Floor Loudspeaker Reviews

The loudspeakers themselves are only part of the dbx 1A system. Also supplied is an active controller, which makes the system immediately suspect for perfectionists, who prefer to have as little active circuitry in their system as possible. The fact that the controller has very little detrimental effect on the sound is likely to be of less importance than the fact that it does, in fact, introduce a small amount of veiling. A perfectionist might think about using the Soundfield 1-A without its controller, to get rid of its slight veiling tendency, but he can just forget about that. The speaker's "raw" frequency response is terrible; all middles, with no highs or lows.

The controller includes response correction for the speakers, so that with all of its front-panel EQ controls set for "flat," the system has the flattest response it is capable of. This may strike the purist as "cheating," but electrical EQing of loudspeakers has always been and still is an entirely legitimate way to go. Some of the most highly respected perfectionist systems use it: KEF and Meridian, for example.)

What about spaciousness? This is almost impossible to judge, because the controller has a front-panel slider which enables one to vary the amount of stereo spread, from exaggerated to mono (none at all). And how does it do this? By a process called stereo "shuffling."

Originally derived from the 1930s work by Alan Blumlein as a means for increasing the spatiality of coincident-miked recordings (see Michael Gerzon's article in the July 1986 issue of Studio Sound), stereo shuffling involves the manipulation of the stereo difference signal to increase (or decrease, if desired) stereo separation between the channels. The stereo signals are matrixed down to a mono sum (A+B) signal and a mono difference (A-B) signal (footnote 2), and the midrange difference signal is processed independently before being remixed with the sum signal to reconstitute the stereo ones. When the difference signal is boosted so as to have greater amplitude than it originally had, relative to the sum signal, and then used to reconstitute the stereo signals, they will have greater separation than they did originally. Attenuating the difference signal produces a narrower soundstage than originally. At the middle setting of the controller's Ambience adjust, the output signal has the same difference content (breadth) as the original signal.

Also on the controller are knobs for LF Cutoff (20 or 32Hz), LF Compensation (boost or cut, to accommodate different room and speaker-placement characteristics), and Rumble Suppression (which blends the channels below 100Hz to eliminate vertical rumble from bad turntables and bad discs). Unlike most rumble filters, this one does not attenuate low bass. All it does is eliminate rumble and acoustic feedback due to vertical vibration of the disc, which are their major source. Yes, it kills LF separation, but most vinyl discs are mono below 100Hz anyway. (It's their overtone content which gives bass instruments their directional locations.)

Other controls include HF Compensation—an across-the-band "tilt" control pivoted around 200Hz—and Auto Balance, which automatically keeps center information at the same level in both channels. The latter constantly measures shared information in both channels over periods of 30s or so and compares its relative amplitudes, adjusting channel balance as necessary to make it the same in both channels. The circuit works remarkably well with most program material, but it does require that the program have fairly consistent center information to begin with. This rules out recordings made with two spaced omni mikes, recordings of string quartets, many small-group jazz combos, and realistically staged operas and theatrical plays, where voices may originate from off-center for many minutes at a time.

Second, the controller must be located in a system before the preamp's own balance control, otherwise it will merely act to compensate for that control's corrections, resulting in consistently unbalanced sound if the speakers themselves need balance correction. Since all preamps have their balance control following the input selector anyway, the latter requirement is met routinely when the controller is inserted in the preamp's Tape Monitor, and the Auto Balance serves only to equalize channel balance between the various signal sources.

At best, though, the Auto Balance often shows some sign of indecisiveness on most musical signal sources, so if you're really critical of channel balancing, you will probably opt for leaving the Auto Balance switched Out, and use manual balancing instead. But one thing the Auto Balance does work superbly for is film soundtracks, where all dialog is routinely recorded as mono information, to keep it firmly anchored on-screen, and to prevent any leakage into the rear Dolby-encoded surround channels. (With Dolby Surround, front/rear separation is greatest for center front signals.)

The Controller also has a built-in power monitor to protect the system from sustained overloads, such as operation at excessively high levels. This circuit, which need not be used if you're confident you won't be abusing the speakers, requires two pairs of unshielded, lightweight wires between the power amp outputs or loudspeakers and the appropriate inputs on the controller. The controller then monitors the output voltage levels, reducing the line-level gain through the controller if those voltages reach potentially damaging levels. (Note that this does not check for amplifier overload; just for signal levels at the speakers. If the amp is clipping, the monitor will sense this as excessive HF energy and will clamp down on it.) The signal attenuation takes place slowly, so as not to effect transient information in program, so it cannot protect the drivers against sudden catastrophic overloads, such as a loud switching "Pop." (dbx recommends fusing the speaker lines with 6-7A fast-blow fuses if you habitually listen at levels exceeding 103dB.)

Set-Up
The system is a snap to set up initially—much more so, in fact, than most because of the 1-A's very wide and very uniform dispersion. Wiring is completely straightforward except for the (optional) wires from the power-amp outputs and the controller, for actuating the overload limiters. I tried the system both ways—with and without the limiter—and I could not hear an iota of difference in the sound. However, I chose not to use the protecting circuit, because I tend never to listen at levels above 104dB, and I feel more comfortable when there's as little extraneous circuitry in the system as possible. I don't care whether it sounds any better with less extra stuff in there; I believe it sounds better, and that's all that counts when I'm the one doing the listening.

Initial setup is just the first step with these speakers, though. They will work "well" with just about any old set-up, as long as they face one another and are reasonably symmetrical to the listening area. But these are capable of working much better than "well."

No other speaker system that I know of gives you the flexibility of placement or the variety of spatial characteristics you can get from a pair of Soundfield 1-As. Without the constraints of precise toeing-in and critical included angles, you are free to go for a shallow soundstage or tremendous depth, tight center imaging or a vast, spacious soundstage. And the so-called Ambience control allows a wide range of soundstage presentations with any given placement of the speakers. This tremendous flexibility of placement is what makes initial setup so simple if all you want is good performance, but it also means you could easily take weeks of experimenting with speaker placement before you get exactly the imaging and soundstaging characteristics you want. But rest assured that, whatever soundstage presentation you prefer, these speakers can give it to you.

Getting good bass performance from the Soundfields can be a bit of a chore, too, because the system's omnidirectional radiation tends to excite every LF standing wave the room can support, and the only way to minimize this is by careful physical placement in relation to the room boundaries. dbx's suggestion that you aim for the most disparate possible ratios of distances to each adjacent room boundary surface is a sound one, but you should still be prepared to do a lot of experimenting, with (perhaps) your listening location as well as speaker location, in order to get the smoothest possible bass response at the listening seat. Most critically important is the 37–55Hz range, as this range provides the underpinnings of all music. If part of it is weak or exaggerated, you will find some recordings to be too thin or too fat, while others may sound reasonably good. A test oscillator is invaluable, if not essential, for low-frequency tweaking of the 1-As and, indeed, of any speaker system.

---

Footnote 2: An M-S stereo microphone delivers separate sum and difference signals, so the latter can be processed without initial dematrixing. Matrixing (S-M, M-S) then produces a pair of conventional stereo signals.