Sound-Lab A-1 electrostatic loudspeaker Page 2
The A-1's radiating surface area is 2200 square inchesa full 30% greater than the Model A-3's. When sound is radiated from such a large, tall surface in close proximity to both floor and ceiling, the vertical dispersion of the sound follows that of a line source. That is, a plane wave is launched from the surface for those frequencies whose wavelength is small compared to the length of the line source. Hence, all large panel radiators are generically referred to as planar speakers. In the nearfield of a line source there is no vertical dispersion of energy. Energy radiates in a figure-eight or dipole fashion from each side of the diaphragm, but each radiation lobe looks like a pie wedge with flat top and bottom. Sound radiation fans out horizontally but does not disperse vertically. A resultant practical benefit is that the vertical listening axis or listening height is not as critical for a planar speaker as it is for a multi-way dynamic speaker. One is not tied down to a critical listening axis on which the driver blending is optimum.
Another consequence of this type of dispersion pattern is that sound intensity for a planar speaker only falls off by a factor of two for each doubling of distance from the source; for a point source, intensity falls off much more rapidlyas the square of the distance from the source.
This has important perceptual consequences. In the real world of musical instruments there are no point sources of sound. The acoustic outputs of a piano, double bass, cello, or any other stringed instrument essentially come from the sounding boards themselves. Wind instruments come closest to fitting the definition of a point source, but even here things aren't as they seem.
If there are no point sources of sound, what happens when you try to reproduce musical instruments with a point-source loudspeaker? Take piano reproduction: For front-to-back head movements, intensity falls off more rapidly than in the concert hall. Sitting close to a piano (in the piano's nearfield) during a recital would correlate with much-smaller-intensity variations as I move my head about. A planar speaker more closely reproduces the natural variation of intensity with distance that one experiences live.
If I attempt to move my head 12" laterally in and out of my favorite minimonitors' sweet spot, the soundstage pretty much collapses to the channel my head is closest to. Well, that's exactly what the precedence effect predicts: a perceived sound will tend to appear to come from the nearest source. What about the analogous live-sound situation of moving from the extreme left of the concert hall to the extreme right? Am I still able to hear the entire orchestra, or am I confined to monitoring only a slice of the stage? The answer is that, while the stage is skewed spatially toward me, I can still perceive the entire stage. Instruments located on the opposite side of the stage are not masked by nearer instruments. Planar speakers such as the Sound-Lab A-1 and A-3 emulate this sort of concert-hall perspective. The precedence effect is not as strong as my head moves laterally because intensity variations with distance are less, by a factor of two, for planar than for point-source loudspeakers.
In a Sound-Lab "White Paper," Dr. West discusses the "reciprocity" principle and its application to Sound-Lab speakers. This principle (in West's version) states that, for a loudspeaker to re-create a soundstage with spatial accuracy, the dispersion angle of the speaker and the acceptance angle of the microphone must be the same. Because most microphones feature a cardioid acceptance pattern, the A-1's dispersion pattern is right most of the time, goes the argument. While the A-1's controlled dispersion pattern does have merit in minimizing floor and ceiling early reflections, two-channel home stereo has its own brand of requirements which have little to do with emulating microphones. However, I firmly believe that planar speakers are more ideally suited for capturing the most natural soundstage illusion in the home. One reason for this has to do with the planar wave launch initiated by such speakers, which closely emulates the wave propagation of musical instruments. The other reason, also a result of the planar speaker's size, has to do with surface loudness.
German acoustician Ernst Petzold was responsible for formalizing the concept of surface loudness in the '20s. The fact that his ideas languished in obscurity for many years in no way diminishes the importance of the concept. He defined surface loudness as the ratio of sound power emitted by a surface to the area of that surface, measured in acoustic watts per unit area. Petzold reasoned that, because the auditory system projects sounds to the point of origin (after all, no one says, "I'm hearing the piano in my ear"), the direct sound not only determines direction, but the surface loudness of the source is responsible, at least in some part, for our perception of its spatial extent.
Petzold argued that our perception of the size and timbre of sounds was fundamentally shaped by their surface loudness. He used the analogy of candlepower to explain our perception of a trumpet's timbre as dazzling. Here is a great amount of acoustic power being radiated from a small surface. He generally ascribed the perception of "dazzling" to a sound source that combines great surface loudness with significant acoustic power. A piano, which possesses a low surface loudness combined with great acoustic output, would in his jargon be described as "conveying," which I take to mean hefty in outline.
There are two other possible combinations of surface loudness and sound energy. Low surface loudness together with small acoustic power is described as "filling" in timbre. Great surface loudness together with small acoustic output (ie, a piccolo) is described as "sharp." Petzold went on to explain that loudspeakers typically falsify the surface loudness of instruments they reproduce. For example, a piano reproduced by an 8" driver would be portrayed at an unnaturally high surface loudnesscertainly much greater than that of the original instrument. From this standpoint, planar speakers more realistically approximate the surface loudness of such instruments as piano, cello, and the human voice.
One may argue that, when it comes to trumpets and piccolos, point-source loudspeakers do a better job of approximating surface loudness. But in my experience, planar speakers have done a far better job of approximating the size and feel of most musical instruments. Living with the A-1s served to constantly reinforce the spatial-fidelity potential of this sort of design.
One of the luxuries of having two listening rooms is that I can stroll out of the Reference Room (where the Sound-Labs have found a permanent home) into Listening Room 2, where I can be dazzled by the finest minimonitor sound. Here either the Ensemble Reference or Black Dahlia loudspeakers are capable of flooding the space between the sidewalls in the front third of the room with a vivid soundstage. When everything else in the chain is right, the depth perspective is simply breathtaking. N-o-o-o-o doubt about itthis is great fun! In audiophile jargon, you'd salivate about the "pinpoint imaging."
So what's wrong with this picture? Yes, images are sharply focused in space, even palpably fleshed out in a 3-D manner, with minimonitors, but instrumental outlines are not realistically proportioned. The word "toy-like" keeps popping into my mind. At best, this is a hi-fiish view of reality. Instruments are arrayed within the soundfield like stars in the night sky: small points of light in a vast field. With some recordings the illusion is better than that because of the presence of lots of hall information, but in general "men" are not "men" when reproduced by a point-source loudspeaker.
It's been said that, since no height information is captured by such purist two-channel recording techniques as a Blumlein crossed pair of figure-8 mikes, image outlines should properly collapse to a point. I reject that notion outright. By approximating the proper surface loudness of many instruments, planar speakers can infuse into the reproduction a spaciousness that significantly heightens the illusion of a natural soundstage. The A-1 was capable of sketching the outlines of a piano or chorus with lifelike dimensions that had me hooked like never before. And I don't mean the sensation of experiencing a 10'-wide violin. That sort of spatial distortion is an artifact of poor recording technique. Instead, I was captivated by the sensation of having my favorite singers standing before me as if in the flesh. The realism of the height perspective made me want to run right up to these phantoms and hug them out of sheer joy.
I'm sure you can appreciate the fact that someone (like me) who cut his audio teeth on minimonitor sound would have a hard time weaning himself of his addiction to "pinpoint" imaging. With the room optimized, though, I found that the A-1s could float a soundstage just as effectively as a pair of minimonitors, the crucial difference being the enhanced believability of image outlines. The problem was that it took me literally months to get there. For most people, minimonitors represent a much easier ticket to acceptable imaging.
The Sound-Lab wings are officially known as "Acoustic Wave Delay Panels." And large panels they are: 79.5" H by 24" W by 0.75" D. Made of high-density particleboard nicely finished in oak or walnut, they attach to the frame of the array via hinges so that they may be rotated to any desired angle relative to the frame. Because the panels rest on the floor, they serve to mass-load the speaker's frame and thus reduce reactive movements in the speaker structure; this helps tighten up the bass response. I ended up removing the little feet on the bottom of the wings so they could sink more deeply into the carpet and couple more tightly to the frame.