Mirage M-3si loudspeaker

What's 1/16" narrower, over 1/2" shallower, and 3/16" higher than the Mirage M-3 loudspeaker? The new Mirage M-3si, that's what. Though the published dimensions for the old and new speakers are the same, my eyes told me there was a difference between them when I had them side-by-side in my listening room in Santa Fe. Being the compulsive type, I got out my trusty tape measure. No, my eyes had not deceived me—the M-3si is skinnier and taller. As I waltzed them into position, I sensed they weighed about the same as their predecessors; close enough that setting them on Arcici Super Spikes is a two-man operation. Though either speaker makes a definite presence in a room, I still find their high-gloss, black finish (the only finish available) unassuming, attractive, and elegant.

In developing the second-generation M Series (footnote 1), Mirage engineers focused on four areas they identified as being critical to the product's improvement: tweeter design, crossover points and slopes, midrange and woofer cone materials, and cabinet construction. Mirage claims their Pure Titanium Hybrid (PTH) tweeter, which took two years to develop, is the "first dynamic tweeter which can claim the transparency achieved with electrostatic or ribbon designs while eliminating the sonic overhang typical of those designs." According to Ian Paisley, the company's VP of Engineering, the new tweeter consists of a feather-light, pure titanium 1" dome on a cloth suspension system. The dome's low mass and rigid, close-tolerance shape is said to ensure near-instantaneous response to signal inputs, while the cloth suspension eliminates post-signal resonances, or "ringing," typical of other metal-dome designs. Mirage feels their PTH has the fastest time-domain impulse of any tweeter over a full start/stop cycle. In addition, the new tweeter's voice-coil is wound on the dome, which is said to result in low distortion, lowered fundamental resonance, and increased speed.

The crossover has also been revised to eliminate crossover distortion, flatten the frequency response, and improve coherency between the low- and high-frequency drivers. The number of crossover parts has been reduced to 23 in the M-3si, from 28, and the crossover frequencies are lower: 350Hz and 2kHz compared with 400Hz and 2.2kHz in the M-3. The rolloffs are second- and third-order: 12dB/octave (electrical) high- and low-pass at the 350Hz crossover point; 12dB/octave low-pass and 18dB/octave high-pass (both electrical) at the 2kHz point.

All woofers and midrange drivers in the new M Series use injection-molded polypropylene cones impregnated with carbon and other proprietary additives for increased strength and rigidity. The woofer surround is still Butyl rubber, but the magnet is heavier: 28oz in the M-3si compared to the 20oz magnet in the M-3. This, along with a dual (or second) winding on the voice-coils of the woofer and tweeter (the latter also sports a heavier magnet), is said to lower distortion and increase power handling. The woofer is designed and built in-house; the midrange and tweeter are designed at Mirage, but built elsewhere. As in the M-3, the M-3si features matched pairs of back-to-back PTH tweeters and midrange drivers (with the rear-radiating sound delivered in-phase with the direct output from the front), but uses a single, front-mounted, ported 10" woofer.

The cabinet was also beefed up; it features thicker MDF outer walls and a completely redesigned internal lattice structure. (Internal baffles are now 1" thick, as in the M-1.) In addition to further reducing unwanted resonances, these changes are said to result in enhanced deep-bass output and the tightest, most vibration-free enclosures Mirage has ever produced. A knuckle-rap on the sides of the cabinets produced a thuck-thuck sound, indicating a very well damped enclosure with a relatively high Q.

The bass ports have also been re-engineered and enlarged for maximum efficiency and minimum vent turbulence (a type of distortion occurring when a large volume of air is passed through a relatively small, circular port). They are also stair-stepped, "in a manner approximating a ramp," to reduce or eliminate the chance of the grillecloth hitting the rear baffle. In the M-3si, internal volume was increased 12% without significantly affecting the outside dimensions. The new speaker has also been tuned 3Hz lower.

I was happy to see that the M-3si offers the provision for solid floor-coupling via ¼" steel spikes, which can be threaded into steel inserts set in the base. Taller, stronger spikes are available at extra cost for those (like me) who need additional clearance and strength. I was also glad to see the plastic five-way binding posts used on the M-3 replaced with much stronger, gold-plated, hex-head, solid metal ones. There are two pairs, so bi-amping or bi-wiring is possible. You can torque these babies down tight with a combination wrench; they won't squeal or crack. Just make sure the nut which holds this assembly in place, located underneath the speaker, is good and tight. The fact that the posts are closer to the rear edge of the base also makes cable connections a whole lot easier. Like the M-3, the M-3si is sold in mirror-imaged ("handed," as JA would say) pairs, the left speaker's drivers hugging the right edge of the cabinet; those for the right speaker hug its left edge.

Listening conditions
The timing of this review coincided with my move from mile-high-plus Santa Fe to sea-level San Diego. I was thus able to hear the speakers in a rather small, dry room (Santa Fe) and a humid room roughly half again as large (San Diego). (If there were any sonic changes due to different atmospheric conditions, they were negligible (footnote 2).) My Santa Fe room had become a refuge where I could immerse myself in whatever music I thirsted for at the time. "Immerse" is the right word, for the system in that room was capable of marvelous soundstaging, pinpoint imaging, and virtually holographic presentations of musical events. It was magic—and addicting. When I invited friends over for a listen, it was often difficult to get them to leave, despite threats from significant others. It was with some trepidation, therefore, that I packed the truck for the move to California and a listening room of unknown character.

My fears were largely unfounded. In all but one respect, the new listening room is an improvement. It is 16½' wide by 21½' long, with wall-to-wall carpeting and solid plaster walls. The 8' ceiling is also plaster. Roughly half the area of one long wall is glass windows covered by adjustable, floor-to-ceiling draperies which I have adjusted to get correct left/right balance at the listening position. As I expected, the highly reflective nature of the walls and ceiling caused a severe slap echo problem, but by carefully "tuning" the room with a couple of handfuls of EchoTunes, four CornerTunes, and half a dozen RoomTunes (footnote 3), this apparent handicap was eliminated. In fact, what had at first been perceived as a liability was turned into an asset: the room's energy and speed lent a vibrancy to the sound which had been absent in Santa Fe. This was accompanied by a significant (and welcome) increase in my system's powers of resolution.

The size of the room also lent more support to low bass (especially organ pedals). The spatial aspects of recordings—depth, height, and width—are exemplary. What my new room doesn't have (and I'm working on this) is the overwhelming sense of recording venue I was privy to in Santa Fe. I miss that "envelope" of air which hung, umbrella-like, over the performers and helped define the acoustic signature of the hall, auditorium, studio, or stage. I've retrieved space in spades, but do not yet sense the ultimate and conclusive details of its character. (I don't attribute this failure to the loudspeakers.)

Room positioning with bipolar radiators such as the M-3si (as well as most ESL or ribbon dipolars) requires a little thought and a lot of placement testing (otherwise known as muscle). After much schlepping to and fro and side to side, I ended up with the speakers firing down the room's long axis, 7' out from the short wall—a position which put them roughly a third of the way into the room (footnote 4). I set the speakers a little over 3' from the side walls and toed them in slightly so that the distance from the centers of the speakers to my ears was 5'. This resulted in 6' separation, center-to-center, between the speakers. With the speakers so placed, I got a seamless, wall-to-wall soundstage with excellent depth and precise image focus. I've always admired the finest minimonitors for their ability to be heard and not seen. On the best of them, music seems to emanate from everywhere but their grilles. This is an enviable quality in any transducer, yet achieving it has been the Nemesis of many larger, full-range loudspeaker systems. I'm happy to report that the M-3sis, despite their size, succeeded admirably in their ability to disappear. On the best program material, these speakers faded from consciousness, letting the performers and music take over.

The M-3s have been my reference speakers ever since I first heard them (and immediately pried them loose from) Tom Norton and the Stereophile listening room. I've enjoyed their virtues and accepted their (to me) minor faults. In my room in Santa Fe, they never failed to satisfy me with their musicality and ease of listening. If a speaker is the least bit aggressive or forward in its presentation of these events, I soon weary of the experience. If I'm unable to exorcise the demon from the system, I make changes in the setup or look for other things to do. In the considerable length of time I've spent listening to music through the M-3s, I rarely had the urge to seek out those "other things."

Having lived with the M-3sis for several months, I have even less of an urge to leave the listening seat. These are wonderful loudspeakers, and should immediately appeal to all music lovers out there whose desires for high-end sound are tempered by worldly realities. (How many of us can afford speakers which cost more than what many of our parents spent for a house?)

To test the M-3sis' power-handling capabilities, I threw a lot of material their way which demands to be played LOUD. From the adolescent primal screams of such groups as Nirvana, Butthole Surfers, Fugazi, and Big Black to the industrial, synth-tech dance music (which must be played even LOUDER) of Front Line Assembly and Front 242, I never got the sense the speakers were even close to meltdown. As I turned the volume knob past 12 o'clock—accompanied by three-digit SPLs, a shaking sofa, rattling windows, and watering eyes—not once did the speakers blat, blare, honk, or squawk. They just, in Coreyspeak, "KICKED ASS!" At any prudent volume level, they seemed incapable of making unmusical sounds in my room (even though some of the music could be construed as, uh...unsound).

I was also struck by their apparent neutrality. By this I mean that what I heard coming out of the loudspeakers made differences and idiosyncrasies in upstream components' sonic signatures easier to recognize. Homogenization may make milk more palatable, but it's best avoided in the reproduction of music. For example, sonic differences between power amps (all else being equal) were more pronounced through the M-3sis than through the M-3s. Though I didn't learn which amp was more "accurate" (and don't much care), I discovered which I preferred listening to. On the earlier Mirages, it was harder to make this decision.

Footnote 1: See Tom Norton's review of the M-3 in Vol.13 No.11, wherein it walked away with a "... solid Class B..." recommendation. That review, along with Larry Archibald's review of Mirage's flagship M-1, in Vol.12 No.6, are required reading for more detailed accounts of the theoretical grounding and technical considerations which led to the development of the original M-series loudspeakers. I will not rehash that material here.—Guy Lemcoe

Footnote 2: In early 1991 I performed an experiment measuring the same loudspeaker under identical conditions in Santa Fe and Washington, DC. The results showed that while there is a reduction in sensitivity with increasing altitude, there are no tonal-balance changes other than perhaps when the wavelength of the sound is smaller than the diaphragm (a 1" tweeter above 20kHz, for example). However, relative humidity does play a significant role in changing the sound of your loudspeaker. If your system sounds harsher than usual or more aggressive in the upper midband, check the humidity; it may be lower than usual. I keep the relative humidity in my listening room at a constant 50% year-round.—John Atkinson

Footnote 3: I consider the use of Michael Green's 'Tunes to be among the most cost-effective improvements one can make to a listening room. Another is running dedicated mains service, but since I'm renting, this option is not practical. One of these days...—Guy Lemcoe

Footnote 4: This is the same relative distance from the rear wall I found optimum for my listening room in Santa Fe. Coincidence? Maybe. Rule of Thirds? More likely.—Guy Lemcoe

3641 McNicoll Avenue
Scarborough, Ontario M1X 1G5
(416) 321-1800