dbx Soundfield 10 Speaker System Review

More than a year ago, dbx, which was known until then only as a manufacturer of signal-processing and noise-reduction devices, entered the loudspeaker field with a revolutionary new product, the Soundfield One. The results of extensive psychoacoustic studies were incorporated into the design of the dbx Soundfield One with the goal of generating a sound field that would present a stable and realistic stereo image to listeners anywhere in the room.

The Soundfield One speakers are constructed as “phased arrays” of multiple drivers driven by signals having carefully controlled amplitude and phase characteristics. Each speaker’s horizontal polar radiation pattern is largely omnidirectional but favors the direction of the opposite speaker. The frequency response of the Soundfield One, unlike that of most speakers, is virtually the same in all directions. The Soundfield One meets its design goals in that it combines a stable stereo image throughout the room with superb overall sound quality.

At $2,600 a pair, however, the dbx Soundfield One is expensive, partly because of the fourteen drivers in each cabinet and an accompanying complex and versatile electronic equalizer.

The Soundfield Ten was designed to have sound qualities similar to those of the Soundfield One but to sell for a much lower price—a little less than half. Among the obvious changes in the Model Ten are the reduced number of drivers, eight instead of fourteen, and a redesigned and simplified equalizer. The cabinet of the Soundfield Ten is nearly as large as that of the Soundfield One, but instead of each of the four sides having a pair of woofer/ midrange drivers, the front and rear surfaces are solid and only the left and right sides have the low- and middle-frequency drivers: a 9-inch woofer and a 4-inch cone midrange operating in sealed enclosures. The dbx engineers found that most of the desired effect at low and middle frequencies could be obtained with the side radiation alone. One benefit of this radiation pattern is that the Soundfield Ten can be placed against a wall without sacrificing its unique spatial properties.

The tapered “top hat” of the Soundfield One, which contains an array of six dome tweeters covering a full 360 degrees, has been retained in the Soundfield Ten, but the number of drivers has been reduced to five. As in the Soundfield One, the tweeter that faces the opposite speaker is about 1 inch in diameter, and the other tweeters are approximately 3/4 inch.

The primary function of the driver and crossover configuration of the dbx Soundfield system is the creation of a specific directional characteristic rather than a uniform frequency response. Therefore, it is necessary to equalize the system electrically to smooth the frequency response. The SFC-1 controller supplied with the Soundfield One is a large unit with several features in addition to its equalization functions. Among these features is a sophisticated electronic protection system that monitors the input power levels separately at low and high frequencies and automatically reduces them to safe levels when necessary, making it effectively impossible to blow out a driver.

Although the new SFC-10 controller is only a fraction the size of the SFC-1, it provides similar equalization and the same response-trimming adjustments as the SFC-1, including an ambience control and both low- and high-frequency compensation controls. In addition, a switch marked wall equalizes the system appropriately for placement next to the wall instead of freestanding. The power monitor has been eliminated, however, and the user is advised to protect the speakers by inserting fuses in their circuits. Like most signal-processing accessories, the SFC-10 is normally connected in an amplifier’s tape-monitoring loop, and the function of the amplifier’s tape-monitor switch is performed by a similar switch on the controller.

The instruction manual for the dbx Soundfield Ten is exceptionally complete and informative, in striking contrast to the many speaker “instruction” sheets that tell you little or nothing about the performance and design features of the product. Because of the unique nature of the Soundfield Ten’s design, it is advisable to read the manual carefully. The electrical and acoustic specifications for the speakers and controller/equalizer are included along with details of the test conditions (almost unheard of in loudspeaker specifications).

The dbx Soundfield Ten is finished in walnut veneer on its front and rear surfaces and has black cloth grilles on its left and right sides. Its tweeter array is covered by a tapered frame with black cloth on all four sides and a walnut-finish top surface. The heavy-duty input binding posts are underneath the cabinet. Each speaker system measures 34 inches high and approximately 15 inches square, and each one weighs a little more than 50 pounds. The black-finished SFC-10 Soundfield Imaging Controller is 8-1/2 inches wide, 7 inches deep, and 1-3/4 inches high. Price: $1,250 per pair.

Lab Tests

The acoustic performance of the dbx Soundfield Ten is derived from the total radiation of both speakers as they interact and the way their interaction is perceived by listeners throughout the room. We would therefore expect the most meaningful measurements to be those of the overall system room response rather than quasi-anechoic, or simple FFT, measurements of a single speaker or driver. This proved to be the case in respect to agreement

between the measured results and our subjective response to the sound of the speakers, although we found the FFT measurements useful also as a check on the room-response curves as well as an indication of the response smoothness of individual drivers and of the crossover frequencies.

As might be expected from the omnidirectional radiation pattern of the Soundfield Ten, the frequency-response curves from the left and right speakers, measured with the microphone on the axis of the left speaker, were essentially identical above 100 Hz (room resonances affected lower frequencies slightly). We cannot recall having ever before measured a high-frequency room response as flat as that of the Soundfield Ten. From 4,000 to 20,000 Hz the average of the two speaker-output curves varied less than ±0.5 dB. As with all our tests of this system except power handling, we made these response measurements through the SFC-10 controller/ equalizer with all its controls set to their nominal (center) positions.

Our low- and middle-frequency measurements were also made with the microphone close to the driver to remove the effect of the room on the response. The woofer and the midrange cone driver were measured separately and plotted on the same graph coordinates. The crossover between them appeared to be at about 300 Hz, and the 4-inch midrange driver crossed over the tweeter array at about 3,000 Hz.

The combined low-frequency curves could be spliced to the room response with little ambiguity, resulting in a composite response flat within 5 db overall from 56 to 20,000 Hz. As with any of these composite curves, the low-frequency portion can be strongly affected by the speaker’s surroundings (room size and acoustic treatment) and its placement in the room as well as the position of a listener with respect to the speakers. The composite curve should be viewed as a rough indication of what the speaker system might do in a hypothetical room, but in most real-life situations the low-frequency response limit would probably be shifted downward and the overall bass output would be stronger but not as uniform.

That is essentially what we heard from the Soundfield Ten. The effective low-frequency response extended to about 35 Hz, and the overall response was as smooth at our ears as it had been at the microphone diaphragm. The bass-response adjustment of the SFC-10 is similar to that of a parametric equalizer, with a range of about ± 8 db at 35 Hz and little effect at 100 Hz and above. Using the control made it possible to flatten the bass response; with the adjustment near maximum, the composite frequency response was within ±2.5 dB from 30 to 20,000 Hz. We were interested—and impressed—to find that the typical power-response specification in the speaker manual was exactly the same as our room-response measurement!

The nominal impedance of the Soundfield Ten is 4 ohms, but the specifications give its range as 2.5 to 8 ohms, as our measurements confirmed. The impedance ranged from a minimum of 2.3 ohms at 450 Hz to a maximum of 7.5 ohms at 170 Hz, but over most of the audio range it remained between 3 and 5 ohms. The sensitivity was measured separately for the front and side directions. From the front, the sound-pressure level (SPL) at 1 meter was 86 dB with an input of 2.83 volts (an octave band of pink noise centered at 1,000 Hz). Measured from the side, facing the two cone drivers, the sensitivity was 93 dB. Since in actual use the sound comes from all directions, dbx’s sensitivity rating of 90 dB appears to be realistic (it is also approximately the average of our two measurements).

Our FFT (quasi-anechoic) frequency-response measurements produced curves very like those of the room-response tests. An FFT measurement of both systems operating simultaneously yielded a smooth response curve that resembled our composite room curve, varying about 3 dB overall from 230 to 11.000 Hz. The group delay varied ±0.4 millisecond from 4,000 to 23.000 Hz and reached its maximum, 1.2 milliseconds, at 180 Hz.

We measured the woofer distortion with a constant input to the controller at a level that developed 2.83 volts across the speaker terminals at 1,000 Hz. The equalizer circuits boosted the actual driver level considerably throughout our measurement range from 100 Hz on down, which caused the distortion readings to be higher, especially at the lowest frequencies, than those of conventional (unequalized) speakers; we measured 5 percent at 48 Hz and 12 percent at 40 Hz. The data from the Soundfield Ten should not be compared directly with data from other speakers, however, since the actual drive level ranged from 6 to 25 watts instead of the 1 or 2 watts used with most speakers.

The SFC-10 equalizer had an insertion loss of 10 dB at 1,000 Hz. Its output clipped at 1.3 volts at 1,000 Hz and at 3.5 to 3.7 volts on the low-bass range. Its distortion when delivering 1 volt at 1,000 Hz was 0.022 percent, and its A-weighted noise output was only 20 microvolts, or —94 dB referred to 1 volt.

Our power-handling tests of the Soundfield Ten were made without the SFC-10, and we measured the drive level at the speaker terminals. At 100 Hz the woofer’s acoustic-output waveform developed a slight “ripple” resembling a high-order distortion at about 78 watts into its 5-ohm impedance. The effect was not particularly audible, and there were none of the harsh sounds that result from a voice coil or cone reaching the limits of its movement. At 1,000 Hz we measured the output of the midrange cone driver. At 125 watts (into 5 ohms) we heard a slight hardness in the sound. The 10,000-Hz capability of one of the small dome tweeters exceeded the 1,000 watts we were able to generate from our amplifier (into a 3.2-ohm impedance).

Comments

The dbx Soundfield Ten must be judged by its principal design criterion: does it produce a stable stereo image throughout the listening area? And the answer is yes, definitely. As you walk around the room, the stereo stage remains effectively fixed (and it extends somewhat beyond the speakers themselves). The apparent sound sources do not shift as you approach either speaker until you are directly against one of them. At that point, surprisingly, the speaker nearest you disappears acoustically, and only the other one can be heard!

Under more normal conditions, nothing changes very much as you move around the room, even to points well beyond the limits set by the positions of the speakers. In A/B comparisons with a pair of very fine forward-radiating speakers, there were striking differences in the stereo stage produced by the two systems when we listened from a point on the axis of one of the speakers. In such cases other speakers show a shift to the side from which they are being heard, but the dbx Soundfield Ten proved to be quite immune to such effects.

However, there is much more to enjoying reproduced sound than its imaging, which is actually a minor consideration to me, though a critical one to many other people. What about the overall sound quality of the Soundfield Ten? It is smooth, wide in frequency range, and without significant emphasis or deemphasis of any portion of the spectrum. That is the way it measures and the way it sounds. It is a sound I can listen to for hours on end, often without awareness of any unusual properties—which is another way of saying that it is quite free of many common loudspeaker aberrations.

If the program contains any deep bass, however, you suddenly become aware that these are no ordinary speakers. I found the most pleasing results in my room with the SFC-10’s wall switch engaged, which reduces the output about 3 dB in the 50- to 200-Hz range and boosts it about the same amount from 300 to 600 Hz, and the lf eq knob set to 2:00, a moderate amount to low-bass boost. The deep, skin-tingling bass that these speakers can deliver, especially from organ recordings, even at very moderate levels sets them apart from many others.

Without being able to compare the dbx Soundfield Ten directly with the Soundfield One, which had a similarly pleasing effect on us when we tested it, we can say that the Ten, at about half the price of the One, comes very close to matching its sound and spatial qualities. We especially appreciate a speaker that can be placed either against a wall or free-standing in almost any room and be heard from any part of the room with all its qualities intact. In short, the Soundfield Ten is an excellent value.

Read all

dbx reviews

Read all