Floor Loudspeaker Reviews

I was lucky to hear a demonstration of the Terra. It added an enormous amount of realism to music reproduction—there was simply more "there" there. On the other hand, it sometimes cruelly revealed studio trickery. The Sonny Clarke Memorial Quartet's Voodoo (Black Saint BSR 0109), one of my February 1992 "Records To Die For" selections, was revealed—in a system composed of the Forsell Air Bearing turntable with arm and Flywheel (footnote 2), a Lyra cartridge, Adyton electronics, some prototype speakers, and just one Terra subwoofer—to have been the object of rather more equalization than I had suspected. On my speakers, which are flat to about 55Hz, the bass had shown a pleasingly natural weight; with the Terra, the recording was revealed to have been heavily boosted in the lowest two octaves—an estimated 4-6dB at 40-60Hz. While the music still sounded superb, sonically I must modify my earlier recommendation: on full-range systems, this record sounds less than natural.

Future Audio Physic developments include a new model that is expected to be slotted above the current top-of-the-line Medea. Probably to be called the Midas, this will differ from the Medea in that it will have even lower bass extension, and a supertweeter to cover the top octave and beyond. Like the Medea, the Midas will employ the Manger driver—a unique planar concentric driver.

Gerhard likes to talk shop with dealers and customers. His enthusiasm for loudspeakers and music seems endless, and he doesn't hide his opinions. Regarding loudspeaker chassis, Gerhard sees a trend toward stiffer diaphragms. There's the eternal tradeoff between internal damping and stiffness—the former important for smoothness of frequency response and natural tonal balance, the latter for sensitivity and for aliveness of sound.

Gerhard's very excited about Audax's new HDA (High Definition Aerogel) membranes, and will surely incorporate them in Audio Physic speakers once Audax gets manufacturing consistency right. But he also sees interesting possibilities with paper cones—especially when they're oil-impregnated. The latest version of the Virgo, the company's middle floorstander reviewed this month by Michael Fremer, employs paper cones for everything bar the tweeter. Since Audio Physic does a lot of business with countries of high humidity, non-treated paper cones are a no-no. Gerhard also believes that suspensions and cone surrounds need a lot of work. A stiff surround means a cleaner transient response, but also introduces dynamic nonlinearities—the so-called hysteresis distortion: the driver doesn't behave identically regarding forward and backward motion of the cone.

As to crossover components, Gerhard thinks that the fuss about the choice of the capacitor dielectric, which has its place in the high-impedance world inside an amplifier, is out of place with regard to capacitors in passive crossover networks. For a low-impedance loudspeaker environment, Gerhard thinks inductance, physical smallness, and mechanical stability are much more important. For inductor coils, too, mechanical stability is more important than slight variations in electrical properties. On the whole, however, the best crossover is no crossover. The crossover of Audio Physic's smallest speaker, the Step, for example, consists of just one series capacitor for the tweeter.

It's impossible to build a totally rigid, non-resonating enclosure for a reasonably priced speaker. Once you accept that, you have to decide where and how you want to distribute the unavoidable resonances. All Audio Physic speaker cabinets are tuned to 344Hz. You must also consider the tradeoff between additional stiffening of an enclosure by internal partitions and the time-delay effect of adding mass to the cabinet. This is an element of the designer's art, but it's based in science. Audio Physic sponsored a university paper that determined optimum placing of stiffening braces and optimum physical shape of the brace itself.

Gerhard thinks that DSP (digital signal processing) equalization for loudspeakers is the only really novel development to come along in loudspeaker design in many a decade. The advent of fast and affordable computing power enables speaker designers to do in the digital domain what has been frustratingly difficult, if not impossible, to achieve via electrical and mechanical means: clean impulse response coupled with flat frequency and phase responses. The mechanical means to achieve those goals have often been in conflict with each other, which is one of the reasons why speakers have tended to sound so different, every designer judging the compromise a little differently.

An early fruit of Audio Physic's efforts in this area was the DSP equalizer for the Tempo (footnote 3), the result of a close collaboration with Norwegian amplifier manufacturer Adyton and the UK's University of Essex, where Gerhard is a Visiting Fellow.

Gerhard laments the fact that many hi-fi companies, absurd as it may seem, don't have proper listening rooms (footnote 4). Audio Physic has a purpose-built listening room at the factory (heavily treated with Michael Green RoomTune devices), so they can show employees why it might be important to do something just the way they were told and not in another way that might be a little more convenient or economical.

In Gerhard's opinion, the widespread use of DRA Labs' MLSSA measurement system has been a mixed blessing for the loudspeaker industry. On the one hand, it's encouraged designers to look at the time-domain behavior of their products; on the other hand, the MLSSA algorithm assumes a distortion-free test speaker. Since many designers use this tool exclusively, the distortion aspect of speaker design has been slipping a little overall.

Audio Physic has a lot to contribute to the High End's common cause: the art of reproducing music at home. I don't believe novel solutions to old problems or high aesthetic appeal will fail to woo the customer's heart.—Markus Sauer

---

Footnote 2: The Forsell is a combo worthy of its Class A accolade in Stereophile's "Recommended Components" listing. Sans Flywheel, it's a little too bland in the pace and rhythm department for my taste, however.

Footnote 3: Reviewed by Robert Deutsch in the November '94 Stereophile (Vol.17 No.11, p.169).

Footnote 4: I too have spoken to and visited with more than one designer who revealed to me that they don't really listen to their stuff anymore. "We used to do that when we were still in the learning phase, but now I know what sounds good and can concentrate on the lab," is the usual line. Typically, the companies in question are on the decline.