Kharma Ceramique CE-2.0 loudspeaker
OLS has been in operation since the early 1980s, and for me, seeing the depth and variety of their product line was like discovering an unexpected continent of new possibilities. Shortly after my first audition, OLS began showing the Kharma Ceramique speakers at various shows, and were they ever seductive. Not only are they drop-dead gorgeous, suggesting a build quality beyond reproach, but they made me happy to hear them. Each time I repaired to the Kharma room, it was a sanctuary of good music played at sensible (but realistic) levels, and mercifully un–hi-fi. At HI-FI '99 one of my buddies prodded me again and again: "When are you going to get these in for review?"
The OLS Kharma Ceramique CE-2.0 is the middle of OLS's Ceramique line, named for the ceramic-coned midrange drivers. OLS could as well be referring to the cabinets' glistening, nearly crystalline lacquer finish. Like the CE-3.0 two-way and the CE-1.0 three-way, the CE-2.0 is a heavily braced and damped rear-port design with a 35mm-thick MDF outer shell. While the general shape is reminiscent of the Avalon speakers, the Ceramique cabinets have more gently faceted front surfaces that make them seem less formal. Their complex shape and tilted perspective are intended to minimize edge diffraction and floor reflections. Every aspect of design execution and construction is of the very highest quality, as befits the $11k asking price. The sheer grace and glowing beauty of the CE-2.0s made them welcome guests in my listening room.
The CE-2.0's three drivers are arranged vertically on the front panel. The central, ceramic midrange is a proprietary unit sourced from Thiel & Partners in Germany [no relation to the Thiel loudspeaker company in Kentucky—Ed.]. Its very rigid diaphragm ensures pistonic displacement but is supposed to be free of the audible consequences of resonance and breakup that can plague metal cones. This driver covers the 200Hz–2kHz range and is mounted in a subenclosure constructed of MDF mass-loaded with lead sheets. In addition to isolating the MF driver, the subenclosure serves as brace and damper for the main enclosure, to which it is affixed fore and aft with elastomer glue. A 9" Eton driver with a diaphragm made of a sandwich of Nomex honeycomb and Kevlar handles the bass, and the high frequencies are radiated by the low-inductance Revelator dome used in all Ceramique speakers.
A "Subtractive" Crossover Network
All of this represents the state of the art, but it is the CE-2.0's crossover network that really demands attention. Charles van Oosterum, designer/founder of OLS, has used a "subtractive," or series, crossover that differs from the parallel networks that are almost universal today. OLS maintain that the transfer function of their network ensures "one part being the exact opposite of the others"—that is, each driver's output is the exact output complement of the rest of the system. They also say that, unlike high-order parallel networks, a series design is capable of "an ideal step function which sums to unity." [The much-hyped DiAural crossover topology is also a variant of a a series crossover.—Ed.]
In parallel networks, each driver is fitted with an R/C/L filter that restricts its response to a defined frequency passband, and the amplifier drives each driver/filter unit in parallel with the others. Because they are in parallel, each unit has the full signal applied to it, and because each filter sees the low output impedance of the amp directly, operates fairly independently of the others. In series networks, each driver is fitted with an R/C/L filter that shunts out-of-passband energy around it, and all of the driver/filter units are connected in series across the amplifier's output terminals. Because the filters are in series, the signal seen by each unit must pass through each of the other filters.
Series networks are thus more difficult to design: modifying one driver/filter in a series system significantly influences the effective performance of the others, changing the signal and the load it sees. The goal of both approaches is, of course, that the filtered acoustic outputs of all the individual drivers should sum perfectly, on- and off-axis, and in both the frequency and time domains. As Buddy Holly said, "That'll be the day..."
OLS posits that, because the "energy is divided without possibility of mismatching" in a series network, "tolerances in components are thus of negligible importance." This is debatable. I grant that the total signal must pass through and be totally dissipated in the chain of series elements. But, more than with parallel networks, the performance of all the components will change with the variation in value of any component.
OLS acknowledges that the quality of those components remains important, and wires its networks with PP-insulated OFD-copper coils, non-inductive PP capacitors, and proprietary silver/gold cables. The networks are potted and sealed in the base of the enclosures, and the bases themselves contain no metallic elements to affect crossover operation. Even the threaded sockets for the spikes are non-metallic. Just above the bass port on the rear of each cabinet are two gold-plated, insulated OFC WBT terminals. But only two—it's impossible to biwire or bi-amp a series network system.
US distributor Trevor de Maat slipped the gleaming black CE-2.0s out of their wooden crates, carefully fitted their spikes, and walked them into place. He leveled the speakers and adjusted their positions as we recycled the balance/phase tracks on the XLO/Sheffield Test & Burn-in CD until the in-phase voice was sharply focused between the speakers and the out-of-phase voice was completely diffuse. After a quick audition to determine that the speakers had survived their travels unscathed, the gracious de Maat was gone, leaving me alone with the black beauties.
Hooked up to any of the amps on hand, the sound of the CE-2.0 was instantly appealing and recalled those brief auditions at shows. But instead of demonstrating a particular prowess or outstanding feature, the Kharmas seemed almost invisible, the music hanging there in the space between them. The sound was quite full-range, with extended lows and highs as the music demanded. In contrast to the Revel F30 that I reviewed in May 2000, the CE-2.0 presented a flat, integrated response—if anything, it tilted down ever so slightly toward the treble. Overall, the Kharma induced me to just relax and enjoy the performance.
I was completely disarmed by the CE-2.0's pure, unforced high-frequency performance. Nothing ever seemed to be coming directly from the tweeter, yet the high-frequency resolution and balance were superb. From my normal listening position, the treble instruments and voices seemed to appear without mechanical support: they were just there. If I neared the speaker, I got the impression that the treble and midrange shared a single source: the ceramic midrange driver. To localize the treble from its discrete source, I had to place my ear within a foot of the tweeter! The consequence of this was an extremely relaxed tonal and dynamic presentation that sacrificed nothing in sparkle and detail. No wonder I and many others had found the Ceramique demos to be such wonderful respite from the "audiophile" assaults that abound at hi-fi shows.
At the other end of the spectrum, the CE-2.0's bass performance was quite respectable for a speaker costing more than ten grand, and remarkable for its size and its use of a single 9" driver. The low-frequency performance was tight and well-delineated, and I never felt that it was lacking, even when genuine music made substantial demands. Though I tried very hard, I could provoke no audible bass misbehavior from the CE-2.0.
Still, there are limits to what one can expect from such a driver. Unlike the EOS Bass Module behemoths that I reviewed a few years back (in October 1997), the CE-2.0 could not peel paint from my walls with torture discs—but that is pretty nearly irrelevant. As with the Revel F30, bass from acoustic sources was unobtrusive yet satisfying. My only caveat is that tom-toms and plucked acoustic bass had a suggestion of upper-bass/lower-midrange bloom through the CE-2.0.
I had no problem with the resolution or power handling in the ceramic driver's 200Hz–2kHz range—or anywhere else, for that matter—and the Kharma's midrange related as smoothly to the woofer as it did to the tweeter. Nonetheless, the Ceramique's idiosyncratic performance in this area resulted in a slightly warm and reticent overall sound. My ears told me that the ceramic driver itself had a moderate but definite emphasis in the 200–500Hz range compared with its output further up.
My first clue was that, as I put the CE-2.0s through their paces, some subtle aspect of visceral enjoyment eluded me. In the best of times, the music or the sound has the power to stop me in my tracks as I go about my other activities. The Kharmas never snared me as I passed through the music room. Moreover, when I auditioned familiar discs, I rarely got that jolt of presence when I didn't expect it, or even when I did. I know that I've written that I prefer neutrality to euphonics, but the Kharmas were just uninvolving. Whether this is due to the inherent performance of the driver, its interactions with the enclosure, or its phase/amplitude behavior with the crossover, I can't say. (But I betcha John Atkinson can!)
The effect of this response bump on wind instruments and plucked strings was detectable but not particularly bothersome. Bowed violins lacked sheen. Electric guitars and tom-toms had weight and power but didn't pop in the ear. Male announcers on WQXR-FM sounded more baritonal than ever. In mono, they were bottled up in the CE-2.0 box with their voices escaping just from the midrange diaphragm. Male singers were just as near as ever, but lacked immediacy and impact.
If you can imagine it, Hans Theesink's voice on "Call Me" (from Burmester Sampler CD II) was big, but more bluff than buff. Female voices suffered, too. For an analytic example, use any of Rickie Lee Jones' recordings and dissect her voice into three components: the pitched tones, the nasal twang, and the breathy overtones. Only with the proper balance of all three does Jones sound like Jones. Unfortunately, the CE-2.0's low-mid bump sapped the breathy overtones and emphasized the nasal twang. Result: Jones sounded just like Jones, but with a head cold.
The consistent mid-low bias, coupled with the de-emphasis of breath and mouth sounds, made most male and female voices rounder and richer than normal, but less explicit. Even such a gorgeous recording as Lorna Hunt's All in One Day (Classic DAD 1015) was rendered unexciting and a bit plummy over the Kharmas. I once had the pleasure of hearing Hunt sing many of these songs as I sat about as far from her as I do from the speakers in my room, and I know that this 24/96 disc captures her voice and presence with accuracy. With the Kharmas, she was somewhere else; consequently, so was I.
The spatial presentation of the CE-2.0s was very much a consequence of their tonal personality. The soundstage was wide, with remarkable stability and specificity in the lateral dimension. Indeed, I have rarely heard such firm placement of individual voices across the soundstage, even though the overall presentation was relaxed. I particularly enjoyed the breadth and space in Salonen's LAPO recordings of Mahler and Lutoslawski. The tradeoff, however, was that the soundstage lacked much depth and height. All of these effects were, like the crossover's, subtractive and subtle, but they could become disturbing under certain conditions.
When I pushed the CE-2.0s in order to assess their dynamic capacity, the lower-midrange bias very quickly became obtrusive. Sure, they played and pounded real loud with no sign of stress or distortion, but it was too easy to play them louder than I cared to hear them.
All of these problems seemed to result from a tonal bias at the expense of presence. I have no reason to believe that these samples were not representative, and the removal of the fabric grilles had no significant audible influence. I've recently added a pair of double-sided panels in the corners behind the speakers, and another pair of single-sided ones on the walls to catch the first side reflections. Repositioning or even removing these failed to change the Kharma's midrange.
Doctor, Doctor, Give Me the News...
In order to confirm the diagnosis from our signs and symptoms, I switched the trusty z-Systems rdp-1 parametric equalizer into the digital stream. With this transparent device I can subtly alter the frequency response; the crime fits the punishment only if the EQ corrects the perceived tonal error. Just a bit of fiddling was needed to arrive at confirmation and a pretty good therapy: a cut of 2.2dB centered at 315Hz, Q=0.4. The very low Q means that the region of attenuation is very broad, extending into the woofer's domain and nearly up to 1kHz. While these specific parameters may not be the ideal correction, they are as close as my ear and the rdp-1's controls permitted. As JA and others have pointed out, low amplitude, low-Q frequency aberrations are much more likely to create an audibly disturbing tonal imbalance than high-Q aberrations of much greater amplitude.
The effect of this equalization on the Kharma's performance was transforming and, to me, essential. Gone was the slight overhang on bass and drums. Gone, too, was Rickie Lee Jones' head cold. Gone, as well, were the over-polite presentation and flattened soundstage. Now, the system grabbed my attention with its presence and dynamics.
The killer demo of this was Classic's 24/96 DAD reissue of the Everest recording of Antill's Corroboree Ballet Suite (DAD 1029). Talk about lifelike and shimmering detail! Talk about power and immediacy! Exciting music and delicious, spine-tingling sound occupied the huge virtual space between, above, and behind the Kharmas. Had the CE-2.0s sounded like this from the get-go, this review would have been briefer and brimming with Bravos.
Remember the tale of "The Princess and the Pea"? After a night spent atop a stack of 20 mattresses, below the bottommost of which was hidden a single tiny pea, the Princess was asked how she slept. "Oh, very badly!" she said. "Heaven only knows what was in the bed, but I was lying on something hard, so that I am black and blue all over my body. It's horrible!" Being hypercritical goes with being a Princess. Being distressed by a little frequency deviation goes with being a passionate audiophile.
The Kharma CE-2.0 is a wide-range, low-distortion speaker of high quality. A pair of them driven by good electronics will make many listeners very happy. Aside from that midrange bump, the CE-2.0's performance and panache are everything one demands from a speaker in its exalted price range. Others of us would not accept the midrange aberration even in a $500 speaker, although application of equalization confirms the Kharma's superb potential. Like noble Achilles, the CE-2.0 has every strength but one fatal flaw.