Manufacturer: EOSS Corporation, 19420 Hallmark Lane, Cerritos, California 90703; 562/809-8630 www.soundpoles.com; eoss@soundpoles.com

Price: Two Soundpole satellites, cables, and two dedicated outboard woofer systems: $4,990

Source: Manufacturer loan

Reviewer: Howard

Jim Frane reviewed an earlier version of this system in Issue 91, and now I have what appears to be the latest version of this product on hand to audition. As it now exists, the Musiko system consists of four

sections: two "SoundPole" satellite speaker systems and two powered woofer systems, which should be connected to the poles by means of two, six-foot speaker cables that are also part of the package.

The SoundPoles themselves are each about six feet high. However, they do not look as imposing and room dominating as most other six-foot tall speakers, because as "poles" they are very narrow in profile, being only 3.5 inches on a side. Each "enclosure" is a metal screen cage that surrounds an interior enclosure that holds eight small, proprietary drivers. Each also has a built-in high-pass filter that limits its manufacturer-specified bottom-end extension to 180 Hz. Each pole sits on a well-finished and heavy cast-metal metal base that also has the electrical connections. The pole sections and the base are extremely well finished in light and medium gray tones.

Each woofer system (EOSS declines to call them subwoofers for good reason, because they cannot probe the deepest frequencies) is about 11.5 inches square and 28 inches high with the mandatory spikes installed. Each contains a downward-facing, 8-inch driver, along with a small, downward-facing port, plus a 150-watt amplifier and a variable low-pass filter that allows for a smooth dovetail with the respective pole on that channel. The cabinets themselves are covered with a full wraparound medium-gray fabric that matches the tones of the metal-cage section of the poles. They have high-gloss tops that match the lighter gray of the pole bases.

When I received this EOSS shipment, I was struck by how well packaged they were. The poles were carefully packed in one, long container, the two bases were in second container, and each subwoofer was well packed in its own sturdy box. The packing job was topnotch, and so the shipment arrived in fine condition.

Assembling the pole sections to the two bases was easy, thanks to the supplied instructions and the required parts. (I did find a few typos and inconsistencies in the owner's manual, but those should be corrected in the next printing.) The physical quality of all of the hardware was very high end, as one would expect with a system package in this price class. It is obvious that a lot of time, effort, and money went into producing these speakers.

Hooking up the poles to an audio system is simplicity itself. They have two sets of connector, s on each base (one pair being wide spaced and the second with the jacks close enough together to handle double-banana plugs), and those connectors themselves are connected together underneath the base and also connect to the circuitry in the poles. You run leads from your power amplifier to one set of connectors on each channel (either on the woofers or on the pole bases; it does not matter), and then connect the poles to the woofers by means of supplied, six-foot speaker cables. The length is deliberate, because each woofer is supposed to be fairly close to each pole, with the optimum distance being about two feet behind.

The above arrangement means that the powered woofers are driven by speaker-level signals. If you are uneasy about doing a speaker-level hookup to a powered woofer, the woofer amplifiers also allow for a line-level hookup from a preamp-out jack. (The owner's manual does not mention the fact that a line-level hookup would require you to use to use a line-level "Y-splitter," since the preamp output would also have to drive a power amp for the poles.) The nice thing about the speaker-level hookup (which is the one EOSS apparently favors) is that the woofers can easily be connected to any receiver, including one that does not have a preamp-out feature. The speaker-hookup distortion issue, is probably of no consequence.

Regarding line-level hookups to the woofer, the manual points out that you cannot use the usual subwoofer-out hookup from a receiver or processor, since those typically make use of an internal crossover. The EOSS woofer's crossover cannot be bypassed to allow for that kind of connection, and even if it could the typical 80-100 Hz low-pass rolloff we have with typical sub-out connections would be too low in frequency.

The woofer amps have the same kind of controls that you find on any number of subwoofer amplifiers: level, variable crossover frequency, phase, and power. The latter may be left on all the time, as the woofer amp has a signal-sensing circuit that will shut it down once the input signal has been discontinued for several minutes. Note that the sub does not have any kind of high-pass filtering, since the poles have that circuitry already built in. This is a dedicated, two-woofer system that can only be successfully used with the poles.

EOSS recommends that the SoundPoles be set up in a fairly strict manner. They cannot, for example, be placed close to a front-wall boundary if you want them to perform as well as they should. While a distance of 3 feet to that wall is OK, according to the owner's manual, it is better to get them out even further than that, with the optimal range extending clear out to 10 feet, or even more, according to the designer, Ed Onglao. You will also want to get them fairly far from the side walls, although the owner's manual was not specific on that issue. The distance from the center axis between the speakers to the listening position can be anywhere from 8 to 12 feet, with the distance between the speakers themselves anywhere from 7 to 9 feet.

The SoundPoles are dedicated systems, with one specifically for the right channel and one specifically for the left. In my 18 x 22 foot main listening room, I placed the test samples 6 feet out from the front wall and about 8 feet apart (and therefore about seven feet from either side wall), and about 10 feet from the primary listening position. Following the instructions in the owner's manual, I toed the poles inward, so that they were aimed right at the central listening position. I do need to mention here that although this is the best way to arrange the poles, according to the manual they are not systems that require a tight sweet-spot listening position. The soundstage will hold up fairly well, even at locations some distance from the preferred central seat.

For all of my measurements and listening, I positioned the woofer sections about two feet behind each pole, but aligned them in such a way that each pole was between the woofer and the sweet-spot listening location. This is basically how the optimal positioning was illustrated in the owner's manual.

As those of you who have read my reviews in the past may remember, when I measure speakers my AudioControl SA-3051 RTA integrates all incoming signals while I move a microphone slowly within a 1 x 1 x 5 foot area at roughly head level at the listening couch. I use uncorrelated pink noise for this test, which allows for a workable bass-level reading and also delivers a good simulation of how a pair of systems will propagate their total output throughout the listening room. More importantly, it registers how uniformly the systems deliver a blend of direct and reverberant field signals to the listening couch. The result is a room curve that does a good job of measuring real-world spectral balance.

Unfortunately, in terms of measured performance using this technique, I did have some trouble with the Musiko system. The problems were twofold.

First, I simply could not get a smooth dovetail between the woofer systems and the poles. The poles themselves could not quite reach down to their specified 180 Hz lower limit, and they were essentially out of the picture by the time they were down to that frequency. While they are supposed to reach uniformly down to 180 Hz, it is possible that part of the dip I measured from the poles was the result of a floor cancellation that would not show up on an anechoic measurement.

On the other hand, the woofers were very attenuated in output by the time they reached up to 180 Hz, and this was the case even with their low-pass adjustment set for maximum high-frequency extension. I got this result with the room curves I ran with both woofers running solo (in both my main room and in my living room) and also with a near-field measurement of one of the woofer cones. The result was a room-response dip of about 6 dB, centered between 160 and 200 Hz that I could not shake, no matter how I adjusted the woofer low-pass control or positioned the woofers.

The second problem involved the overall smoothness of the systems above the crossover point. Relative to the admittedly hard-to-determine average levels, there was a one-octave wide peak that reached +5dB at 1 kHz, a sharp 5-dB dip centered at 2.5 kHz, and a one-octave wide dip that reached -6 dB at 6.3 kHz. While this sounds pretty bad, it actually looks worse on the response plot, because there were no uniformly flat areas between these peaks and dips, and the plot also included the dip centered at 200Hz.

Except for a relative degree of smoothness between 250 and 800 Hz, everything was up and down, as if a comb filter was driving the systems. Between the 2.5 and 6.3 kHz dips there was a sharp rise centered at 3.15 kHz, and above the dip at 6.3 kHz the output rose to a sharp point at 12.5 kHz, with a fast rolloff above that frequency.

This is in stark contrast to a number of exemplary systems I have measured in the past, including the Dunlavy SC-II and Cantata models, the Waveform MC/MC.1s, the AR Phantom 8.3s, and my own Allison IC-20s. While this was not the most erratic curve I have ever plotted, it was the most irregular I have ever plotted from a stereo speaker package that had a price above $2,000. I also measured both channels driven with mono pink noise, measured each channel one at a time, and measured the satellites separately, without the woofers, and the results were essentially the same as what I got when measuring in stereo with uncorrelated noise. This is not a pancake-smooth pair of speakers, at least when compared to those I mentioned above.

Now, an EOSS representative indicated to me that the SoundPoles cannot be measured by ordinary techniques, because to the way they interact with each other. I have no problem with this if we are talking about soundstaging, imaging, focus, projection, etc. However, my rule has long been that first thing a speaker system must be able to do before we deal with anything else regarding its performance, is deliver reasonably smooth response at the listening position, be that response dominated by the reverberant field or by a well-focused direct-field output. The Musiko System did not do an impressive job of achieving this result.

The bass response was pretty much what one would expect from a pair of ported, 8-inch drivers. From the top of its range, it slowly rose due to room reinforcement down to about 80 Hz and was uniform from there down to about 50 Hz, with a rolloff below that frequency that reached average midrange levels at about 40 Hz. With test tones, the bass was clean to 50 Hz, but if you listened to material with a lot of low bass or played test tones moderately loud at very low frequencies, the port wind noise was quite audible. Since EOSS does not call the woofers subwoofers, I can easily forgive the noise artifacts and low-end rolloff, and simply recommend that those wanting deeper extension invest in a good subwoofer to augment the output of the two EOSS units. Of course, when you do that you will end up with three outboard bass enclosures in your listening room.

OK, now perhaps flat room response is not what we are after here. After all, according to the owner's manual, this speaker package is able to do something that few other systems can do: create a believable soundstage somewhat behind the speakers, with a uniform and realistic spread of instrumentation in the open space between the poles and the front wall.

I did some close listening to the systems playing by themselves (powered by a pair of Sherbourn 1/ 300MB monoblock amplifiers) and was initially impressed with what they could do. My first audition was with an excellent recording of Martinu's Third and Fifth symphonies (Naxos 8.553350), and I was thoroughly taken by the ability of the SoundPoles to deliver a realistic soundstage. As the information in the owner's manual stated, the presentation did indeed form up somewhat behind the speakers. This impression continued when I listened to three other fine recordings: Muffat's Concerti Grossi, Nos. 7-12 (Naxos 8.55743), Bach's Art of the Fugue (Dorian 90297), and Orlando di Lasso's Lieder, Chansons, and Madrigals (CPO 999855). However, I still sensed a lack of body and fullness in the middle bass region that was not evident when I listened to other speakers. The artifact tended to make instruments such as cellos move further back into the orchestra.

Still, with all of the above releases, the SoundPoles were impressive. I also gave them a try with an excellent jazz release, Brooke Sofferman's The Green Between (Summit 323) and Dire Straits' On Every Street (Warner 26680). With the Sofferman disc, I could detect that lack of mid-bass richness I mentioned above, and with the Dire Straits album I began to notice that the systems were not able to deliver the kind of rock-music punch some individuals might care to hear.

Indeed, it began to dawn on me that these speakers are more at home with lightweight jazz and small-ensemble classical material than they would be with larger-scale presentations or chest-thumping rock transcriptions. It also was clear that their fine soundstaging was very dependent upon the way the recordings were made, with some not being any more impressive than what I had previously heard on a number of other good systems.

It is pretty difficult to do an A/B comparison against other speakers with the Musiko System, because the close proximity of large speaker enclosures will probably compromise their soundstaging. Nevertheless, I did give it a try, using tracks from the above-noted recordings, as well as from sections of my standard A/B comparison disc, Engineer's Choice II (Delos 3512). The pairs were driven by separate amplifiers, which allowed me to get their relative levels fairly closely matched.

The result was revealing, because when compared to speakers such as the Dunlavy Cantatas and AR Phantom 8.3 models, it was easy to see that one's opinion concerning the abilities of the Musiko System would have to boil down to a matter of taste. As best I could tell, the Dunlavy and AR models were richer-sounding and had a perceptibly superior sense of smoothness all up and down the frequency range, particularly the middle frequencies. This edge in smoothness was particularly noticeable with female vocal material. Other listeners might prefer the somewhat leaner sound of the poles, however, particularly with small-ensemble transcriptions.

This was perhaps not fair to any of the speakers, since the Dunlavy and AR models were, as I noted, fight next to poles and probably compromised their soundstaging and spectral balance. Consequently, I did another A/B comparison against my Allison IC-20 models, which were in their usual, low-profile locations against the front wall. This put them six-feet behind the Soundpoles, and pretty much lined up with them in relation to the sweet-spot listening position. In other words, the Soundpoles, six feet out in front, were partially blocking the IC-20s. To keep the EOSS subwoofers from also blocking, I placed them somewhat outboard from their previous locations. The wide dispersion characteristics of the Allison models made the blockage from the EOSS satellites not all that much of a problem.

In this comparison, I again listened to the above program material, particularly the Art of the Fugue, which has a degree of consistency that makes it a good item for speaker comparisons. I also auditioned very fine recordings of Haydn's String Quartets, numbers 4, 5, and 6 (ASV 1084) and Boyce's Pindar's Ode (ASV 232). It was easy to see that the SoundPoles could indeed throw a soundstage between their out-front locations and the front wall. However, the IC-20 systems, six feet further back, could put a wider soundstage in the same relative front/back location. What's more, the Allison models could do this without having to be placed six feet out into the listening room.

In addition, like with the Dunlavy Cantatas and AR Phantoms, the IC-20 systems had a richer and smoother overall sound, in combination with a wider and more spacious soundstage. I even hauled my wife in to do a blind comparison while playing a segment from Vaughan Williams' Second Symphony (Chandos 9902) and she immediately noted that the SoundPoles had a thin, sometimes inarticulate sound that did not do justice to that large-scale classical material. With small-scale stuff they were fine (although still inferior to the IC-20 systems), but with the heavy stuff they lacked a sense of proper spectral balance and body, their fine soundstaging notwithstanding.

I also tried to better pinpoint any possible advantages the EOSS systems might have in comparison to the IC-20 models by doing an A/B comparison of their respective abilities with the first Chesky jazz sampler and test disc (JD37). I compared using the LEDE vertical imaging test and the tambourine depth test and found that although the SoundPoles did a great job, so did the IC-20s.

I am going to conclude here and note that while the EOSS speakers have fine soundstaging and imaging, they are not really exceptional when compared to other good systems that I have auditioned, some of which cost considerably less than the EOSS ensemble. They sound thin and lacking in body compared to those other systems, primarily because of the octave wide dip centered between 160 and 200 Hz. With the samples I reviewed, I could not get rid of that dip, even when I adjusted the woofer low-pass setting as high up in frequency as it could go.

The Musiko System also lacks clarity and detail on certain kinds of music, no doubt because of the less-than-stellar frequency response smoothness. That lack of smoothness accentuates the sound of some instruments and suppresses the sound of others. The bass performance is OK, but to get the sound up to high-end standards in that area a subwoofer would be necessary.

Topping things off is the need to position the systems, including the outboard woofers, well out into the listening room. While this might not bother some enthusiasts in the least, my guess is that it would bother a lot of serious housekeepers plenty. Given that many other systems I have reviewed can do everything the SoundPoles can do, and also do a lot more, I have reservations about them as a sensible purchase for serious music lovers, at least at their list price. These speakers will probably appeal to well-heeled enthusiasts who appreciate their striking appearance and fit and finish as much as they favor their sound.--HF