Dynaudio Accent 3 powered loudspeaker Page 2

The woofer (24-W-100) is nominally a 240mm drive-unit with a plastic cone and a 4" aluminum voice-coil featuring Dynaudio's patented Hexacoil winding technique. This process improves not only the density of the winding within the magnetic gap, but also reliability and heat dissipation. Power handling for 10-millisecond transient bursts is an astounding 1000W. The range between 500 and 5000Hz is handled by a 3" (model D76), soft-dome midrange which is capable of excellent dispersion and dynamics. The tweeter is Dynaudio's time-honored D28, a 1" soft-fabric dome with ferrofluid damping in the air gap. The basic construction of this tweeter is more than 10 years old. Dynaudio is still enamored with this design, despite a parade over the years of "new and improved" domes made of hard plastic, soft plastic, beryllium, aluminum, copper, and most recently titanium composites. I would agree that a soft-fabric dome, done correctly, can be a very musical performer; well-behaved, forgiving, and clearly preferable to soft or hard plastic domes. But such a dome, in my experience, fails to offer maximum transparency and resolution of treble transients. The soft-fabric dome can be very good and easy to live with, but is it still state-of-the-art? I don't think so.

Setting Up
According to Dynaudio, their speakers require a break-in period of several days in order to achieve their full sonic potential. The review samples were broken-in continuously for at least 72 hours, using CD program material at medium volume levels before attempting any serious listening. This, I imagine, would hold true for any dynamic loudspeaker, and I did notice a significant increase in smoothness and treble liquidity after the break-in period. For the audiophile, Dynaudio recommends that the amplifiers be left on at all times. Generally, solid-state electronics sound better when idling continuously, and since the quiescent power consumption is only 30W, that's exactly what I did. Wisely, interconnects are not included as part of the package; it is suggested that the user experiment with a variety of interconnect cables to lock in on a synergistic combination. The Cardas Hexlink interconnects (14' runs) were used throughout my listening sessions.

As usual, my front end consisted of the Threshold FET-10 preamp (I've yet to hear a more accurate preamp—cost no object), the SOTA Star 'table/SME-V arm/Virtuoso DTi cartridge, the Kinergetics KCD-30 CD player, and the Apogee-enhanced Sony PCM-F1 digital processor. Also as usual, most of my listening time was spent with my favorite black vinyl.

Hearing is believing
The first formal listening test involved the Lesley Test (which involves master tapes of my wife's Lesley's singing voice). Some serious problems surfaced even this early in the evaluation. The overall character of Lesley's voice was pretty smooth, but I had a tough time getting her voice to gel spatially. The best integration was achieved with a substantial toe-in of the speakers so that their axes crossed in front of the listening seat. Even here, the sense of image focus was not up to my expectations. Instead of a tightly focused and palpable vocal outline, which is readily achievable with, say, the old or new Quad ESLs, the result with the Active 3 could be described as bloated, somewhat veiled, and imprecise—as if the image was being pulled apart toward each channel. In addition, the lower midrange was significantly colored. My initial guess was a resonance in the range of 200–300Hz that added extra weight and chestiness to Lesley's lower registers. Once triggered by particular vowels, the effect might be described phonetically as imparting an "ah" resonance to the lower registers.

These findings were disturbing enough that I decided to return to the Lesley Test after several days of amplifier warmup, just to confirm my findings. The chestiness was still there, as was the reduction in image focus. It occurred to me that the focus problem might be related to absolute phase reversal. (Incidentally, I did check the relative phasing of the two channels and found them to be correct.) There was no easy way to verify this because neither the Threshold FET-10 nor the Active 3 possess a polarity-reversal switch. However, as I discovered subsequently, focus was consistently poor with a variety of program material, both digital and analog. Therefore, it is unlikely that the problem is attributable to polarity reversal. In fact, even in summed-mono mode I was unable to get a tight central image.

Digging deeper, I uncovered more dirt. In a supplicant posture before the speakers, on hands and knees, I observed a significant change in tonal balance as I moved my head vertically up and down the front baffle. There was even a sweet spot, precisely ear-level with the midrange driver, where much of the aforementioned chestiness disappeared. This finding points an accusing finger at off-axis acoustic interference between the drivers. Even more disturbing was the finding that the two channels did not sound exactly alike. One channel sounded better than the other, the midrange and treble being smoother, with reduced levels of that lower-midrange coloration.

Analog Listening
The analog listening session started off with the mandatory Test Record 1 by Opus 3. The depth perspective appeared to be compressed. Transparency was just OK, and instrumental outlines within the soundstage were unfocused. I still liked the tonal balance of these speakers, but they were sure having a tough time fleshing out a soundstage. There is plenty of centerfill, but extreme right and left information clings too much to the speakers, interfering with the realization of a seamless soundstage. The chorus on Laudate! (Proprius 7800) was portrayed with broad brushstrokes; it was difficult to pinpoint individual voices. The size of the hall was difficult to visualize, as was the location of the back wall. The treble quality, at least at moderate volume levels, was pretty natural—smooth and spacious, without a hint of shrillness. However, as was evidenced on Belshazzar's Feast (EMI SAN-324), the treble can turn grainy when the speakers are pushed hard. Normally, speakers with an impressive dynamic range manage to convey the power of the orchestra effortlessly without compression or stress. In this case, however, the results were mixed because tweeter resonances, and a slight glare through the upper mids, both of which were audible at high volume levels, conspired to produce a sense of stress when the going gets tough.

Halfway through this session, I was compelled to start experimenting with speaker placement again, so bothersome was the imaging. I patiently varied both the listening position and the toe-in angle, trying to find a combination of the two that would snap the soundstage into focus. I'm sad to report that this never happened, although the experiment did yield some improvement.

Nor did the midrange fare all that well under close scrutiny. Violin overtones sounded slightly dry and grainy. I also noticed a slight glare in the upper mids. Resolution of inner detail was decent, but generally textures lacked the smoothness of, say, a decent ESL.

It struck me about this time that, in many respects, I was reacting to and describing the sound of an average solid-state amplifier. The upper-mid glare, the losses in transparency and spatial resolution, all of these could very well be solid-state irritants.

A pretty good GESR
This is a true story. J. Gordon Holt, our resident guru and Golden-Eared Subjective Reviewer (GESR for short), popped in during a recent visit to Santa Fe. Here I was sitting at my PC, painstakingly trying to flesh out this review, when JGH interrupted this delicate process. After I explained what I was doing, JGH quickly glanced over the speakers and said, "Here's what I think they sound like: the treble is somewhat soft at low volume levels, becoming grundgy when pushed hard; the mids are veiled, and the bass is pretty ill-defined." I was amazed! Two out of three, without ever listening to the speakers! Well, according to JGH, a soft dome plus active solid-state electronics added up to that sort of mental image. Not bad for an old GESR (footnote 1).

Negative reviews are much more difficult to write than positive reviews. The trick is to place the flaws in perspective and weigh the plusses and minuses in a balanced fashion so that the prospective buyer at least has a clear-cut picture of what the product can and can't do. In view of the asking price of the Active 3, any major flaws are, however, simply unacceptable. I'm tempted to say that in this realm of competition a product should not have any significant weaknesses. This is not strictly true. All of our Class B recommendations fall short in some respects, but at least they offer significant slices of reality. The Active 3 fails that acid test. Except for its bass quality, tonal balance, and dynamic range, there is essentially nothing else that I like about it. Spatial resolution and soundstage cohesiveness are disappointing for a speaker in such an exalted price range. There are some significant midrange colorations, and the treble does not hold up well during loud passages. Nor are midrange transparency and liquidity competitive with other speakers in this price class.

Many of the Active 3's active irritations may be directly traceable to its solid-state electronics. If that's the case, then it is a question of inherently good drivers sabotaged by indifferent electronics. The sound of the Active 3 does appear to be dominated by solid-state sound, and as such I'm not about to invest over $5000 per side just so I can enjoy the sound of what appears to be a mediocre $1000 solid-state amplifier.

Footnote 1: I hasten to add that the venerable JGH does listen to products before he commits himself to print with opinions on their sound qualities.—John Atkinson
Dynaudio North America
1140 Tower Lane
Bensenville, IL 60106
(630) 238-4200