"In-Wall Speaker Test, Part 2"

Within the audio industry, but especially on the sales floor, getting an in-wall speaker to sound competitive with a conventional freestanding speaker is seen as a task somewhere between an uphill battle and an impossibility. There's the problem of the acoustically errant wall elements—wood studs, gypsum board and thermal insulation—smudging and blurring the sound by "singing along" out of sync and out of tune with the drivers themselves. Then, at an even more basic level, there's the hole problem: Unless you're willing to give a wallboard hanger full-time employment patching your walls, you pretty much have to settle on a location for the speakers before you've actually heard how they sound there—a restriction that doesn't apply to freely movable bookshelf or floorstanding speakers. And then there's the speaker's physical location at a room boundary, something bound to alter the bass response.

These and other hurdles have led some speaker designers to abandon any plans their marketing departments may have had for "making a killing" in the in-wall market. Richard Vandersteen summed up the feelings of many when he declared, "I'll start making in-wall speakers when Steinway starts making in-wall pianos."

Doomed?
These problems, and there are others, are real, but not insuperable. Let's take them in order. First, the in-wall speaker's enclosure doesn't have to be fashioned from acoustically inferior materials. Wood studs come in sturdier two-by-six and two-by-eight varieties; real speaker-box materials like medium-density fiberboard (MDF), birch plywood, constrained-layer damping sheets and synthetic composites like Corian can be used in place of flexy wallboard; a layer of gypsum board can be applied over the top to provide a surface that can be taped, textured and painted. The cavity itself can be strategically braced and fitted with padding to control internal low-frequency standing waves and stop higher-frequency enclosure reflections from bouncing back through the woofer diaphragm. (I hope to present the results of my experimentation with various wall/enclosure construction in these pages soon.)

The hole problem only underscores a recurrent theme in the "Well-Tuned Room" series, namely, that some thought should go into the placement options before the speakers are specified and purchased. For example, if a high mounting position is the only alternative, a speaker with an especially good vertical off-axis response should be sought. While in a laboratory context it's true that some in-walls outperform others, in the real world of walls, windows and furniture one "pretty good" in-wall will often prove to be more acoustically appropriate, and therefore sound better than an otherwise "superior" speaker.

The allegation that the in-wall speaker's coplanarity with a major room boundary fundamentally dooms it to poor sound is serious and common. One high-end speaker manufacturer is fond of demonstrating the charge by slowly moving closer to a wall while talking; his voice, of course, tonally grows boomier and more "colored" as he moves to within a few inches of the wall surface. The conclusion drawn from the demonstration—viz., that there is no worse placement for a speaker than at a room boundary—is patently bogus. It is true that, whether live or reproduced, when the lowest frequencies hit a nearby wall they fold back into the room in phase, adding an unmistakable bass emphasis-often heard as boominess.

Indeed, if you were to try to install into a wall a conventional speaker whose response was tuned for freestanding placement, that is, well away from walls, you would hear a noticeable reinforcement (roughly 6 decibels) in the mid- and low-bass range. But no credible manufacturer of in-wall speakers uses a 4-pi (free space) tuning for a product destined for a 2-pi (hemispherical) environment. Further, all other things being equal, an in-wall speaker enjoys a decided advantage over non-in-walls precisely because one of the biggest variables in the performance of freestanding speakers—namely, the woofer's relationship to room boundaries-is replaced by a certainty that can be factored into the speaker's response in the design stage.

This ability to make the wall work for, not against, the speaker is not a trivial or academic advantage. Whereas a freestanding speaker might, in an anechoic chamber, achieve response that is flat, ±2 decibels, down to 25Hz, the same speaker's bass response in a real room will show peaks and valleys of up to 18 decibels, depending simply on where it's put. Even placing a freestanding speaker against the wall does not offer relief: At long wavelengths (i.e., low frequencies) the rear-radiated sound will bounce back in phase, reinforcing the bass, but at shorter wavelengths (higher frequencies) the direct and bounced-back sounds grow out of phase, until the very opposite of reinforcement occurs: cancellation. (This is typically seen as a dip in the 150-200Hz region.) All of this makes a roller-coaster out of the speaker's advertised "flat" response, and helps make the case for the in-wall speaker.

a/d/s/ C700i/r
The C700i/r features an 8-inch woofer and a ribbon tweeter approximately 7 inches long by 3/4-inch wide. The ribbon's placement to the side of the woofer—not above it—makes this a/d/s/ model the only entry in the survey that, when installed, is wider than it is high. Neither the owner's manual nor the current Audio equipment directory provides frequency response or sensitivity information. The owner's manual does state that the speaker should be oriented with the ribbon running vertically, and placed so that its center is 48-52" above the floor.

Other than the extra weight contributed by the ribbon's large magnet structure, the C700i/r installed much like a conventional in-wall. The bezel/mounting frame is not as robust as some, and I had a bit of trouble getting the perforated metal grille to seat properly, but these were minor problems.

On audition, the overall impression left by the C700i/r was one of crispness and clarity, especially at low to moderate playback levels, and a full, but slightly heavy upper bass. The speaker generally handled voices quite well, and did a credible job of recreating the church ambience in the medieval Cant de la Sibilla selection. The Parkening (classical guitar), Prokofiev (orchestral), Pogorelich (piano) and Bartoli (vocal and piano) tracks were particularly successful, portraying an attractive tonal balance and good spaciousness.

A trace of excessive sibilance detracted from the Mary Black number, a little boominess slipped into the Lyle Lovett vocal, and a touch of hollowness crept into the winds on the El Cancionero de Palacio track, but on balance, the speaker's tonal colorations were relatively benign. Driven hard on the Chris Rea and John Hiatt cuts, the C700i/r loses some of its trademark crispness on snare drums and cymbals, suggesting that the ribbon has trouble keeping up with the woofer above 95dB or so. The listening tests confirmed the need to mount the speaker at the recommended height: If installed much lower or higher than ear-level, the ribbon's output drops off considerably, leading to a dull, lifeless sound.

The unsmoothed on-axis responses appears fairly well behaved, just missing a ±4dB window throughout its range due to a peak at 140Hz and a broad, but slight, elevation in the 2-5kHz region. It's this latter "wrinkle" that's probably responsible for slightly highlighting sibilance. With one-third octave smoothing, the C700i/r measures flat, ±3dB, from 160Hz to 10kHz.

The averaged off-axis response rolls off at a mild 3dB per octave; note that this doesn't show the true picture, which is that the tweeter's lateral response stays relatively flat out to 15kHz-even at a remarkable 60 degrees off-axis-while the vertical response begins to drop off fairly steeply as soon as the ribbon takes over at around 2.5kHz. This confirmation of the broad horizontal and narrow vertical coverage pattern reinforces the manufacturer's insistence that the speaker be installed in "landscape," not "portrait" orientation. The 1-watt/1-meter sensitivity would appear to be in the 87-88dB range.

Apogee Ribbin-Wall
Apogee's entry utilizes a 6.5-inch woofer in a sealed enclosure and a 3/4-inch wide, 26-inch high ribbon tweeter. A three-position switch is provided for adjusting woofer output; the ribbon tweeter's mounting arrangement allows for a modest amount of aiming both horizontally (i.e., toeing in/out) and vertically (i.e., canting up/down). Frequency response is given at 40Hz to 20kHz; no sensitivity spec is provided.

My installer found the speaker to be well designed, solidly constructed and easy to install.

The listening tests made one thing strikingly clear from the first track: The Ribbin-Wall focuses treble energy in a very narrow pattern vertically. If you aim the ribbon so that high frequency coverage is strong at the seated listening position, you will almost certainly hear a pronounced drop-off of treble when you stand. Interestingly, the vertical "window" seems about 12 inches high-about half the height of the 26-inch ribbon itself. In the "sweet spot," the Ribbin-Wall did a creditable job with all program material. From the violins all the way down to the tympani, orchestral balance and instrumental definition scored highly in the Prokofiev, and even the pedal-to-the-metal rockers by John Hiatt and Chris Rea came off well. A mild case of hollowness was heard on the winds in the Cancionero selection, but otherwise the Apogee didn't stamp the music going through it with a strong personality. Moving left and right in the listening room didn't alter sound quality appreciably; moving up and down altered it dramatically, and for the worse.

The measurements don't look quite as good as the speaker sounded. Ignoring a sharp notch at 4.1kHz—probably the result of cancellation off the cabinet edges near the ribbon—the Ribbin-Wall needs a ±6dB window to accommodate its 100Hz to 18kHz on-axis response. As for the off-axis components, response shapes at 30 degrees and 60 degrees laterally bear a very strong resemblance to the direct, on-axis sound, while the 30 and 60 degree up and down measurements show a marked drop-off-typically 10 to 20 decibels-in the treble.

These measurements confirm the listening experience, which found a very narrow vertical window for good sound quality. The Apogee Ribbin-Wall might be thought of as a good choice only for applications in which the listeners are always seated and in-line with the ribbons. But I suspect that they could prove ideal in more casual environments, too, where the sound needs to be sharp and focused for those seated, and less sharp and attention-demanding for those who are walking around and through the room. A sensitivity rating of about 86dB for a 2.83 volt input at 1 meter would seem warranted.

Bohlender-Græbener Radia RD75
The RD75 is what NBA guards used to be: quick, skinny, and about 6'3" tall. The 1.25-inch wide ribbon has a rated frequency response of 125Hz to 18.5kHz, ±3dB; sensitivity is specified at 87dB/1 watt/1 meter. The manufacturer sees the RD75 not as a "full-range" speaker, but rather as a "satellite" to be partnered with a subwoofer. Bohlender-Græbener offers several subwoofer drivers for in-floor mounting. These were not practical for my tests, so I opted to augment the RD75 with the Bag End ELF processor and 18" subwoofer instead, driven by a second Soundstream DA-2 power amp. Note that the Bag End subwoofer's contribution is not included on the accompanying graph.

The RD75 installed relatively easily, aided by a well-written and well-illustrated manual.

Without a subwoofer, the RD75 sounded almost unlistenably thin. I wasted no time in connecting the aforementioned Bag End system, setting the crossover for 125Hz, and calibrating the result with the microphone and TEF; the whole exercise, which took about an hour, proved to be the tonic that was needed. The drums from the Cancionero disc sounded clean, with no perceptible smear or overhang. The Cecilia Bartoli vocal proved silky and captivating, making even the best conventional in-walls sound colored and closed-in, although her piano accompaniment lacked some of the characteristic lower-midrange and upper-bass richness needed to convey the sonic grandeur of the instrument. A similar reticence in the "richness" part of the spectrum held the Prokofiev, Christopher Parkening and John Hiatt tracks back; my notes likened the effect to trying to mash down on the pedal with the parking brake still engaged. This effect was especially noticeable in the lower midrange, an area that needs to track dynamic swings faithfully if many instruments are to "speak" convincingly, and gave the speaker a bit of a "high-class background music" feel.

Despite this disappointment (and after all, you expect big results from a speaker over six feet tall) the RD75 did some things remarkably well, and especially at volumes below 90 decibels at the listening position. As advertised, sound quality proved largely independent of listener position and room acoustics; it seemed to flood the entire room with sound. (Physics accounts for the effect this way: Whereas the sound pressure level from a conventional point-source speaker is inversely proportional to the square of the distance, a line-source like the RD75 generates SPL that is inversely proportional to the distance—at least up to a distance of about twice the height of the line-source.) The resulting evenness of the coverage pattern horizontally is not a trivial advantage; whether it offsets the somewhat softened sense of dynamics will be up to the tastes of the prospective purchaser.

The on-axis frequency response is somewhat flatter and smoother than the Apogee's, its logical competitor. The basic response profile holds up quite well off to the sides, although by 60 degrees left and right the high frequencies are beginning to roll off-an artifact of the width of the ribbon. (The Apogee ribbon is narrower, yielding a still-broader coverage at very high frequencies.) The Radia is so tall—over 2.5 times as tall as the Apogee—that it beams vertically at all frequencies from the midrange on up. (Unless you listen perched at the top of a ladder, or are taller than one of those NBA player, you'll never hear the beam effect.) This lack of up-and-down coverage will prove to be a decided advantage in rooms with hard floors and/or reflective ceilings. The 87dB sensitivity rating is only slightly inflated; 85dB would seem more descriptive.

B&W Signature 7
One of the more high-tech-looking speakers in the survey, the Signature 7 deploys a 7-inch woven Kevlar woofer and 1-inch, fluid-cooled metal dome tweeter, both patterned on the drivers found in the company's freestanding $8,000-per-pair Silver Signatures. Rated frequency response is 40Hz to 20kHz ± 2dB, and sensitivity is given at 88dB for a 1-watt input at 1 meter. A three-position tweeter level switch is provided on the baffle to allow compensation for overly bright or dead rooms. An onboard "APOC" circuit protects the drivers from potential damage when overdriven.

My installer described the Signature 7's front baffle as "awesome"—stout, nonresonant and superbly finished—and found installation easy though a little more time-consuming than average on account of the tab-type mounting design.

The adjective that kept popping up in my listening notes of the Signature 7 was "natural." The relative freedom from "spotlighting" certain instruments at the expense of others fostered a balanced, likable sound from all test tracks. There were no tonal idiosyncrasies to detract from Cecilia Bartoli's warm mezzo-soprano, or Ivo Pogorelich's big piano tone on the Scarlatti sonatas.

The B&W's even-handedness enabled a minor, but telling, discovery: On Lyle Lovett's "North Dakota" it only became clear that the backing vocalist was Ricky Lee Jones when I switched to the B&Ws. The Signature 7 sailed through the Prokofiev orchestral "torture track" with barely a stutter: Rapid-fire tympani strokes emerged crisp and clean, brass surged convincingly, and strings glowed with the sweetness and woodiness of the real thing.

Only during the most vigorous crescendos did the Signature 7 lose its footing, taking on a bit of congestion and a trace of steeliness. This effect was also heard on the rock tracks by Chris Rea and John Hiatt, which grew coarse and shouty when playback level approached 100 decibels or so—admittedly louder than most users will probably ever assay. I suspect that the Signature 7 is limited more by its gypsum board enclosure than by any intrinsic shortcoming in the woofer, tweeter, crossover, baffle or fittings. A boxy, "gyp-board-y" coloration tethered La Capella Reial's baritone and bass singers from soaring majestically with their tenor, mezzo and soprano counterparts in the 10th century "Cant de la Sibilla" or "chant of the Sibyls."

The test data confirmed the essential neutrality of the Signature 7, with a frequency response that, aside from a slight bump centered at 150Hz (probably attributable to a standing wave inside the 4-foot-long wall enclosure) fully met its manufacturer's ambitious ±2dB specification from 60Hz all the way out to 18kHz. This is commendable performance by any loudspeaker standard, whether in-wall, freestanding or bookshelf. The claimed extension to 40Hz was not met within the ±2dB window; response at 40Hz measured 81dB, or 7dB below the 88dB reference—still a respectable figure—while output at 30Hz, specified at -6dB in the owners manual, proved to be decidedly optimistic, with the measured result coming in at almost exactly -12dB.

Interestingly, B&W issues a specification for the Signature 7's off-axis response, viz., within ±2dB of the on-axis response within 40-degree lateral and 10-degree vertical arcs. And indeed, the off-axis average tracked the on-axis response remarkably closely: Even by expanding the specified horizontal arc to 60 degrees (i.e., 30 degrees left and right of the frontal axis) the maker's ±2dB tolerance was met from the deepest bass all the way to about 9kHz. Even the vertical off-axis behavior—especially important in installations where the speakers are mounted 7 feet or more above the floor—shows remarkable fidelity to the on-axis profile. These measurements suggest that, within reasonable limits, the speaker will sound substantially the same throughout a typical listening area. The manufacturer's 88dB sensitivity spec would appear to be warranted.

Canton InWall 9
Like the KEF CR200Qr reviewed in Part I [February '96], the InWall 9 is a coaxial design, in this case nesting Canton's German-made 1-inch metal dome tweeter in the center of a proprietary 9-inch woofer. Frequency response is given as 34Hz to 22kHz; sensitivity is rated at 89dB/1 watt/1 meter.

Patrick found the Canton to be an unusual product to install, with the driver unit attaching directly to the gypsum board without a conventional mounting frame. Beefier wire termainals would have been welcome, but the speaker otherwise mounted tidily and achieved a good fit and seal.

On audition, the InWall 9 proved to be a mixed bag, with some tracks—including, most notably, the Sting, Counting Crows, John Hiatt and classical guitar cuts—brought off as energetic and involving, while other selections fell decidedly flat. The Mary Black cut showed an unmistakable nasal coloration, and a decidedly hollow sound marred the brass passages on the Prokofiev orchestral track. The drum on Cancionero showed some blurring, and Montserrat Figueras' soaring vocal on Cant de la Sibilla took on an unattractive piercing quality. Chris Rea's "Auberge" rocked along stirringly, but before the tune was over the in-your-face vocal presence had become fatiguing. Like In-Wall Audio's Vörtex, the Canton model proved adept at generating an infectious "party spirit" but came up short on tonal and spatial finesse.

Looking at the accompanying graph, the InWall 9 needs one of the largest windows-±11dB-to contain its unsmoothed on-axis response, the result of a strong peak at 4.2kHz and a deep, though thankfully narrow, notch one octave higher, at 8.4kHz. With one-third-octave smoothing [not shown], the window can be reduced to a more reasonable ±6dB, although even this charitable act hardly produces a graph that would qualify the InWall 9 as a "low coloration" device.

The comparison to the aforementioned KEF is instructive: Both of these coaxial designs produce a similar spectral shape, with a small peak around 700Hz, then a more pronounced one in the 1-2kHz area, followed by an even higher one around 4.2kHz. I suspect that these shared anomalies arise from destructive interference between the wave energies individually radiated by the tweeter and its surrounding woofer cone. When playing music, as opposed to test signals, the movement of the horn (i.e., the woofer cone) at low frequencies may well exacerbate these spectral problems, as the movement can be substantial compared to the short wavelengths handled by the tweeter. Like the KEF, the Canton proved somewhat less erratic off-axis, although the aforementioned peaks remain. The 89dB sensitivity spec was very nearly met; 88dB would probably be represent a more realistic broadband average.

In-Wall Audio Vörtex
Unlike the vast majority of in-walls, the Vörtex is supplied with an enclosure that gets sealed up into the wall. Constructed like a freestanding speaker of medium-density fiberboard (MDF) and internally braced to reduce panel resonances, the Vörtex employs audiophile-grade, Focal-brand drivers: a 1-inch Kevlar inverted-dome tweeter flanked by a pair of 6.5-inch woofers. The purchaser has a choice of 100 fabric or paintable grille options. Frequency response is given at 58Hz to 20kHz; sensitivity is quoted at 92dB, presumably with 2.83 volts input and measured at 1 meter.

Received without mounting instructions, the Vörtex proved relatively time-consuming to install. Actual construction quality appeared to be excellent, with the exception of the grille, which had a tab break off during installation.

Clearly aimed at audiophiles looking for freestanding speaker performance in an in-wall package, the Vörtex initially turned in somewhat mixed results in the auditioning sessions. After the second track, it began to remind me of the long-discontinued JBL L-100, the mega-selling king of "rock" speakers back in the 1970s. The up-tilted treble balance emphasized the shawm on the El Cancionero de Palacio selection, and made Cecilia Bartoli's piano accompanist sound like he was playing a spinet instead of a grand. Similarly, voices on the Mary Black track were too forward, upsetting their relationship to instrumentalists.

There were other tracks, however, where the Vörtex's stylized presentation worked magic. "Auberge" very nearly drove me to a rare bout of terpsichorean abandon, and the Counting Crows track sounded much like what I suspect was heard in the band's studio. The rhythm section on the John Hiatt rock track propelled it forward with conviction, with tight kickdrum and crisp snare, and a vocal line that grew shouty only when the decibel count hit the century mark.

As a general matter, male voices (Lyle Lovett, Sting, and the tenors, baritones and basses on the Elgar and Cant de la Sibilla selections) came across convincingly, while female voices suffered from a too-forward balance that introduced a "pinched" coloration. These effects were mild but consistently audible. If your tastes include a healthy dose of male-vocalist led pop, rock or country, and you're looking for an in-wall speaker that excites, the Vörtex could prove irresistible. My lone caveat would be that the broad depression in the 1-3kHz range may affect speech intelligibility; be sure to audition them with a dialog-based soundtrack before choosing them for a home theater application.

The on-axis data corroborated the tonal idiosyncrasies noted in my listening sessions. The broad rise in the 80-300 Hz range is probably responsible for the spotlighting of male vocalists, while the even-higher hump in the 3-6kHz region is probably responsible for the enhanced "snap" of snare drums and the brightening of female voices, violins and other high-pitched instruments. The "two-hump" shape of the Vörtex's response fills a wide, ±5dB window from 70Hz to 7kHz; above 7kHz the on-axis response rolls off at the rate of about 6dB per octave.

The same basic character is seen in the off-axis curve, which only serves to underscore the "rock speaker" tonal presentation. The 30- and 60-degree up and down measurements [not shown] show particularly strong departures from tonal neutrality, strongly suggesting that the speakers be installed such that the tweeters are at ear-level. The claimed 92dB sensitivity would appear to be overly optimistic; 89dB would seem to represent a realistic broadband average.

MB Quart 90M
The German-made Quart 90M relies on an 8-inch woofer and 1-inch titanium dome tweeter. Quoted frequency response is 45Hz to 32kHz; sensitivity is specified at 87dB/1 watt/1 meter.

Patrick found the 90M easy to install and get a good fit and seal.

My listening notes for the 90M identified a slightly overdone "fullness" in the mid- and upper-bass, but with scant mark-downs elsewhere. On the pop and rock test tracks, the effect generally worked to drive the music forward: Rock, after all, is largely a rhythmic music form, underpinned by instruments (e.g., bass drum and bass guitar) with strong mid- and upper-bass output. The "good chemistry" between the 90M and popular music extended to a few older music forms, too, giving just enough boost to the baritone and bass vocalists on Cant de la Sibilla to let them fill up the big church acoustic convincingly, and underscoring the body and woodiness of Christopher Parkening's classical guitar.

However, when handling other classical selections—Prokofiev, Bartoli, Pogorelich—the effect was less successful, sometimes injecting a heaviness that often thickened and occasionally blurred intricate passagework. A similar thickness was noted on Lyle Lovett's voice on "North Dakota," where the "fortified" vocal seemed too heavy to play against the more delicate, airy percussion effects. This bass quibble aside—and it is a quibble more than a complaint—the 90M proved to be one of the more consistently listenable, natural-sounding speakers in the survey.

It turned out that the microphone and TEF system "heard" the speaker largely the same way. A +3dB hump centered at 150Hz was the only impediment to fitting the on-axis response within a ±3dB window from 60Hz to 13kHz. With one-third octave smoothing added, the 90M's response fits within a commendable ±2dB window from 250Hz to 20kHz. The titanium dome tweeter shows only a modest bump around 18kHz, refuting the view that "big spikes" are part and parcel of the breed. The off-axis picture is generally smooth, though speech intelligibility may suffer a bit more than with other in-walls for listeners postioned 60 degrees or more away off the speaker's frontal axis. Lastly, a look at the graph might suggest that the manufacturer was overly conservative in specifying a sensitivity rating of 87dB for a 1-watt input. Because the 90M is a 4-ohm, not 8-ohm, device, it yielded 3 decibels more output when fed the 2.83 volts used throughout this test. A bit of math reveals that, given a 2.0 volt signal (i.e., 1 watt into 4 ohms), the 90M would meet its maker's 87dB spec right on the nose.

McIntosh WS220
Like the InWall Audio Vörtex and Triad Gold systems in this survey, the McIntosh WS220 uses its own MDF enclosure (or "backbox") to create an optimal acoustical environment for its drivers-in this case, a shielded 8-inch woofer and a 1-inch aluminum dome tweeter that is set into a portion of the front baffle angled inward 20 degrees. Quoted frequency response is 60Hz to 22kHz; sensitivity is given at 87dB with a 2.83 volt input (i.e., 1 watt into 8 ohms), measured at 1 meter.

The solid look and feel of the cabinet and the clear instructions inspired confidence. The frame and grille proved less impressive, bending slightly under tension.

The WS220 proved to be one of the least "in-wall"-sounding speakers in the survey, with a refreshing freedom from most of the breed's sonic trademarks: stridency in the treble, honk in the midrange, and smearing in the upper bass. Whether due to the nonresonant enclosure (which clearly reduces the "grunge" otherwise produced by vibrating gypsum board) or the selection of drivers and parts, the McIntosh sounded cleaner, better focused and more articulate than the vast majority of its competitors. This was most obvious on the difficult Prokofiev track, but also on the introduction to the Lyle Lovett number, with its airy sound effects, and on the rhythm section on Sting's "Mad About You."

There were just two noticeable, though admittedly mild, departures from tonal "rightness:" A tendency to thin out some of the warmth and "body" of voices and instruments (heard on the Pogorelich piano, and Mary Black, Counting Crows and Chris Rea vocal tracks); and a tendency to accentuate that portion of the midrange where the "a" (as in "cat") sound lies (noticed on the Cant de la Sibilla, Mary Black, and Chris Rea tracks). I made a note during the listening sessions to check into whether there might be a shallow depression in the frequency response in the "warmth" region-from about 200 to 300Hz-and a bump at around 600Hz that would explain that sometimes-pesky "a".

A subsequent graphing of the response curves bore out my hunches. A bump at 620Hz turned up in the on-axis response, neatly accounting for the "cat" effect, and so did the expected dip beginning at about 200Hz. Note that the "on-axis" response (red trace) in the accompanying graph was derived as if the WS220's tweeter were firing straight ahead. Because the tweeter is angled inward, toward a listener positioned between the speakers, it is actually aimed some 20 degrees away from the microphone's "on-axis" position.

The dark-grey trace on the additional graph shows the unsmoothed response with the microphone moved 30 degrees to the left, where it is more nearly on-axis with the tweeter. Note that this trace is somewhat smoother than the putative "on-axis" response, fitting a ±3.5dB tolerance window from 57Hz to 20kHz. Applying a little smoothing results in a response that, between 130Hz and 8.5kHz, falls within a commendable ±1.5dB window. The three additional traces on this detail demonstrate how the speaker's response changes when the listener moves away from the direct "line of fire"; taken more generally, they also indicate how the response of front-firing speakers typically varies depending on horizontal angle. The WS220's 87dB sensitivity rating was confirmed.

Sonance D6000
The three-way D6000 is Sonance's top-of-the-line offering, featuring an 8-inch polypropylene woofer, 2.5-inch midrange, and 3/4-inch aluminum dome tweeter. The tweeter can be swiveled to direct high-frequencies toward the listening area. Comprehensive tonal adjustability is provided by three banks of switches on the front-baffle that allow the individual driver responses to be tailored to the acoustical environment; a little arithmetic reveals 256 possible combinations. The tests were conducted with all settings at the factory defaults (i.e., "flat") and with the tweeter aimed straight ahead. Frequency response is given as 33Hz to 21kHz ±2dB. Sensitivity is quoted as 91dB/1 watt/1 meter.

Patrick found the D6000 a cinch to install, achieving a good fit and tight seal in a matter of a few minutes.

Despite its nominally flat setting, the D6000's tonal balance appeared to emphasize the mid-bass and the upper reaches of the treble. These departures from tonal accuracy were not as dramatic as with some of the other speakers in the survey, but they proved to be consistently audible nevertheless. The drum on the Cancionero disc was overly prominent, though to the D6000's credit it was rendered cleanly, without tubbiness. A trace of mid-bass heaviness crept into the Sting, Hiatt and Prokofiev tracks, though on these cuts the D6000 countered with a spacious, crisp sound that ultimately flattered the music.

As with most deviations from flat response, an "etched," airy-sounding treble isn't always flattering: On the Pogorelich selection the piano showed good "bite," though a sense of excessive "tinkliness" set in after a few minutes. The D6000 did a first-rate job of portraying the ambience and reverberation of the church acoustic in the Cant de la Sibilla selection, where the speaker's natural mid-bass prominence enhanced the deep, cavernous sense of space. The Sonance also sounded sure-footed when cranked up to levels that drove other speakers into conniption fits.

The measurements buttress the subjective evaluations. The on-axis frequency response shows an unmistakable rise in the 100-200Hz region, and another centered at 12.8kHz. Response between these two "bookends" is less than ideally smooth, with notable valleys at about 425Hz, 900Hz, 2.3kHz and 5.3kHz. The ±2dB window quoted by the manufacturer must be relaxed to about ±7dB to accommodate the actual, unsmoothed response. Even with one-third octave smoothing—which minimizes the severity of the peaks and valleys—the D6000 requires a full ±6dB window to contain its 40Hz to 18kHz response. The Sonance flagship behaves rather better off-axis, where the high-frequency peak is essentially gone, although the presentation will still be dominated by the "plummy" mid-bass. The 91dB sensitivity is clearly not met: Except for the mid-bass and upper-treble peaks, the 1-watt/1-meter response does not exceed 87dB, averaging perhaps 84dB.

None of the foregoing should discourage the seeker of good in-wall sound: The sheer adjustability of the drivers' output levels, plus the aimable tweeter, make the D6000 the most "tailorable" speaker in this survey. This tunability, coupled with a strong woofer and a knack for remaining unflappable at high drive levels, will likely ensure its suitability in many installations that will humble the D6000's less flexible competitors.

Triad InWall Gold Satellite/PowerSub
The InWall Gold Satellite and InWall Gold/4 PowerSub were tested as a system to see what enhancement an in-wall subwoofer might bring to an in-wall satellite. The IWG PowerSub is a complete package—enclosure, 10-inch woofer, and crossover/power amplifier—with a 20Hz to 180Hz rated operating range (the upper limit is set by the user). The IWG Satellite houses a 6.5-inch woofer and a forward-firing 1-inch tweeter in an internally braced MDF enclosure somewhat similar in appearance to the McIntosh WS220. (Triad reportedly builds the McIntosh unit.) Frequency response is given as 72Hz to 20kHz, ±2dB; sensitivity is quoted as 87dB with a 1 watt input.

The Triad satellite and subwoofer installed like the McIntosh, q.v.

The Triad system proved to be a winner from the outset. The drum strokes on the Cancionero disc were tight and well-defined, and other instruments sounded more focused and realistic than I'd heard from any in-wall design. The Mary Black cut, which had elicited a smorgasbord of unsavory sounds from other speakers—screeches, smears and honks—came across as if all those strumming guitars, close harmonies and swirling accordions posed no challenge at all.

The ability to keep densely-scored passages from blurring was especially striking on the Prokofiev orchestral selection, achieving sharp string definition without adding upper-midrange glare. At the other end of the musical scale, John Hiatt's rocker, "Riding with the King," showed more drive and snap than any speaker in the survey, representing a viable option to freestanding speakers in the price range. As expected, the subwoofer increased the "physicality" and "realness" of the presentation, yet sometimes the enhancement worked a surprise or two. When Sting sang the line "I'm lost without you" he could have been referring to the bass drum: without it, the tune doesn't work—the parts don't fit—and it was only with the Triad system that the bass drum got properly incorporated and the tune "clicked."

Although I had to go outside the dozen core test tracks to do it, I managed to dig up two selections that caused the Triad system to falter: Sir Edward Elgar's "Where Corals Lie" for mezzo soprano and orchestra (EMI CDC7-47329-2); and Rachmaninov's "Symphonic Dances" (London 410-124-2). In both cases the cellos didn't come across as full and rich as they should have.

Reviewing the on-axis frequency response reveals a narrow "V" centered at about 220Hz, probably the source of the aforementioned thinness of cello tone. The shape and location of the notch arouse my suspicions that it's the work of a standing wave inside the satellite cabinet. (If so, a fix can probably be easily effected by placing a wedge of dense absorbent foam or fiberglass horizontally inside the cabinet about half-way between the top and bottom.) Putting aside this dip, and another, narrower one at about 3kHz, the IWG system's on-axis response manages to stay within a ±3dB window from 40Hz to beyond 18kHz. The off-axis response is strikingly flat and smooth throughout the critical midrange and lower treble. After the familiar dip where the crossover phases out the woofer and brings in the tweeter—at around 2.2kHz by the look of the graph—the off-axis average begins to roll off fairly smoothly at about 6dB per octave. The 87dB rated sensitivity was confirmed.

Sidebar: Test Setup

Each speaker was mounted in a four-foot high wall formed of two-by-four wood studs set 16 inches on center, and sheathed with a single layer of half-inch gypsum board on the back side and two layers on the front. The two front layers were separated by a paper-thin, sticky constrained-layer damping sheet (Omni-brand dB-Rock) to control drum-like wall resonances. The 2,436 cubic inch cavity-a volume typical in residential wall construction with requisite fire-blocking-was filled with conventional R-11 fiberglass insulation.

Each speaker was installed by Patrick Calderone, a professional A/V installer presently based in Southern California, per manufacturer instructions. All speakers were checked for good "seating" in the wall, as well as for buzzes, rattles and air leaks. Both gypsum board layers and contrained-layer damping sheet were changed whenever the cut-out requirements for a speaker did not closely match the hole left by the preceding speaker tested.

To get the most accurate ground-plane measurements possible, each speaker/wall system under test was set flush into the ground, speaker aiming skyward, in a quiet, rural setting-my backyard. The joints between the ground and wall edges were filled and smoothed to essentially eliminate secondary radiation. If the speaker had a level control it was set to the "flat" or "0" position. If, as in several cases, the speaker's tweeter was pivotable, it was aimed straight ahead and left there.

An ACO Pacific instrumention microphone and a Techron TEF 20HI connected to a PC were used to generate, capture and display test signals. The test signals were amplified by a Soundstream DA-2 power amplifier set to deliver 1 watt at 8 ohms (2.83 volts) and then routed to the speaker under test via 12-gauge low-oxygen speaker cable. Each speaker was tested with the microphone at 2 meters away directly on-axis, as well as at 30 and 60 degrees off-axis laterally, and at 30 and 60 degrees up and 30 and 60 degrees down. Lateral responses for those speakers with non-vertically aligned drivers were taken at 30 and 60 degrees both to the left and to the right. Six decibels were added to the curves to show the plots as if they were derived at 1 watt/1 meter.

I exported the on-axis and off-axis response files to WaveMetrics' Igor Pro scientific analysis and graphing program, where I averaged all eight off-axis curves for each speaker at each of 4,096 data point to provide a composite picture of how the speaker distributed its energy into the environment. (This is a critical element of the sound of in-wall speakers in particular, due to the fact that off-axis energy predominates the sonic presentation in most casual residential environments.)

Prior to taking frequency response measurements, an energy-time curve (ETC) test was run for each speaker to verify the accuracy of the test setup. Reflections off structures, etc. received within the 58 millisecond measuring time window were never higher in level than -40dB, and typically registered about -60dB out to 120 milliseconds, enabling an extraordinary level of performance detail to be captured, stored and analyzed.

I'd like to thank the following individuals for their help in establishing what amounts to the most rigorous regimen ever used for comparative testing of in-wall speakers: Dr. Floyd Toole, director of research at Harman International; Don Keele, technical editor of Audio magazine; and Farrel Becker of Techron, Inc.

SIDEBAR: Software Used

The following compact discs were used in the core listening portion of the tests:

Francisco de la Torre: "Danza Alta" from El Cancionero de Palacio: 1474-1516, performed by Hesperion XX, Jordi Savall, director. Astrée.

Alessandro Scarlatti: "Già il sole dal Gange" from If You Love Me: 18th Century Italian Songs, performed by Cecilia Bartoli, mezzo-soprano, and György Fischer, piano. London.

Domenico Scarlatti: "Sonata K.20 in E major," Ivo Pogorelich, piano. Deutsche Grammophon.

Mary Black: "Still Believing" from Babes in the Woods. Gifthorse/Curb.

John Hiatt: "Riding with the King" from Riding with the King. Geffen.

Sting: "Mad About You" from The Soul Cages. A&M.

Sergei Prokofiev: Introduction to "Romeo and Juliet, op. 64," Kirov Orchestra, Leningrad, Valery Gergiev, conductor. Philips.

Counting Crows: "Mr. Jones" from August and Everything After. Geffen.

Chris Rea: "Auberge" from Auberge. Atco.

Federico Torroba: "Castles of Spain" from A Tribute to Segovia, Christopher Parkening, guitar. EMI Classics.

Performance-Based Room Design (sm)

Building your own dedicated A/V room?
Wondering how to get your room & A/V gear to work with each other?
Questions about Home Theater or Listening Room acoustic design?
E-mail us!

Copyright (c) 1998-2001 Keith Yates Design Group, Inc. All rights reserved.