Audio Artistry Dvorak Loudspeaker System Continued...

Dipole Basics: A dipole can be modeled as two independent point-source transducers separated by a gap measured between the front and rear center of the transducer, including the width and depth of any baffle. The two sources radiate opposite-polarity but equal-amplitude soundwaves, giving rise to a moderately directional "figure-8" dispersion pattern.

Some of the earliest known speakers were dipoles; by the mid-'80s several companies were producing large-panel dipole designs, many of which were, and still are, considered among the best transducers in the world. The midrange performance of the Quad ESL-63, for example, closely matched Siegfried's expectations in terms of transient response, timbral purity, and transparency.

However, using a large panel to implement a dipole radiation pattern is fraught with limitations: curtailed dynamic impact, the difficulty in achieving realistic bass reproduction, and the critical nature of speaker-placement requirements in order to achieve acceptably low levels of higher-frequency colorations. In addition, many such designs present a touch, reactive load to the amplifier. Large panels can also visually dominate a listening room, making their acceptance more challenging for the less audio-inclined members of a household.

Unfortunately, since almost all dipole speakers are also planars, the inherent positive attributes of dipole radiation are often mistakenly credited as being unique to panel speakers, while the negative characteristics that stem from the use of a physically large transducer are often falsely blamed on the dipole concept itself! If I had a buck for every time I've heard that "dipoles" are a bear to set up properly, I could easily buy a second pair of Dvoraks (which, by the way, are relatively easy to set up).

The best panel speakers share a number of positive qualities to greater or lesser degree: a crystal-clear and timbrally pure midrange, a freedom from boxy colorations, excellent transient response, and the ability to convey realistic image size. Dipoles also differ markedly from monopoles in a moderately directional radiation pattern over their effective bandwidth, and a reduction of low-frequency in-room reinforcement.

However, linear excursion capability is not among the intrinsic attributes of panel drive-units. Dipole panels that attempt to reproduce genuine bass require a very large transducer in order to move sufficient air; even then, the results aren't always satisfactory.

This need to move large amounts of air to reproduce the low-frequency foundation of music with convincing volume, dynamics, and definition is the fundamental problem with panel speakers. An unequalized dipole's response rolls off with decreasing frequency as the two opposite-polarity soundwaves increasingly cancel each other at frequencies where their separating path length is short compared to the radiated wavelengths. Progressively larger excursion is required, therefore, to maintain a constant sound-pressure level as the frequency drops. For example, for a dipole driver of any given size to generate a 50Hz tone at the some volume at the listening position as a 500Hz tone requires 1000 times the excursion.

Compare this to the performance of a conventional monopole woofer, which needs only 100 times the excursion to maintain the same volume at 50Hz as at 500Hz, and it's easy to see why dipoles put such serious demands on driver quality at their operating extremes. This limitation has given rise to many hybrid designs using conventional woofers to reproduce bass -- with varying degrees of success. Yet taking this course means sacrificing the genuine dipole advantage of low-frequency directionality.

On the opposite end of the spectrum, panel transducers large enough to provide reasonable bass extension typically create a serious compromise in upper-midrange and treble reproduction. All speakers become more directional at higher frequencies, because the radiating area (including the driver and baffle dimensions) becomes equal to, then progressively larger than, the wavelengths of the sound. When the ratio of driver size to wavelength increases beyond a certain point, multiple lobes form in the radiation pattern, producing an uneven off-axis response (see polar-pattern diagrams in sidebar). Reflections from these irregular soundwaves can blend with the speaker's direct sound, adding coloration. The beamy treble also forces the listener to sit with his or her head in a virtual vise to get decent imaging.

These problems are potentially significant, considering the large radiating areas featured by many traditional dipole designs. Manufacturers like Martin-Logan, Sound-Lab, Magnepan, and Quad have developed clever partitioning schemes to minimize treble beaming and off-axis colorations in their designs. Nevertheless, these dispersion problems still exist to some degree with panel speakers.

Despite these very real limitations, the communicative powers of the best panel dipoles can be so stunning that their flaws are accepted, or at least overlooked. Once you've heard how such designs can portray music's scale and clarity, it's hard to do without it.

Moving-Coil Basics: Moving-coil box speakers dominate both high-end and consumer audio sales; the best examples are viewed as real-world benchmarks of excellence. However, no matter how well full-range moving-coils perform in an anechoic chamber -- or a huge ballroom at a trade show -- they're usually bought for use in moderate- to normal-sized living rooms.

When these speakers reproduce music with wavelengths significantly larger than their radiating surface and baffle dimensions, they radiate equally in all directions and excite many of the room's resonances. Since a wavelength of 50Hz extends approximately 22.5' and 20Hz is equivalent to 56.5', the drivers and cabinets of even the largest box speakers are tiny by comparison. As a result, all music below roughly 250Hz radiates from monopole woofers as omnidirectional, spherical soundwaves which then reflect off all adjacent boundaries in the listening room. As the wavelengths of these spherical soundwaves are generally larger than the distance from the speaker to the nearest boundaries, the reflections combine in-phase with the direct sound, resulting in a broad band of low-frequency reinforcement, the so-called "room gain."

In the worse case, up to 9dB can be added to the natural volume level in the bass, with peak-to-peak differences of around 20dB between the nodes and anti-nodes of specific room resonances! In practice, the actual frequencies, and the degree to which reinforcement of standing-wave excitation takes place, will depend on the speaker's low-frequency response, the specific absorption characteristics and dimensions of the room (including acoustic treatments), and where the speaker is placed with respect to the nearest corner and room boundaries.

The sonic result is often an unnatural fullness or, in severe cases, a turgid or bloated "one-note" character to low-frequency reproduction that obscures low-level detail and adds coloration clear up into the midrange. When this excess bass energy is combined with high levels of broad-band reverberant energy, often generated by cumulative reflections from monopole midrange drivers, overall resolution and dynamic contrast suffers. Omnidirectional bass can also skew the overall tonal balance of a recording because it conflicts with the increasingly directional dispersion of a speaker's midrange and treble.

Such problems may be lessened somewhat by careful low-frequency design and intelligent application of acoustic treatments. Use of room-placement computer programs -- such as Visual Ears, described by JA in this issue's "Industry Update" -- or a lot of educated trial and error, can also be of real help in reducing the amplitude of resonances excited by a conventional speaker. Those audiophiles who have taken the time to optimize speaker placement for a given room have no doubts about the importance this offers for increased sonic realism. However, even under near-ideal circumstances, genuine base extension from a monopole speaker in an average-sized room can pose a significant barrier, curtailing the resolution of natural timbres and low-level decay that are captured on good recordings.

the dvorak

The fundamental limitations faced by both dynamic box speakers and large planar designs underscore why Siegfried Linkwitz has spent the past nine years designing a speaker that attempts to minimize the weaknesses of both while building on the strengths of each.

[1] [Page 2] [3] [4] [5] [6] [7]