Mirage M-1si loudspeaker Of Monopoles, Bipoles, & Dipoles
For the benefit of those who may not have seen GL's review of the similarly conceived M-3si (or our earlier reviews of the original M-1 and M-3), a bipolar design differs from both conventional front-radiating loudspeakers and dipole designs.
The former, of course, describes 90% of the loudspeakers on the market. One or more drivers radiate from the front of an enclosure, which may be configured as a rectangular box (usually), or—if the designer is adventurous and the budget open-ended—as any one of a myriad of other shapes. The midrange and treble sound is sprayed from the front of this cabinet in a generally fan-shaped pattern while the low frequencies are omnidirectional—equally strong in all directions due to the wavelengths at these frequencies being much larger than the loudspeaker dimensions.
In a perfect design the frequency response of the system would be the same at every point on this frontal arc. Since perfection does not exist in the real world, this "ideal" is compromised in as many ways as there are designers. To make matters more complicated, there is, in fact, no real agreement among experts as to what this "ideal" radiation pattern should be. How wide should the dispersion be? Or should it, in fact, be limited at all—physical constraints permitting? The most-often–heard criterion is uniform response within a 60° (±30°) forward arc—but this is hardly universally accepted gospel. As I stated in my review of the original Mirage M-3, the lack of symmetry between the record and playback processes, plus the essentially artificial nature of two-channel stereo (it is really amazing that it works as well as it does), conspire against any one pattern being more "accurate" than another with real-world program material.
Still, the reality of the matter is that the wider the dispersion, the more the room is brought into play. The resulting reflections from the off-axis radiation can affect the perceived balance in any number of very complex ways—the loudspeaker, room, loudspeaker position, and listener position all play a part. A wide dispersion with significant off-axis response variations is probably worse than a narrow, well-behaved dispersion pattern. And the real world intrudes in other ways on any attempt at "ideal" dispersion. As stated earlier, woofers are essentially omnidirectional radiators at low frequencies (below about 300Hz), and all drivers are increasingly directional at high frequencies (most definitely above a frequency at which the diameter of the driver equals the wavelength of the frequency being reproduced). Loudspeaker designers have devised not only a nearly limitless set of variations on the front radiator theme, but have branched off widely into other directions as well. Ignoring for the sake of brevity the various omni- and multi-directional radiators which have found strong adherents in some quarters (the Shahinian Diapason reviewed last month by JGH is perhaps the most visible current high-end loudspeaker striving for a limited form of full-range omnidirectionality—pardon the contradiction in terms), the most common exception to the front-radiating design is the dipole.
Dipole radiation is the natural mode of a panel loudspeaker—though by enclosing the rear of the driver it is possible to make a panel which radiates only from the front (footnote 1). Nearly all dipoles are electrostatic or ribbon (or pseudo-ribbon) loudspeakers, though a whole new subset of conventional-driver dipoles has arisen recently for use in the surround channels of Home Theater systems. The primary radiating characteristic of a dipole entails the equal production of sound from the front and rear, with the energy generated from the back of the loudspeaker 180° out of phase with that from the front. This results in little net output at the sides, especially at low frequencies; the front and rear signals simply cancel on this axis. This unusual pattern produces both benefits and problems for the user: potential room reflections from the sides (and top) are minimized, while reflections from the rear are increased. The popularity of dipole panel loudspeakers in the High End would seem to indicate that many audiophiles find the benefits worth the added complication (footnote 2).
To the best of my knowledge, the term bipole (as contrasted to dipole) was either unknown or obscure prior to Mirage's introduction of the original M-1. A bipolar pattern differs from a dipolar one in that the front and rear outputs are in phase with each other. There is no cancellation at the sides, and the output from a bipole resembles that of a pulsating cylinder. It is not entirely dissimilar to a dipole, however, in that the radiation is still stronger directly to the front and back than to the sides, especially at higher frequencies. And while it is certainly possible to construct a bipole from electrostatic or planar magnetic drivers—perhaps by mounting two of them front-to-back and enclosing and suppressing the rear output of each—no one to my knowledge has yet felt an irresistible urge to do so.—Thomas J. Norton
Footnote 1: If you really want to get picky, even dynamic (cone and dome) drivers are inherently dipoles, though we don't think of them as such because their rears are (nearly) always enclosed.
Footnote 2: There is a lot more to this story, of course. The radiation pattern of dipoles is only one aspect of their appeal.