a loudspeaker can sound different in different frequency ranges. You can detect this by listening to a particular instrument that covers a wide range and has a rich tonal palette. A simply recorded cello works well, perhaps in the Bach solo cello suites, or those by Britten. If the speaker is altering the sound of the cello from one octave to another, and you're used to listening to cellos, you can tell.

I'm listening to Edgar Meyer's recording of the Bach Suites performed on double bass, a FLAC rip of Sony Classical SK 89183, a CD. The double bass doesn't get high in frequency—it's a double bass after all—but it covers the lower half of music's tonal range, and over that range it is all of a piece—recognizably the same instrument from note to note.

Sounds Of Instruments And Recordings Are Well-Differentiated.
There are pleasures to be had with home listening that are not available in live performance, such as the ability to listen to a variety of recordings and performances, one after the other. Right now, Roon Radio is playing piano music, chosen by their algorithm, based on my tastes. I'm hearing one solo-piano piece after another, often short movements, performed on different pianos, recorded with different microphones, different mike placement, and different engineering. Every piano recording I'm hearing sounds dramatically different from every other piano recording.

Hearing these differences—the many varieties of piano sound, pleasing and otherwise—provides me with a surprising amount of pleasure: I must be an audiophile. Who would have known?

This ability to differentiate also extends to components under review. Differences in amplifiers and sources are easily heard.

A component that can reveal differences among components and recordings can also reveal differences within a recording: Such a component (including loudspeakers) does a better job expressing what the musician endeavored to express, and that the engineer worked to capture: It delivers more of the music. I am also able to hear, very clearly, how certain pianists alter the piano's sound with pedaling and touch within a particular work. You can hear it too, I'm sure, on your speakers. It's a matter of degree. The M2s differentiate tone and touch very well.

Do I have anything negative to say about the M2? The binding posts are just above a shelf that sticks out half an inch or so on the back of the cabinet, so large spade connectors don't fit when connected from below. Solution: Use banana plugs, or feed the cables in from above. No problem.

My listening chair, a Poäng from Ikea, with the matching ottoman, is too low for a proper listening chair; I really should replace it. With the M-Pods installed, the M2s—though not overly tall—won't tilt forward far enough to put my ears on the optimal axis while I'm sitting in my Poäng with normal posture. I find myself sitting up straight, craning a little, unconsciously seeking out the optimal sound—this at 9'–11' from the speakers. I'd buy a height-adjustable office chair to replace the Poäng if I could find one with a matching, height-adjustable ottoman.

There are many very good loudspeakers out there, all capable of communicating music effectively and delivering profound pleasure. Only a few aspire to get all the details right and even fewer succeed. Magico's M2 is on that short list.—Jim Austin

Footnote 2: At the time of writing, the M2 cost $56,000/pair and the MPod three-point stands cost $7600/pair. In January 2021, the M2 cost $63,600/pair with non-optional MPod stands.