Plateau Camber 3.5 loudspeaker
So far, I have yet to hear a Canadian loudspeaker that redefines the state of the art from the top down, but, as Will Hammond of KPFK's "In-Fidelity" radio program has pointed out, the impact on the Canadian industry of having the benefit of the NRC's laboratories to hand results in better affordable loudspeakers. To generalize perhaps a little too sweepingly, for the same price the customer used to pay for a not very distinguished performer, he now gets a very much more competent design, with perhaps some areas of performance hinting at true high-end sound.
Does this contender fit this categorization? We shall see.
In Vol.10 No.7, I reviewed the least expensive model in the Canadian Camber range, the 1.5, and found that, while not perfect, it offered a respectable performance for just $339/pair. The Camber 3.5A ($669/pair) is considerably larger but retains the basic reflex-loaded two-way formula. The tweeter is a 1" plastic-dome unit from the Danish Vifa company, with the dome recessed behind a very short flare and with ferrofluid in the voice-coil gap.
As with all the Camber models, the woofer is made by Camber and is constructed on a substantial diecast aluminum basket for maximum rigidity. The polypropylene cone is driven by a high-temperature, black-anodized, aluminum-wound voice-coil, and the woofer is reflex-loaded by a port, 2.5" diameter and 3.5" deep, offset to the nearside edge of the cabinet. The tweeter is also offset, but to the outside edge, Camber recommending that the 3.5s, which come as a mirror-imaged pair, be positioned this way. This asymmetry spreads out in frequency the deleterious effects of diffraction and reflection from the cabinet edges. Both drivers are rebated into the front baffle.
An unusual amount of design attention has also been paid to the enclosure for what is basically a relatively inexpensive loudspeaker. The 20mm high-density particle-board walls and baffle are braced by two horizontal H-braces, one just below the tweeter, the other just below the woofer, while a third brace is wedged between the rear of the magnet woofer assembly and the rear baffle. As this is oversized by 0.5mm, it pushes the front and rear baffles into a slightly bowed shape, which is said to reduce the level of vibrations in these boards by 90%. The positioning of the braces divides the cabinet walls into three unequally sized portions, to spread the frequencies at which resonances would otherwise occur. Internally, all the walls are covered with 25mm thick acoustic foam to further damp vibrations. The only sign of cost-cutting is the vinyl finish, available in simulated walnut or black: a real-woodveneer finish quadruples the cost of the cabinet to the manufacturer, and the use of vinyl represents a saving which will not adversely affect the sound quality.
The crossover is hard-wired and uses custom-wound inductors, and plastic-film capacitors rather than the usual reversible electrolytics. Slopes are second- and third-order Butterworth. Signal connection is via knurled binding posts inset on a sloping panel on the cabinet rear.
William C. Taylor of California, the winner of the drawing at our Santa Monica show in April, visited Santa Fe during the review period and I called upon his ears for a formal listening test, auditioning these and other loudspeakers. I have indicated in the text where he and I are in agreement and where we differ in our views on the speakers tested.
Plateau-Camber recommends that stands be used; not having the appropriate Camber stands ($69 extra) to hand, I used 16" open-frame, spiked Heybrook stands which placed my ears at tweeter height, the axis I felt to give the best balance between midrange and treble. I usually remove speaker grilles when auditioning. However, Camber has bevelled the edges of the particle-board grilles to minimize diffraction effects and recommends that they be kept in place. This I did, though there was slightly more top-octave "air" without the grille, and the already good imaging improved further.
Taking the good aspects of the sound first, the 3.5s' stereo soundstage was wide, deep, and well-defined. Centrally placed vocalist images were stable with frequency, while the layers of the orchestra, on appropriate recordings, were reasonably precise in definition. The midbass was also well defined for a ported design, with good differentiation between bass guitar and kick drum. The only times the midbass seemed to lose control was when hit with high levels of plucked double bass, when it became too loose. It was also easy for the left hand of the piano, particularly on recordings made with spaced omni microphones, to become too rich. Subjectively, the bass seemed more extended than was suggested by the in-room measurements, with good weight apparent down to the bottom notes of the double-bass and bass guitar, around 41Hz.
The upper bass, however, was less clear, and there also seemed to be a relative lack of weight to the sounds of tenor instruments such as the cello, though, paradoxically, male voice had a little too much chest tone, a gruffness that seems inescapable with bass-reflex designs. Bill Taylor was more bothered by this than I was, feeling that male voice became too "rumbly," though he did comment favorably on the strong low bass.
The midrange seemed relatively low in coloration for what is basically an inexpensive design. A midrange warmth below 1kHz, however, lent trumpet more of a cornet tonality, and also pushed some piano notes forward, making them a little too clangorous. This was particularly noticeable in scale passages that also went above 1kHz, where the forwardness was contrasted against a lack of energy in the region just below crossover.
The treble was the area where I was least happy. Not that it was unpleasantfar from itbut there was a presence-region hardness which made level-setting problematic. Below a certain threshold, strings were a little wiry but acceptable; above that threshold and the sound became too hard. This also accentuated the throat sound of female voice a little too much for my tastes: listening to Kiri Te Kanawa's collection of Auvergne song arrangements by Canteloube (London 410 004-2), I was drawn in to the music by the wide, deep soundstage but was then forced to turn the volume down when the singer entered, so forward was the presentation of her voice. On typically multimiked orchestral recordings, this hardness led to additional confusion and a lack of treble transparency.
The high treble was characterized by a slight emphasis at the top of the penultimate octave, which added both a little sibilance to voice and a not unpleasant sparkle to triangle and tambourine. However, it also emphasized LP surface noise, suggesting that the Camber 3.5A would be a better choice for a CD-based system. Above that region, the response fell off quite rapidly, particularly off-axis.
Camber's 3.5A is a well-engineered and solidly constructed example of the classic two-way bass-reflex loudspeaker, with what appears to be an excellent bass/mid driver and a well-braced, rigid enclosure. Its high sensitivity and easy impedance suit it for use with relatively low-powered amplifiers, though these should have a tight low end to provide adequate control of the mid and upper bass. It offers outstanding stereo imaging capability for its price, and decodes a considerable degree of depth in appropriate recordings. I was less impressed by its Vifa tweeter, however, finding, as I have found in other models using versions of it, that it can be a little coarse in the lower region of its passband, to the detriment of treble transparency, and rather wispy in the bottom of the top octave.
The Camber 1.5 was an easy recommendation, due to its low price. Its bigger brother is harder to assess, due to the presence of some very strong competition in its price class, notably from the Spica TC-50 ($550/pair), Snell Type Q ($780), Magnepan SMGa ($495), Siefert Maxim IIID ($599) and British Fidelity MC2 ($595), all of which have smoother, less aggressive, but more transparent highs. The 3.5 will be less system-fussy than any of these thoroughbreds, however, and will do better with inexpensive ancillaries. It also has better low-frequency extension; if that, coupled with its excellent imaging, is important to you, you should definitely check out the 3.5A.