Loudspeaker design is often an interesting competition of ideas. Equally intelligent and competent designers may embrace very different philosophies. Even when they embrace the same philosophy, they may choose very different implementations. And even among speakers which are practically identical at first glance we may find very divergent yet creative thinking.
We call the general philosophy we embrace the “Two-Streams Paradigm”. The Two-Streams Paradigm holds that “presence” and “envelopment” are both highly desirable; that the direct sound conveys “presence” and the reverberant sound conveys “envelopment”; and that a reflection-free time gap between the direct and reverberant sounds enables both presence and envelopment. Concert hall acoustics and psychoacoustics expert David Griesinger on the subject:
"Envelopment is the holy grail of concert hall design. When reproducing sound in small spaces [home listening rooms], envelopment is often absent."
"Envelopment is perceived when the ear and brain can detect TWO separate streams: A foreground stream of direct sound, and a background stream of reverberation. Both streams must be present if sound is perceived as enveloping."
“When presence is lacking the earliest reflections are the most responsible.”
"The earlier a reflection arrives the more it contributes to masking the direct sound."
Earl Geddes on the same general topic, focused more specifically on home audio:
“The earlier and the greater in level the first room reflections are, the worse they are. This aspect of sound perception is controversial. Some believe that all reflections are good because they increase the listener's feeling of space – they increase the spaciousness of the sound. While it is certainly true that all reflections add to spaciousness, the very early ones (less than 10 milliseconds behind the direct sound) do so at the sake of imaging and coloration... These reflections must be considered in the loudspeaker design and should be also be considered in the room as well.”
One implication is that later reflections (those arriving after 10 milliseconds) are desirable because they convey spaciousness without degrading imaging or introducing coloration, assuming they are spectrally correct.
The basic concepts of the Two-Streams Paradigm have been around for decades, but have usually been considered a room acoustics matter, rather than a loudspeaker design matter. In effect we are designing into the loudspeaker benefits comparable to fairly extensive and well-executed acoustic treatment. While there are other loudspeaker topologies which are effectively compatible with the Two-Streams Paradigm, our deliberate focus has resulted in a unique combination of attributes which sum to a unique solution.
Here are some of the design concepts we use in the Bohemian 215:
The way we see it, amplifiers + loudspeakers + room = a “system within a system”, so we want a fairly high and smooth impedance curve, along with high efficiency, for compatibility with specialty tube amps like OTL and SET amps. Not to leave the solid state guys out, we have developed a version specifically optimized to perform well with solid-state amplification.
Since we manufacture highly successful disributed multi-sub systems which do a better job in the bottom two octaves than a pair of loudspeakers can, we don't need to build a great deal of bottom-end extension into the Bohemian 215. Nor do we want to – that which multiple subs can better do, let them do. So we are freed up to use midwoofers which are better in the midrange region, and we don't need the enormous enclosures which would be required to combine high efficiency with deep bass. Custom subwoofer options are available for those interested in extension down into the single digits, and of course the Bohemian 215 can be used with other manufactures' subwoofers.
Since most of our design efforts focus on the loudspeaker/room interface, we pay a lot of attention to radiation pattern control. In the Bohemian 215 we use two 15” midwoofers with very powerful motors and very light cones, crossed over to a large horn where their radiation patterns approximately match. Our 15” midwoofers' motor-strength-to-moving-mass ratios surpass those of high-end 5” and 6” midbass drivers, and are competitive with high-end 5” midrangecones. So right off the bat we have impact that small midwoofers cannot dream of, along with excellent articulation and superior radiation pattern control. The excellent articulation is partially a result of the reduced early reflections (which are the most detrimental ones, according to Griesinger and Geddes).
Our midbass enclosure has a wide front baffle which pushes the baffle-step frequency down low enough that it is of little consequence. The two 15” cones in a vertical stack result in a very weak floor-bounce anomaly. The net result is an authoratative midbass which lends realistic body to instruments like cello and double bass. The enclosure uses extensive bracing and both constrained and unconstrained layers of complementary materials for the panels, and is divided into an upper and a lower chamber (of different dimensions) to reduce the length of the longest standing wavelengths the internal dimensions allow, resulting in more effective management via damping material. The only internal parallel walls are the side walls, and those are well treated.
The horn we use is unique. We designed it using Earl Geddes' equations and nobody else has anything quite like it: A large Oblate Spheroid designed for a 1.4” throat compression driver. The Oblate Spheroid profile is mathematically the most benign curvature for a given radiation pattern angle. Ours has a 75 degree constant-directivity radiation pattern which gives good coverage over a wide listening area with minimal early sidewall interaction. We paid attention to the details: The horn's entry angle matches the compression driver's exit angle, and the round-overs use purpose-optimized curvatures large enough to be effective in the frequency ranges which matter most.
Now one of the tradeoffs involved with the use of a constant-directivity type horn has to do with the on-axis efficiency at high frequencies. Because the high frequencies are spread across a full pattern width (rather than being concentrated on-axis, as is the case with exponential, spherical, tractric, hyperbolic, Le Cleac'h, and most other horns), the on-axis SPL at high frequencies is not as high as with most horn types. The same amount of high frequency energy is going out into the room, but its distribution is consistent across the horn's pattern width, instead of being “beamed” into an angle which gets progressively narrower as we go up in frequency. This places a practical upper limit of about 100 dB as far as system efficiency goes, for constant-directivity horns. Horn systems with considerably higher on-axis efficiencies exist, but their radiation patterns are correspondingly narrower at high frequencies.
Most high-efficiency horn systems use at least two upper-frequency horns, one for the midrange and one for the tweeter. This imposes a juggling of tradeoffs: Do we align the horn mouths, or the compression driver diaphragms? And if we align the horn mouths, do we use DSP delay to align their acoustic centers?
Recent developments in compression driver diaphragms have made it possible to cover a wider range with a single driver, and the compression drivers we use are optimized to go high enough without needing a separate tweeter.
A great deal of consideration went into choosing the horn size. The horn needs to go low enough and high enough, have large enough roundovers to minimize the mouth reflection and diffraction, but not be so large that the distance to the midbass drivers is an issue. As part of our optimization, we use a different curvature for the mouth's inside round-over than the one we use for the outside round-over, each being chosen for its role.
There is a worthwhile advantage to using a single driver to cover the frequency range from 700 Hz on up. According to David Griesinger, it is particularly important that the phase of the overtones above 1 kHz (and ideally above 700 Hz) be preserved. The format we use enables this, and goes one step further: The depth of our Oblate Spheroid horn corresponds with the phase-rotation-induced delay imposed on the midwoofers, so the acoustic centers of compression driver and midwoofers are effectively aligned. This avoids the time-domain smear often present in conventional speakers which have highpass and lowpass drivers technically “in phase” in the crossover region, but with the woofer actually delayed by one wavelength relative to the tweeter. We don't have to choose between the elegant simplicity of woofer and tweeter on the same baffle, and the superior time-alignment of stepping the tweeter back by a sufficient distance to effectively align their acoustic centers. The depth of our horn is our tweeter set-back. So the fundamentals and overtones arrive at the same time, or much more nearly so than is normally the case without DSP.