Focal Aria 5 loudspeaker kit Page 4

And sing it did, and much more sweetly than before. The upper mids this second time around were sweeter and timbrally more accurate. The Lesley Test, track 13 on Stereophile's Test CD 1, represents an excellent upper-midrange test (at least for me). Lesley's vocal sweet spot is very much entrenched in this region. Thus, any upper-midrange aberrations in the reproduction chain are clearly highlighted by timbral inaccuracies through Lesley's upper registers. Much of the sense of space and palpability inherent on the master tape has been obliterated from this track. I'm not sure why, but at least the final version of the CD has got Lesley's timbre quite right.

The older Aria had sounded dark and dry through the upper registers. The latest Arias managed to come much closer to reality, and the essence of Lesley's voice came through unscathed. In general, the later Aria painted string overtones and soprano voices with a sweeter, lighter sheen. The old Aria was distinctly darker, more brooding in character, which made it all the more difficult to put up with the treble grain and sizzle.

JA measured the new pair of speakers at the same position on the identical axis, in order to compare with the old. Fig.17 shows the MLSSA-derived anechoic response of one of the new pair (top) compared with one of the old pair. (The pair-matching of the new pair was excellent, he felt.) The measured improvement through the entire treble range offered by the new speaker is obvious, as is a superbly smooth midrange.

Fig.17 Focal Aria 5, anechoic response on tweeter axis at 48", corrected for microphone response, of D'Appolito sample (top) and of original sample (bottom).

Finally, the lower-treble grain and roughness that so irritated me earlier was somewhat reduced. The level of grain went from coarse to fine sandpaper. But—and here we come to the Achilles heel of this design—the lower treble still sizzled. Therese Juel's sibilants (Opus 3, Test Record 1) splashed out of control. Soprano upper registers were tinged with an overly bright character. Nylon-stringed guitar's natural brightness became accentuated. The treble brightness became a glare, then a scream during loud passages. There were moments when I was captivated by the glory of the midrange, only to be irritated a few seconds later by the relentless quality of the upper octaves. During quiet passages there were always moments of great joy when the treble coloration was sufficiently low to be masked by the program material. But as soon as the music got going, the treble intruded.

Further experimentation with speaker placement again confirmed the wisdom of listening to the treble driver off-axis: using a significant toe-in allowed the tweeter axes to cross over well in front of the listening seat. Still, the sizzle and grain continued to annoy me. The bright lower treble also cut into the long-term listenability of the Aria. Fig.18 shows the "waterfall" plot for the new sample of the Aria 5. The treble hash remains, while there is some resonant overhang at 4.5kHz which might correlate with my feeling of brightness.

Fig.18 Focal Aria 5, D'Appolito sample, cumulative spectral-decay plot on tweeter axis at 50" (0.15ms risetime).

To judge from the popularity of bright tweeters, it seems to me that bright and sizzly treble is very much in demand these days. Treble sizzle is a sensation that should be in the public domain as plain as that of wind on your face. There should be no mistaking it. Yet many audiophiles actually like it hot. Is it a case of shifting cultural tastes? Have we been conditioned by electronic music, with its ability to produce copious amounts of treble, to actually crave it? Is excess treble (footnote 1) a "sugar fix" for some folks? Let me go on record as rejecting all forms of hot, sizzly treble on musical grounds. I'm unimpressed with Kevlar or any other plastic dome that breaks up in the audible bandwidth. In the name of high-end audio, we should be able to do better than that.

Will the real Aria please stand up?
The gap in performance between the two samples of the Aria 5 raises a serious quality-control issue. Which sound quality can you expect? Being a kit, the drivers and crossover parts or boards you receive are not likely to have undergone any testing or matching. Will your Aria be sonically behind door #1 or #2? Certainly, the improved bass detailing could be explained on the basis of the second sample's better cabinet. The maple enclosures are available for a premium of about $100 compared with the standard MDF offering, and many speaker builders would opt to build their own cabinets anyhow. But what about the sweeter mids and smoother lower treble elicited by pair #2? Here one would have to point a finger at possible driver differences, internal wiring and crossover part choices, and crossover component tolerances. The care with which the crossover components are matched for tolerance is important—especially for complex networks. Rarely are caps and coils sold off the shelf better than ±10%. For fourth-order networks it is critical to match components to at least 5%. I would imagine that suppliers would be reluctant to do so themselves, and that this task will fall on the kit-builder's shoulders.

It should be realized that although its builder saves a considerable amount of money, any kit-building project presents inherent quality-control problems, some of which may be beyond the average builder's ability to solve. This is a risk you'd better be prepared to take. This caveat can be leveled at any advertised kit. I would hope that fully tested crossover boards would at least be available from Focal America.

To label the Aria 5 a sonic mixed bag would really fail to capture the true flavor of this design. It's more a case of an otherwise excellent speaker with a significant fault: a Mona Lisa with too big a nose. The ultimate question to ask yourself about the Aria is: can you focus on its smile without being distracted by what I feel to be a blemish in its balance?

There's no question that the design succeeds in serving up exceptionally fine imaging within a spacious and highly transparent soundstage. Hall ambience is clearly resolved, and image outlines are floated within the soundstage with almost palpable focus. The midrange is sweet and detailed, and eminently capable of communicating the joy of music. That this is accomplished without sacrificing an authentic tonal balance through the power range of the orchestra is all the more amazing. I can't think of anything else under $1000 that can touch the Aria's strong suits.

If only that T90K tweeter could fly! As things stand, it sinks the ship for me. The bright, sizzly treble balance is bad enough. But couple that with a grainy textural quality, and you have a tweeter that I would not let into my house.

There's a lot of potential lurking within the confines of the Aria 5. With a soft-toned tube amp and an extreme off-axis listening seat, it's possible that many out there could learn to coexist with the Aria's treble. Personally, I would very much welcome a second iteration of the Aria 5 sans the Kevlar dome.

Footnote 1: I believe bad digital sound is also reforming the sonic tastes of listeners. I've talked to so many who wax rhapsodic re. the "wonderful brightness!" of CDs.—Richard Lehnert
Focal America, Inc.