Epos ES12 loudspeaker

Blind loudspeaker listening tests are hard work, not least because usually, most of the models being auditioned fail to light any musical sparks. But back in the spring of 1991, when a small group of Stereophile writers were doing blind tests for a group speaker review, one speaker did light up smiles on the listeners' faces, including my own. (We don't talk during our blind tests, but it's more difficult to keep body language in check.) Once the results were in, we learned that the speaker that got the music right in that test was the diminutive ES11 from Epos in England (footnote 1).

The ES11 featured a major innovation in speaker design: the use of mineral-loaded structural polypropylene front and rear baffles connected by four tensioned steel rods. The fiberboard side, top, and bottom panels are then wrapped around this rigid assembly. The front baffle also acts as the frame for the woofer. Like the earlier ES14, the ES11 didn't have a woofer crossover filter, the drive-unit rolling off naturally above its passband. And the tweeter crossover consisted of a single series capacitor. According to designer and Epos founder Robin Marshall (see Stereophile, February 1989, p.59), this maximizes a loudspeaker's transparency.

Marshall subsequently left Epos to join first Spendor, then Verity, the parent company of Mission, Roksan, and Quad [and in 2006 works for Harman International in Europe]. Along with Tannoy, Mordaunt-Short, and Goodmans, Epos is now a division of the TGI group [it was acquired by Mike Creek in 1999—Ed.] but seems to have kept its quality mission—the Epos ethos, if you will—intact.

Superficially, the '12 looks identical to the ES11, other than the slight profiling of the side panels around the baffle. A small, two-way, reflex-loaded design finished with wood veneer, it marries a 1" metal-dome tweeter to a small plastic-cone woofer. However, it appears that all that has been carried over from the earlier design is the front panel and the woofer magnet—everything else is new.

Starting at the top: The tweeter, made by Epos, features an aluminum-alloy 1" dome, 50µm thick, with a nitrile-rubber suspension and rear-cavity loading. The magnet gap is a little longer than the voice-coil, to maximize linearity, and there is ferrofluid in the gap. Beneath it on the front panel is the 6.5" woofer, the chassis of which is integral with the baffle. Also made by Epos, this features an injection-molded polypropylene cone, its outer edge thinner than the neck and center region to optimize the strength/weight ratio. An inverted nitrile-rubber surround was chosen because of its good termination properties and low hysteresis. The woofer's voice-coil is wound on a vented aluminum former to give good power-handling.

Unusually, there is no woofer dustcap; instead, a tapered, stationary "phase plug" is fastened to the end of the central pole-piece. This is said to optimize the woofer's dispersion at the top of its passband, to better match that of the tweeter at the bottom of its passband. An important aspect of the cabinet's construction is that there are none of the usual acoustic obstacles—bolt heads, screws, etc.—around the flush-mounted drive-units. This, too, should aid dispersion.

The woofer is reflex-loaded with a rear-panel port 1.5" in diameter and only 3" long. This is flared to reduce wind noise at high levels. Electrical connection is via banana plugs only. The two sets of 4mm sockets on the rear panel make a tight connection with the banana sockets supplied by Epos or used on the CableTalk cables distributed by Music Hall. They're a little oversized for Monster X-Terminators, however.

A few aspects of setup need to be discussed before I describe how the ES12 sounded. The tweeter is protected by a wire-mesh cover, held on magnetically. It can be (carefully) pried loose with, preferably, a nonmagnetic blade of some kind. A trace of tizziness that I noticed before the grillectomy disappeared with the tweeter domes exposed. The downside, of course, is that the tweeter diaphragms are very vulnerable—perform this tweak at your own risk.

Second, the four internal tensioning rods connecting the front and rear baffles are covered by caps on the baffles. The dealer from whom you purchase your ES12s should ensure that the end nuts have not worked loose in shipping, which will reduce the speaker's sense of pace. However, don't attempt to tighten these yourself.

Third, the ES12 seems very transparent to the cables with which it is used. Changing from the CableTalk Concert 2.1 to Cardas Cross gave a noticeable improvement in the sense of top-octave air and overall treble refinement. The CableTalk bass, however, was both well-defined and had an excellent sense of pace. The Cardas in these areas was perhaps a little more rich, sacrificing transient speed for luxurious bloom. Both of these cables were used single-wired.

I ultimately used a bi-wired set of AudioTruth Sterling, which improved the treble presentation compared with the Cardas but, more important, gave stunningly tight, well-defined, deep low frequencies. It was not possible in the review period to thoroughly explore whether this was due to the bi-wiring or to the cable itself. However, a brief experiment with single wiring with Sterling suggested that both factors were at work.

If you feel it lunacy for me to use a set of speaker cables costing more than three times the price of the speakers, then you, of course, are welcome to stick with CableTalk, which I can recommend as an excellent value for the money. The Law of Diminishing Returns is as true with cables as in any other field of audio. The point I wanted to make was that the ES12s' intrinsic performance was good enough both to allow the differences to be perceived and for it to benefit from the better cables.

Fourth, the Epos is very critical regarding how it is connected to its stand. Music Hall's Roy Hall recommends coupling the speaker to its stand with upturned cones. I found this maximized the sense of pace, rock recordings boogieing down in an impressive manner. But for classical orchestral recordings, I found that replacing the cones with small pads of Blu-Tack added a welcome sense of low-frequency bloom—though this did somewhat compromise the ES12's rhythmic quality, adding a sense of lethargy on rock. I recommend that ES12 owners experiment for themselves—as long as they understand that they should first get the best speaker stands they can afford.

Finally, the ES12's designer has optimized its bass balance with the assumption that the speaker will be positioned relatively close to the wall behind it. While I started off with the Eposes about 15" from the wall, I found that moving the speakers out to about 3' maximized clarity and articulation, the only downside being a slightly lean lower midrange. Again, ES12 owners should experiment for themselves.

The Epos ES12 was a big hit at HI-FI '96 among Stereophile's writers for its easy, unexaggerated sense of musical communication. And that was my immediate impression when I first set the speakers up in my room: there was very little getting in the way of my music.

That's not to say the speaker was without character. The midrange initially sounded a little nasal, though this appeared to diminish with time, either because I was accommodating to it or because prolonged break-in actually was reducing its level. I also found the high frequencies to be a little mellow—"sweet" was the word I found sprinkled through my listening notes, particularly when I was driving the speakers with the little Pass Labs amplifier. However, this did not mean that I found the speaker to sound dull or uninformative. In fact, its high frequencies were remarkably transparent, as was its midrange. Although I could hear that there was something not quite right using the 400Hz and 500Hz warble tones on our Test CD 3 (see the "Measurements" Sidebar), music in this frequency region seemed to emerge relatively unscathed.

During the ES12s' review period we mastered the LP release of Stereophile's Sonata recording (see the March 1997 issue, pp.75–89). I therefore used the ES12s to do some level-matched listening tests comparing the LP test pressings both with the 20-bit master tape and the CD version, noise-shaped to 16-bit resolution with the Meridian 518. (For these comparisons, the same Mark Levinson No.30.5 D/A converter was used as had been used to master the LP.) The differences between the three media were readily audible via the Eposes. The CD lacked a little soundstage depth compared with the master. Even though its frequency balance was identical in the midrange and highs to that of the CD, the 20-bit version sounded as if it had more low-frequency extension. Though the LP was not quite as close, tonally, to the master, it did sound more involving than the CD. No way did the fact that the Epos speakers were the least expensive part of the system get in the way of the effectiveness of the listening tests—a tribute to its clarity and lack of coloration.

In fact, other than the nasality I initially noted, I found very little to criticize during the auditioning. While this can be frustrating for the reviewer looking for juicy flaws upon which to wax caustic, it does mean that a speaker is doing a lot of things right. When I put on favorite "torture" tracks to investigate a specific performance area, I tend to end up just grooving on the music—particularly when the speaker in question, like the ES12, has a superbly articulate midrange.

For example, a current favorite recording of mine for detecting upper-midrange aberrations is Hugh Masakela's 1994 Hope CD (Worldly/Triloka 7203-2). The voices are recorded close, meaning that any tendency for a speaker to sound bright will both add hardness and push the vocals too far forward in the mix. I generally select track 3, Masakela's tribute to Nelson Mandela, where I also listen for a speaker's ability to delineate the close harmonies in the celebratory chorus. With the articulate Eposes, I ended up playing all the way through to the end of track 12, "Stimela," the plaintive song about the train that takes migrant workers to the mines and factories of Johannesburg: The Eposes were fazed neither by the singer's whoops, echoing the train's whistle, nor by the eight-note percussion crescendos.

The ES12s' imaging was well-defined in both width and depth planes; recordings capable of producing a soundstage were well-served. And dynamics, both in the sense of preserving quiet detail in the presence of competing, louder sounds, and in the ability to go from soft to loud without compression or blurring, were excellent. Bass drum started and stopped without noticeable overhang, appropriately kicking the music along.

But this is a small speaker. While bass instruments were well-defined, and bass guitar in particular was presented with a believable signature in the upper bass, there was no real energy present below 60Hz or so. And the cap on loudness was about 95dB. The Epos ES12 is best suited to relatively small rooms, where the room boundaries can give the appropriate low-frequency reinforcement, and extreme sound-pressure levels will not be called for.

Summing up
Balance. That's what you need from a relatively inexpensive loudspeaker, where the designer can't max out in any one area of performance without compromising another. Balance is what you get from the Epos ES12. Its faults are small, and balanced by excellent performance across the board. At $1095/pair plus stands, the Epos ES12 might well be the loudspeaker bargain of the year.

Footnote 1: Stereophile's rave review of the Epos ES11 appeared in the July 1991 issue (Vol.14 No.7, p.111), with a measurement Follow-Up in October 1991 (Vol.14 No.10, p.205).
Epos Acoustics
US distributor: Music Hall
108 Station Road
Great Neck, NY 11023
(516) 487-3663