Testing In-Wall Speakers

by Keith Yates
Excerpted from "Five Easy Pieces"
in Audio-Video Interiors magazine, June 1998

For the objective portion, the speakers were tested in my 40-acre "backyard" where freedom from acoustic reflections makes for a better test environment than even the special echo-free test chambers used by the larger manufacturers. This is especially true in the bass range, where the wavelengths are long enough to pass through the foot or two of acoustic foam lining the walls of typical anechoic chambers, then bounce back toward the microphone, distorting the accuracy of the measurements. (Most anechoic environments are accurate down to about 200Hz or so; my setup yields clean measurements all the way down to 20Hz, generally acknowledged as the low-frequency limit of human hearing.)

The test signals were generated and analyzed by a Techron TEF20SHIP, a laboratory-grade device for analyzing the behavior of sound in the domains of time, energy and frequency (hence, "TEF"). Manufactured under license from the Jet Propulsion Lab at Cal-Tech, the TEF has the unusual ability to listen for the test signal being played by the speaker, almost completely ignoring environmental noises--rustling leaves, conversation, vehicular traffic, even the din of the factory floor.

As with the previous rounds, each speaker was installed in its own 4-foot high test wall formed of 2-by-4 wood studs set 16 inche on center and sheathed with a single layer of half-inch gypsum board on the back side and two layers on the front. The 2,436 cubic inch cavity--a volume typical in residential wall construction with requisite fire-blocking--was filled with standard R-11 thermal fiberglass insulation. All speakers were checked for good "seating" in the wall, as well as for buzzes, rattles and air leakes, and every speaker except for the Vandersteen [see "In-Wall Speakers Part 3"] was custom cut into its enclosure per manufacturer's instructions. (The Vandersteen is an on-wall, not in-wall design.)

Well Grounded
To get the most accurate ground-plane measurements possible, each speaker/wall system was put in a 5-inch deep trough in the ground, with the speaker aiming skyward. The joints between the ground and wall edges were filled and smoothed to eliminate the "spraying" of sound off the edges of the enclosure (something known as "secondary radiation" among speaker engineers). If the speaker included provisions for altering its tonal balance, it was configured for "flat" and left there for both the objective and subjective tests that followed.

For the conventional passive speakers in the test, the TEF test signals were amplified by a Carver TFM6CB 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. The active speakers--B&W and Linn--were driven by their own manufacturer-supplied electronics. The output of each speaker was captured by an ACO Pacific instrumentation-grade microphone placed two meters away, directly on-axis (that is, straight in front of the speaker) as well as at 30-degree and 60-degree off-axis locations, both laterally and vertically. An offset of six decibels was added to the curves to show the plots as if they had been derived through the standard 1 watt/1 meter measurement.

In the Lab
The microphone's output was routed to the TEF which was in turn connected to a Toshiba laptop PC running Techron's DOS-based "Sound Lab" software. The resulting data were exported in ASCII format to an Apple PowerMacintosh running Wavemetrics' Igor Pro scientific analysis and graphing software for number crunching and to generate the graphs illustrating all three installments. I had Igor Pro average the various off-axis measurements to derive the single off-axis trace, displayed in blue. As such, there is some built-in smoothing that applies to the off-axis trace; the red, on-axis component was left unsmoothed to show greater response detail, though when viewing the curves on the computer screen after the subjective portions had been written up, I applied one-third and one-sixth octave smoothing to check the overall "trajectory" of the speaker's frequency response. While not included in most published speaker reviews, the composite off-axis curve show a critical element of the sound of all speakers, and in-walls in particular, as it is the off-axis energy that predominates the sonic presentation in most casual residential environments.

Prior to taking frequency measurements, I ran an energy-time curve (ETC) for each speaker to verify the accuracy of the test setup. Reflections off structures, people, and so on 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.

In the Living Room
After saving my TEF measurements to computer disc, but before analyzing the data, I brought the speakers, still in their individual 4-foot walls, into my living room, where I placed them against the wall between a pair of Genelec 1037A professional studio-type monitors. To minimize secondary radiation I wedged a slat of absorptive, 3-pound density fiberglass between the adjoining speaker enclosures and then connected all speakers to a playback system that I considered to be representative of the quality that consumers would use to drive speakers of this performance category. CDs were played on either a Pioneer FD-F1004 100-disc changer or a Linn Karik single-disc unit, and routed to a Linn Kairn preamplifier driving either the Carver power amp or, in the case of the B&W AWM70 and Linn Sekrit, the manufacturer's own amplifiers. Using pink noise, a sound-pressure level meter and the Linn preamp's digital volume control readout, I matched playback levels to within 0.5 decibels.

In order to build a "bridge" among the three installments in this series, I had B&W send me a pair of the Signature 7 in-walls that scored so highly in the March 1996 test. I built an extra wall for the Signature 7, ran it through the same battery of measurement and listening tests as the others in other two installments, and referred back to my original listening notes. Doing so enabled me to "benchmark" the five speakers tested in Part 3 to the 20 tested in the previous installments.

With hardwood floor, area rugs and lath and plaster construction, my room is slightly more reverberant than the average American living room of similar volume. I listened from a variety of locations on the sofa and, indeed, by sitting or standing in various locations around the living room and the adjacent dining room.

Thanks
I'd like to thank my helper for the first two installments, Patrick Calderone, a veteran installer located in Southern California [website at www.speakerguy.com]. My helper for the third installment was Brandon Danieli, a student at the University of San Diego, and all-around fine citizen. Special thanks to three individuals who helped with the methodology employed throughout the survey: Dr. Floyd Toole, corporate vice-president of engineering for Harman International; Don Keele, technical editor of Audio magazine; and Farrel Becker of the Techron division of Crown.

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.