Pawel/Ensemble PA-1 & Reference loudspeakers Page 4
Because the PA-1 is so revealing of spatial information, it ruthlessly revealed the artificiality of multi-miked recordings. An indifferent studio production like Sergio Mendes & Brazil '66's Fool On a Hill (A&M SPX4160) became unraveled through the PA-1. Artificial reverb and bits of space grafted together, sounded just like that: horrible.
The specter of subsonics
While the lower mids always sounded smooth, I became aware over time of grain and roughness through the upper mids and lower treblebut only with LP program material. Violin overtones lost sweetness and smoothness. A sense of strain would creep in that was not volume-related. The Ensemble PA-1 appeared to be performing better with digital than with LP program material. I decided to investigate this discrepancy using the Wilson Audio recording of the Beethoven Sonata for Piano and Violin, Op.96. David Abel's Guarnerius on the LP version of this was not as pure-sounding as CD. By comparison, the LP was beset with noticeable levels of grain and roughness riding along with the violin's harmonic envelope. The CD also managed to develop a better sense of space, the Guarnerius occupying an almost palpable space within the soundstage. What! CDs sounding better than LPs? I can't have that.
Naturally, I began to strongly suspect subsonic energy as the culprit. With the grille off, it was easy to see the wild gyrations executed by the woofer; the introduction of FM and IM distortion became real possibilities. The fact that the PA-1's passive radiator is tuned high means that it is vulnerable to subsonic energy. The woofer is on its own in the deep bass and below, without any damping from the air spring of the cabinet.
I decided to test this theory. Not having a subsonic filter handy, I introduced the Threshold PCX crossover into the chain, using only the high-pass feed above 75Hz. The deep bass and subsonic frequencies were attenuated at the rate of 18dB/octave below 75Hz. This really did itthe transformation was dramatic. The Guarnerius began to sing sweetly and with excellent focus. All grain and strain were removed from the upper mids. At last, the proper balance was restored: the LP sounded better than its CD equivalent. I missed the lost deep-bass information; there wasn't that much to begin with, but when you have none at all you notice it. But the accrued benefits were so great through the upper mids that I could not imagine listening to the PA-1 full-range again
Apparently, the designer is aware of his speaker's subsonic vulnerability, because in the Owner's Manual the use of a subsonic filter is recommended as it "banishes all those non-musical signals (such as record warps) below the lowest musical spectrum, thus allowing a very clean bass." To be fair, I should point out that most minimonitors, and especially vented designs, would greatly benefit from the use of a subsonic filter. Properly executed, the real benefits of such a filter should greatly outweigh the potential disadvantages of introducing another active device into the signal path.
It turns out that Pawel Acoustics actually manufactures a subsonic filter, of which Bob Graham sent me a sample. I don't know what the US retail for it might be. It consists of a tiny chassis with a small outboard transformer that plugs directly into the wall. From the looks of it, it does not appear to be an all-out design, its circuitry being contained in an encapsulated module that plugs into a mother board.
The frequency response is said to be flat to 39Hz, 3dB down at 25Hz, and 40dB down at 5Hz. The output impedance is very low, so that long cable runs are possible. Besides the obvious benefit of eliminating record warp and cartridge/tonearm resonance energy from the amp and speakers, Pawel Acoustics points out that any subsonic grundge picked up by the master during the recording session will also be filtered out, thus allowing higher dynamic range without stressing the amp or speakers.
Sonically, this is a good- but not great-sounding device, sounding its best after cooking for 24 hours or so. It had a minimal sonic signature, characterized by a slight case of transistor hardness through the upper mids and lower treble. Still, the benefits during LP playback far outweighed the faults it introduced.
Bob tells me that he's hard at work on an "audiophile grade" subsonic filter/crossover unit that would also allow the use of a subwoofer with the Ensemble speakers. I heard a report that the Janis subwoofer works very well with them. This may be so, but to be perfectly honest, I don't really have a burning desire to experiment with a matching subwoofer. The way the PA-1 is balanced, I don't see a subwoofer as an immediate priority. But it's nice to know that you can tackle an Ensemble minimonitor-based system in a modular fashion. First buy the minimonitor, and at a later time you've got the option of adding a matching active crossover and subwoofer of your choice to round things out.
It's ironic that, after spending so much time and effort on the development of a very rigid woofer, designer Harry Pawel opted to use a soft-dome tweeter. I basically never cared much for soft domes. They break up at much lower frequencies than, say, metal domes. To their credit, they are usually well damped, so they don't scream at you the way plastic hard-domed tweeters do. As a family, though, they are not particularly detailed, featuring a fuzzy sort of sound. JA, a genius at assigning sounds to various colorations, describes soft domes as having a "fffff" coloration. And that's probably a lot more colorful than calling them fuzzy.
The PA-1's soft dome did much to change my bottom-line opinion of such devices. It is certainly well-behaved and reasonably detailed; it didn't offend me even after prolonged listening sessions. It's just that I expected more here, especially in view of the magnificent resolving power of the mids. What I missed mostly was the textural smoothness and transient delicacy of the real thing. All you have to do is listen to a good ESL, then return to the PA-1. The PA-1 is then heard to be coarser, less resolving, and without the treble finesse one hears in live music.
I trudged out my old pair of Quad ESLs for just this sort of comparison. (By the way, I find the sound of my old (but Koval-modded) Quads preferable to that of the new Quads, and I would advise those of you lucky enough to own a pair of the originals to hold on to them: they're still one of the best.) When it came to sheer dynamics and guts, the Ensemble PA-1 easily outdistanced the Quad. And when it came to imaging skills, the PA-1 painted a far more convincing soundstage. The PA-1 also had more "air" in the extreme treble. In terms of midrange resolution and transparency, the PA-1 more than established parity, exposing the soundstage to a brighter light source and revealing more detail. However, when it came to the lower treble, the Quad began to outshine the PA-1. Textures were simply more delicate and lifelike. Violin overtones shone more sweetly and soprano voices soared more effortlessly. The Quads continued to occupy center stage in my listening room for a couple more days, before returning to business at hand.
It may be said that I'm being unfair in comparing box speakers with planar designs; a case of apples and oranges, if you will. But the Ensemble PA-1 is good enoughcertainly expensive enoughto merit this sort of comparison.
The Ensemble Reference
The Reference arrived after I had spent several weeks with the PA-1, getting to know it intimately. Visually, there is very little to tell the two apart. With the grille off, the pure cone design of the Reference's woofer is the only distinguishing feature. You sort of expect the two to sound alike, and I was told by Bob Graham to expect a strong family resemblance sonically, but with the Reference being about 25% better. So I really had no pre-warning for the major improvement wrought by the Reference. Don't misunderstand me: the PA-1 is very good, it's just that the Reference turned out to be considerably better, by a factor of two. The degree of improvement I experienced is very likely related to my front end and choice of power amp. I would expect that in different environments the differences between the two Ensembles could shrink to as small a proportion as 25%
It's important to note that, broadly speaking, both Ensemble speakers perform pretty much the same; there is, indeed, a strong family resemblance. It's just that in the areas where the PA-1 excels, the Reference is even better.
The gutsy tonal balance was all there, but without the occasional chestiness of the PA-1. The treble, probably because the Reference's tweeter is hand-selected, sounded a bit cleaner and smoother. The retrieval of low-level detail was just as awesome. The Reference was also capable of pumping out detail naturally without being overbearing in the treble. For example, Reference Recordings' new release, Tropic Affair, lost none of its luxuriant detail through the Reference. But instruments were more precisely localized within the soundstage. Bass lines were also a bit tighter.
The soundstage generated by the Reference was consistently crisper and better defined. There was less confusion about where instruments were located. Massed voices were more readily resolvable. There was a further enhancement in soundstage transparency; soundstage veiling was reduced to the point where the sensation of being able to reach out and touch someone became startlingly real.
The arrival of the Reference coincided with a switch in my phono front end from the Monster Cable Genesis 1000 cartridge to the Rowland Complement. The Complement, a master of the spatial nuance, has opened new doors for me. It makes the Genesis look like a high-schooler trying to compete with a college senior, carving out spatial relationships with a conviction I've not experienced before. The Ensemble Reference kept pace with these front-end developments, faithfully delivering the imaging goodies. At no time did I feel as though the Reference was interfering with or limiting the recreation of the original soundfield.
The Ensemble minimonitors break new ground in that they afford the traditional imaging strengths of a minimonitor with the gutsy balance of a much larger loudspeaker. Both the PA-1 and the Reference mimicked the tonal balance of a large loudspeaker very well. The power range of the orchestra, rich in fundamentals, from about 100300Hz, was reproduced with a weight and power that almost defy belief. The orchestral foundation below about 70Hz is largely missing, but in most cases that did not detract from their ability to paint a convincing full-bodied balance. The Reference was a bit more neutral and less chesty than the PA-1 through the lower mids.
The level of midrange detail these speakers retrieved is competitive with the finest planars money can buy, and is a tribute to the woofer technology promoted by designer Harry Pawel. It's nice, for a change, to see an expensive box with more than inexpensive off-the-shelf drivers.
I'm not enamored with the treble. It's certainly acceptable, but not in the same league as the mids. If you're accustomed to ESL-quality treble, the Ensembles will no doubt disappoint you in this respect.
The woofer proved very linear and dynamic, navigating high-powered program material without stress or strain. It rarely ran out of headroom, and expanded linearly as if it were a much larger piston. After living with the Celestion SL600s for quite a while, I can tell you that it was a relief to come across a speaker than not only imaged better but avoided the congestion that the Celestions are guilty of. The Ensemble minimonitors, like many other vented little boxes, were vulnerable to subsonic frequencies. Vinyl playback was greatly enhanced with the addition of a subsonic filter to the chain. Pawel Acoustic's own subsonic filter was pretty good; Graham Engineering should shortly unveil its own filter/crossover network that will also allow the addition of a subwoofer.
Spatial resolution was outstanding. The Ensembles were capable of recreating an illusion of the original soundfield better than any other speaker I know of. The size of the soundstage and the instruments, and spatial relationships within the soundstage, were recreated with almost palpable conviction. The imaging properties of these speakers, as with other minimonitors, are dependent on placement within the room. The designer's recommended placement compromises imaging in favor of a deeper bass extension. I prefer to locate the Ensembles in the imaging sweetspot in my room, which is well removed from room boundaries. Imaging height is also a factor. The 24" Celestion SL stands worked very well, but I can't wait for Bob Graham's integral stands, which should provide not only the optimal listening height but also better coupling to the stand.
These speakers require considerable break-in to sound their best. A minimum of 50 hours is recommended; be sure to take the time to do so. The treble quality, midrange detailing, and pitch definition all improve as the drivers age.
The Reference is sufficiently better than the PA-1 to justify the price differential. In fact, in my system I would estimate the differences in favor of the Reference to be about a factor of 2. Simply put, the Reference is the best-sounding minimonitor I know of. It is better balanced and more revealing than the competition without becoming hi-fi-ish through the upper octaves. The Ensemble speakers are never overbearing, but extremely musical and natural sounding. The Reference is here to stay in my system. It is a true reference.
If you own another minimonitor, take my advice: sell it. You owe it to yourself to own a pair of PA-1s or References at least once in your lifetime. Your education as an audiophile will not be complete unless you take the Ensemble plunge. You'll be glad you did!