Listening #60
I don't want a symphony orchestra in my room: That's crazy. I want their music, played with enough realism that I can hear how it's done. My room is 12' wide by 19' long, with an 8' ceiling, and it's made of wood and Sheetrock. There's no way any sane person would want to sit there and hear the same loudness level they'd hear in a concert hall, in the presence of a full orchestra. Count me among them.
I also wouldn't want to sit in that room and hear quite the same physical scale as a real orchestra —which is neither here nor there, since there doesn't appear to be any way of quantifying such a thing. But because it is possible to measure and quantify loudness peaks and the amount of voltage or current required to reproduce them, some people in our hobby have become obsessed with amplifier power, and many others are overly tolerant of merchants who would have us believe that a nice, neat number of watts can correlate with one's enjoyment of records —that, and not timing or momentum or flow or correctness of pitch relationships or anything else that might actually have to do with music, for God's sake.The appeal of their message is easy to understand: When a consumer is led to believe that the enjoyment of music is quantifiable, he or she is left with the idea that that enjoyment is easy to attain. That conclusion also carries a hope in the existence of audio egalitarianism, of affordable satisfaction. Countless hi-fi companies have been built on little other than the promise of cheap power —more watts for the dollar —along with the suggestion that the more power you buy (from them, of course, since all other manufacturers are notorious liars), the more satisfying your music system will be.
It's rare to hear a thrifty-minded company offer more tone for the buck —or more tunefulness, more momentum, more flow. Not unheard of, but still —rare.
Like other people, I enjoy listening to music systems that can startle me with the strength and suddenness of their dramatic swings —or hold my attention with the sense of sheer physical touch they find in every recording they play back. But from what I've heard in years of serious listening, none of that requires an extraordinary amount of power: just a wisely designed amplifier mated to an appropriate and similarly well-thought-out loudspeaker. My own system, in which a 10Wpc Shindo Cortese amplifier drives 98dB-sensitive Audio Note AN-E loudspeakers, does quite well (footnote 1).
Would more power make my system even better? Maybe. And maybe not. Last month, while preparing a product review for the November Stereophile, I observed that swapping out my Shindo Cortese for an amp five times as powerful made my system sound flatter and less compelling. Nor was that the first such counterintuitively disappointing amplifier "upgrade" I've experienced.
A powerful amplifier may be capable of doing one or more things better than an amp with a tenth as much power, all else being equal. And there may well exist some disagreeable quality in music replay that's more surely banished by a high-power amp than a low-power one, and that I'm simply less sensitive to than other hobbyists —just as I'm less sensitive to being deprived of knowing every performer's precise location on something called a "soundstage." Whatever. The real question, as always, is: Which variables, of the hundreds that exist, are the ones that have consistently proven most important to me, and to my listening enjoyment? Me, me, me.
It's all about me
A memo to the electronics industry:
"I don't want you to sell me more power, any more than I want you to sell me more voltage regulation, more power-supply capacity, flatter frequency response, greater channel separation, or more damping factor. None of those things can guarantee that an amp will be good at the thing it was supposedly designed for —playing music —and no single one of them is necessarily more important than the others. What I really want is for you to sell me more music. When you think you're ready to do that, give me a call.
"But because the finest-sounding amps in my experience have mostly been low-power things, I dare say you'll help your own cause as much as mine by forgetting the watts-per-dollar thing for at least a little while, and concentrating on something else. You needn't worry: There's a growing number of new loudspeaker manufacturers that specialize in high-sensitivity, high-efficiency designs. Thank you."
One of my favorites is DeVore Fidelity —not just because they make very good loudspeakers, but because it's impossible to spend more than a few minutes talking with company founder and chief designer John DeVore without realizing: This guy just plain gets it (footnote 2).
A performer in his own right (on drums), DeVore also has great taste in music: the focus of virtually every conversation he and I have had since we met five or six years ago. During those years, two other things have dawned on me: Most DeVore Fidelity loudspeakers combine higher-than-average electrical sensitivity with sensibly high impedance curves, and most DeVore dealers also carry one or more amplifier brands noted for low power and high music quality, among them Shindo and Audio Note.
Footnote 1: In the technical measurements that accompanied my review in May 2006 of the similar Audio Note AN-E Lexus Signature, John Atkinson observed lower sensitivity than Audio Note had claimed in their published specs: 92.5dB vs 98dB. Still, John described the AN-E as "one of the more sensitive dynamic speakers I have measured."
Footnote 2: DeVore Fidelity, Brooklyn Navy Yard, 63 Flushing Avenue, Unit 259, Brooklyn, NY 11205. Tel: (718) 855-9999. Fax: (718) 855-9998. Web: www.devorefidelity.com.
The opportunity for my first in-home DeVore Fidelity experience came last summer, when John DeVore offered to deliver and install a pair of his brand-new Gibbon Nines ($6500/pair). Not every loudspeaker in the DeVore line is referred to as a Gibbon —the current exception is the flagship Silverback Reference ($15,000/pair), which Mikey raved about in the March 2006 Stereophile —but they all benefit from an evidently unique approach to crossover technology called the Gibbon Circuit, the precise workings of which are a closely guarded secret. "It isn't a first-order, second-order, or third-order filter," DeVore says, "although it uses resistors, caps, and coils [of the usual sort]. It isn't a Band-Aid: Notch filters are offensive. Zobel networks are offensive..."
As DeVore explains, the first step in creating a good crossover circuit is to create good drivers. "Any speaker with a very stiff, undamped diaphragm that rings like crazy will create an enormous spike —and the standard practice is to design a circuit to compensate, and make it flat. The problem is, no matter what you do, the amplifier won't see it as flat." That's one reason DeVore gave the Gibbon Nine a 0.75" silk-dome tweeter, built in Europe to his specifications: It has "response to 40kHz, and no breakup." The other two drivers in the Nine are 6.5" plastic-cone woofers, also designed by DeVore and made for him by SEAS. In the 2.5-way Gibbon Nine, the lowest-frequency driver —the upper range of which rolls off before 100Hz —has a treatment applied to the inner surface of its cone, to alter its resonant frequency.
Back to DeVore's design overview, which he calls the Gibbon Philosophy: "What makes the speaker easy to drive? Sensitivity and impedance are important, sure. But you also have to look at the entire loudspeaker as part of the amp, as part of a complete circuit. I want that interaction to be such that the amp stays as happy as it can, whether it's a Shindo or a Krell." Thus the system requires a well-behaved cabinet to go along with the well-behaved drivers —and DeVore accomplishes the former in a number of ways. The cabinet is made from two different densities of MDF, assembled with three different types of adhesive: "The selection of the glue for each part depends on whether the idea is to conduct resonances or isolate them," DeVore says. Damping treatment also involves three different materials; the result is a cabinet in which every panel exhibits a different resonant frequency from every other.
As to acoustical loading, the Nine is a bass-reflex loudspeaker, but with different ports for the two different low-frequency drivers. The Nine comprises two separate chambers; the upper one is, for all intents and purposes, a complete DeVore Gibbon Super 8. The lower port is longer than the upper one, and is thus tuned to a lower frequency. "Ports put out a lot of energy at a narrow band of frequencies," DeVore says, which can contribute to disturbing the relationship between amplifier and loudspeaker. By breaking them apart, so to speak, DeVore aims to spread out the peaks and minimize their effect on the amp.
The Nine's internal wiring is stranded copper, using what DeVore describes as "very thin conductors, very heavily silver plated." The crossover itself is potted in soft epoxy —as much to guard DeVore's trade secrets as anything else —and suspended within its own separate chamber. The lowest low-frequency driver is recessed slightly on the (inner) side of the cabinet, to provide room for a removable circular grille. And the generously sized support spikes are adjustable, with enough travel that the entire cabinet can be tilted one way or the other by several degrees.
The result of all that work is a narrow, medium-size floorstander (38" tall without spikes) with a nominal impedance of 8 ohms (5.6 ohms minimum), 91dB sensitivity, bass extension down to 31Hz, and a relatively high degree of placement flexibility; in other words, a loudspeaker that seems able to do just about anything.
All for the Nines
The weekend of John DeVore's arrival was my fourth anniversary of living in this house —and, as you can imagine, dozens of other loudspeakers had been here before the Nines. All have been positioned carefully; my hardwood floor bears the marks of countless pointed feet, and the residue of countless pieces of tape that show where each speaker worked the best.
On each side of the room, for each channel, most of those bits of tape seem clumped together in two areas. One is about 5.5' from the front wall and less than 2' from the sidewall, and the other is very close to the sidewall and just 1' away from the front wall.
DeVore tried the Nines in all of those positions and settled on the ones that were well away from the front wall, but with a key difference: He moved the left-channel speaker about 7" farther from the left side wall than is normally the case, and moved the right speaker about 7" closer to its respective side wall. The placement had now become, in a sense, an extension of the speaker's own design, swapping one resonant peak for two less severe ones, and thus spreading our hurdles much closer to the ground.
It took a day to get used to it —I moved my listening seat, even moved my rug off center, so I could make the installation appear more symmetric and thus more audiophile-approved —and I liked what I heard.
So, for that matter, did my spectrum analyzer, which found bass extension that was only 2dB down at 31.5Hz and 6dB down at 25Hz, with the Nines arranged as described (footnote 3). The most notable departure from bottom-end perfection was that the midbass was a bit shelved down (between 40 and 63Hz), including a persistent and at least partly room-related response notch at 50Hz. With the Nines toed-in about 30 ° (I could still see the inside edges of the cabinets from the center of the now off-center listening area), their treble rolloff was the gentle sort I've come to associate with a natural, unfussy sound.
As I discovered weeks later, more bass could be had by putting the Nines all the way back into the corners of the room. But the positions John DeVore selected gave a more enjoyable and less colored performance overall, and let the speakers do their best in a variety of ways while delivering good bass heft. Away from the walls, the Nines could shiver my timbers with the loud orchestral drums in the classic Aldeburgh Festival recording of Britten's Noye's Fludde (LP, Argo ZK1), yet still catch all the timbral complexity and sometimes precarious tunefulness of the chantey-like recorder parts. And they did a very convincing job with the spatial aspects of the recording, from left to right and from front to back, including the almost insufferably cute sound of children, representing cats and mice, scampering from stage right to stage left while singing "Kyrie eleison" at the tops of their lungs. Nice!
The Nines were wonderfully nuanced. Listening to them in place of the audibly more sensitive but undoubtedly more colored Audio Note AN-E/SPe HE speakers, it was easier to hear subtle vocal inflections in particular. On Larry Sparks' classic "John Deere Tractor," from the album of that title (LP, Rebel REB 1588), the Nines made it easier to hear how Larry draws the word John into a descending line of three separate notes —albeit very quickly and subtly, more like Hank Snow than Hank Williams. It was also easier to follow and enjoy Monk's cantabile toward the end of "Blue Monk," from Thelonious Monk Quartet with John Coltrane at Carnegie Hall (LP, Mosaic MQ1-231).
Even at their best, the DeVore Nines weren't capable of sounding as big as the Audio Notes —a quality related to the corner placement for which the latter were designed. (In a good AN-E installation, the corners from which they play tend to function as horns, and the early reflections combine with the obvious room gain to enhance the speakers' sense of scale.) It also took some coaxing for the DeVores to deliver the kind of force that seems commonplace with the Audio Notes. But they did well enough after I put the Miyabi 47 phono cartridge and Shindo Cortese amplifier back in my system —in which case the Nines had a fine sense of touch with both small- and large-scale classical music.
I'm not a huge Artur Rubinstein fan, but I like his recording with the Guarneri Quartet of the Brahms Piano Quintet in F Minor (RCA LSC-2971), and the forcefulness that makes this recording such a good one was put across well by the Nines. The DeVores also left the plucked cellos in the first movement of Solti's great recording of Mahler's Symphony 1, with the London Symphony Orchestra (LP, London CS 6401), with just enough naturalness or randomness of touch, for want of a better expression, that the playing sounded enjoyably human and nonmechanical.
The DeVore Gibbon Nines were, simply, consistently fun to have around: They delighted me as much this morning as they did three months ago. The Nines responded clearly to the differences between various triode output tubes in my Fi 2A3 Stereo amp, and to the changes in scale between the different Shindo amps I've tried. More important, they responded to real music. On their first day here it was Stevie Wonder's uncannily modern-sounding "Pastime Paradise," from Songs in the Key of Life (LP, Tamla T13-34062); this morning it was "Semjase" and "The Bells of Love," from Big Dipper's Craps (LP, Homestead HMS 122-1). Every time out, they've made my music sound involving, impressive, and right.
Not to get too Zen about it or anything, but the Nine's greatest strength was its multitude of strengths —that and the manner in which it dispensed with all of my expectations. Compared to the typical high-sensitivity loudspeaker, the DeVore Nine was more open and less colored, with a significantly greater degree of spatial performance. At the same time, compared to the typical high-end speaker, the Nine was not only easier to drive, it was easier to love. It had more drama and sheer humanity than I've ever heard from such an outwardly conventional loudspeaker, and it never sounded boring or constricted.
To put it more bluntly: Horns and such are a great deal of fun —but if you don't have the money or the space or the patience required for a good horn speaker, and if you're not willing to sacrifice openness and transparency and decent imaging in order to experience the world of very-high-quality, low-power amplification, there is now an alternative. And it's a good one, and it's a nice, neat number.
Footnote 3: In an effort to make my copy sound a little less Peabody-and-Sherman than is usually the case when discussing such things, I've removed all the standard disclaimers —that the spectrum analyzer is an Audio Control Industrial 3050, that the test signal came from its the analyzer's built-in pink-noise generator, that I averaged the results from a range of microphone positions, and that the numbers I've reported are all best-case examples —and put them down here instead.
