An Interview with Duke LeJeune of AudioKinesis

I found something interesting. Back in the 60’s and 70’s, there had been a religious war of sorts between two factions: The on-axis response faction, and the power response (or summed omnidirectional response) faction. Tests were conducted, papers published, heretics condemned, and when the smoke had cleared the on-axis people had apparently won… but, not by a large margin.
I figured they were both right, but at the time they lacked the necessary tools to examine what was happening in the time domain. Seemed like the thing to do was get the first-arrival sound right, and also get the reverberant field right, but I didn’t know how to go about designing such a speaker. That was to come later.

In the early 90’s I bought a pair of used Quads, and quit building speakers for a while. Later I stumbled across SoundLab, and reading about the design mesmerized me: This speaker used a faceted-curved diaphragm to get a uniform radiation pattern across a 90 degree arc, both front and back. In other words, both the first-arrival and reverberant sound would be correct! I bought a pair, absolutely loved them, and subsequently became a SoundLab dealer. While there was much more to these big fullrange electrostats than clever radiation pattern control, I believed a large part of their sonic success was due to just that.

The next stage of the journey was a visit to the Classic Audio Reproductions room at T.H.E Show, 2001. This introduced me to what a good horn system could do, and I saw a window of opportunity in matching up a constant-directivity horn with a large woofer, crossing over in the region where the woofer and horn patterns matched. Now this was hardly an original concept – JBL had done exactly that 20 years before in their landmark 4430 studio monitor – but I was unaware of the JBL design, so I thought I’d come up with something special. My search for a horn designer led me to Earl Geddes and his magnificent waveguides, and he and I worked together for a little while on the project that later became his Summa. Along the way I learned a few things about crossover design, both from Earl and from Wayne Parham of PiSpeakers – someone else who had been quietly doing constant-directivity and pattern-matching in the crossover region for years.

For a short time I assembled Summas for Earl, but then I moved from New Orleans to Idaho on short notice and no longer had ready access to a facility where I could continue with that. I decided to open a small high-end store and was all ready to do so when my store burned down before I had insurance (bureaucratic holdup on my business license). Watching my store burn from across the street was when the thought came to me, “now I either have to become a manufacturer, or get out of audio altogether and get a real job.” So I sold what components I had left at home and bought test equipment and software.

LB: I’m terribly sorry for your misfortune, but sometimes there truly is a silver lining within the cloud. But let’s continue: In the most basic terms (we’ll get into the details later), what are the distinguishing features of your speakers?
DL: I pay a lot of attention to getting the reverberant field right, but my designs are hardly unique in that respect. Well, maybe my bipolars have a somewhat unique combination of attributes. My designs are more room-adaptable than most, probably more dynamic than most, give a wider sweet spot than most, and I shoot for compatibility with a wide range of amplifiers. Many of these features go back to the pursuit of natural timbre, which again is hardly an original goal, but I think a worthy one.

LB: As you know, some take the view that the reverberant field should be in the recording, and that anything added by the playback system is a coloration. How would you respond to that?
DL: Excellent question! And I will readily admit that near-field listening does some things extremely well, such as allow the development of enormous soundstage depth. I have customers who use, or have used, my speakers in a nearfield setup. Recordings are engineered with the expectation that they will be heard in a semi-reverberant environment. Microphones are placed closer to the performers than we would normally sit at a concert, and those recordings are mastered on loudspeakers, not on headphones.

We are accustomed to thinking of reflections as causing coloration and degrading clarity, and philosophically we don’t like the room adding to the recording something that was not originally there. But if the reverberant field is done right (which is something we can come back to), timbre is more natural and clarity is actually improved! That’s right, controlled tests have shown that speech intelligibility is improved by normal in-room reflections. Apparently the ear is better able to decipher complex sounds when it gets multiple “looks” in the form of reflections. The direction that reflections arrive from plays a role as well. Reflections that arrive from the same direction as the direct sound are more likely to be perceived as coloration than are reflections that arrive from the sides. And, reflections that arrive from the sides are more effective at imparting a sense of spaciousness and envelopment. One benefit of my recommended 45-degree toe-in is that it ensures a relatively large proportion of the reverberant energy will be arriving from the sides. The ear derives tonal balance from a weighted average of the incoming sounds, so the reverberant energy plays a significant role there. When the spectral balance of the reflections is very close to that of the first-arrival sound, perceived timbre is richer and more vivid. This is why we listen to grand pianos and choral groups in lively recital halls rather than in thickly-padded rooms.
In my opinion the goal of high-end audio is to recreate, as closely as is practical, the perception of listening to live music. If we define something that serves this end as “coloration,” then perhaps our priorities are misplaced.

LB: That last statement is very insightful, and one I wish more audiophiles would pondert. Now, please tell us more about the factors that lead you to believe in the importance of the reverberant field?
DL: As I alluded to earlier, my paradigm owes a great deal to Roger West and his big full-range SoundLab electrostats. To my ears they are do something right that cannot be readily explained by the first-arrival sound alone, and I think their fairly wide dipole pattern is a key element. More recently I’ve had the pleasure of reading Floyd Toole’s excellent book, “Sound Reproduction: The Acoustics and Psychoacoustics of Loudspeakers and Rooms,” which expanded my awareness of the beneficial role of reflections not only regarding timbre, but also clarity and intelligibility. This actually falls more under the banner of “confirmation,” but I’m learning useful information from Toole. Now among designers who pay a lot of attention to the reverberant field, there are differences of opinion regarding what the spectral balance of that reverberant field should be. In a concert hall, the spectral balance of the reverberant energy is somewhat rolled off in the highs relative to the first-arrival sound, so many designers shoot for a corresponding rolloff of the highs in the in-room reverberant field. I do not, and my choice goes back to the reduced path lengths (and thus reduced time delay) of early reflections in a home listening room, as compared with a concert or recital hall.
Reflections arriving before about 20 milliseconds have far more influence on the tonal balance than reflections that arrive later, with the tonal influence of reflections peaking at about 2 milliseconds. In order to preserve correct timbre, I want to minimize the spectral difference between the first-arrival sound and reflections arriving during that first 20 millisecond.

LB: What are your thoughts on why this has not been more widely accepted?
DL: Primarily lack of awareness. Audiophiles understand that frequency response is of fundamental importance, but under-appreciate that the frequency response they hear includes the reverberant sound within the room.
Prejudice also plays a role. Designs that get the reverberant field right tend do be highly unorthodox, and that creates a hurdle from a marketing standpoint. For example, one of my biggest challenges is overcoming a fairly widespread (but fortunately declining) prejudice against anything that looks like a horn. Poly-directional designs (omnis, bipoles, and so forth) have to overcome being conceptually associated with oldschool direct/reflecting designs, which (rightly or wrongly) are routinely disparaged in audiophile circles.
Again from a marketing standpoint, it is difficult to package off-axis response in terms that come across as sexy and enticing. Exotica is a lot more alluring, be it in cabinetry, driver materials, crossover type, whatever.
Now in all fairness many designers do take the reverberant field into account in their juggling of tradeoffs. A two-way cone-‘n-dome design may well have an on-axis dip at the lower end of the tweeter’s range, intended to somewhat compensate for the off-axis pattern flare in the same region. But the marketing department is not going to mention this because “tradeoff” implies “compromise,” which is not something you want to admit when your competitor makes no such admissions.

LB: Now let’s turn to the details. What design elements in your speakers provide the reverberant field?
DL: The most obvious is the constant-directivity horn or waveguide. The idea is to funnel the output of the compression driver into a constant angle, and to do so while imparting as little disturbance (coloration) as possible. The angular coverage of this device establishes the radiation pattern we want the woofer to have at the crossover frequency, so if we have a 90-degree horn or waveguide, we want to cross over where the woofer’s pattern has narrowed to 90 degrees. There’s a limit on how low the horn or waveguide can maintain pattern coverage, so we don’t want to cross over any lower than that limit. From these two considerations we come up with a maximum woofer size, which will be the diameter that gives us 90 degrees at the minimum practical crossover frequency. With a smaller woofer, we would cross over higher up. Now in my opinion the in-room reverberant field is often weaker and less diffuse than would be ideal. So some of my designs use a second set of rear-facing drivers, in a bipolar configuration, which increases the reverberant-to-direct energy ratio. This places some constraints on setup, as adequate distance from the wall behind the speakers plays an important role. In general I recommend 5 feet or more out from the wall, with 3 feet as a minimum.

LB: I note that you refer to your speakers as using “wave guides.” Do these differ from horns - and if so, how? - or is it a simply an alternate nomenclature?
DL: “Wave guide” makes me sound high-fallutin’! :) Seriously, the prejudice against “horns” apparently doesn’t extend to “waveguides.” A waveguide is a type of horn, in this context anyway, specifically a constant-directivity type that avoids diffraction slots and vanes, and uses gentle contours as much as possible. Some of my designs use a round waveguide, and some use a rectangular “waveguide-style horn”. The latter is not quite what I’d call a true “waveguide,” but it is a very low-coloration device especially for something that looks a lot like a traditional horn. I actually “discovered” it while working on a bass guitar cabinet design, and its performance impressed me so much that my wheels started turning.

LB: Most would agree that audio design is all about trade-offs. What did you give up in going the route you chose?
DL: Briefly, my general approach calls for fairly large woofers that operate well up into the midrange region, and the only suitable woofers are fairly high efficiency models, which means that a large box is called for in order to get adequate bass extension. I’m not necessarily a high-efficiency aficionado; that’s just where my radiation pattern priorities send me. Now with my bipolars, we introduce another set of trade-offs. In exchange for a richer, more spectrally-correct reverberant field, we must put up with a larger enclosure, usually a larger encroachment on our living room space, and our driver cost must be spread out over twice as many drivers compared to an equal-cost conventional speaker.

LB: While we’re on the topic of drivers, without giving out any trade secrets, what type of drivers do you use, and why?
DL: I use prosound or prosound-style drivers, and have sourced woofers from Pioneer’s TAD division, AE Speakers, Eminence, and Ciare of Italy. My compression drivers have come from Beyma and Celestion and one other source that I’m not ready to reveal yet. My underpaid marketing department would have me put in some flowery prose here, but basically I use drivers that start out close enough to my goals that I can coax them into doing what I want via the crossover. Unfortunately, I have not been able to avoid a fair amount of trial-and-error in sorting out which drivers meet that criterion.

LB: Speaker positioning can have a dramatic effect on the overall sound in a room. Do any special rules apply to your speakers?
DL: Obviously my bipolars have a distance-from-the-rear-wall requirement, but my monoples are adaptable to a wide variety of positions within the room. The bass tuning can be adjusted by changing the length of the port, to take into account variation in boundary reinforcement. It is even feasible to tune each speaker independently, depending on its location relative to room boundaries. When set up as recommended, toed-in by 45 degrees, the relatively well-controlled radiation patterns reduce the room-to-room variability of the early reflections. Combined with the pattern control, that 45-degree toe-in results in a very wide sweet spot. Let me explain.
The ear localizes sound sources by two mechanisms: Arrival time and intensity. The well-controlled radiation pattern of my speakers means that the intensity falls off smoothly and fairly rapidly as we move off-axis. So visualize a setup with the speakers toed in by 45 degrees or so, such that their axes criss-cross in front of the normal center listening position. As we move off to one side, the near speaker “wins” arrival time but the far speaker “wins” intensity because we are more on-axis of that speaker but well off-axis of the near speaker. The fairly rapid and smooth falloff of the near speaker as we move off-axis is the key to this. Now it’s not perfect, and some off-center locations will be better than others, but soundstaging for off-center listeners holds up better than it does with either conventional speakers or omnis. Credit to Earl Geddes for teaching me this technique.

LB: Audiophiles are increasingly recognizing the importance of room treatments, and a variety of products - both absorbers and diffusors - are more widely available than in years past. Do your speakers have any unique requirements for room treatments? In particular, would absorption on the wall behind the speakers negate their benefits vis-a-vis the reverberant field?
DL: One of my design goals is for my speakers to sound good in any reasonable room without requiring room treatments to “fix” the sound. Some rooms are brighter than others, and some more bass-heavy than others, so my speakers have some adjustability in the high treble and low bass regions. That being said, room treatments done right are quite likely to be beneficial. As a general principle I prefer diffusion to absorption, but it depends on the specific room – in some rooms, absorption will be more beneficial. And remember where I mentioned that the direction of the reflections matters? This is an important consideration when deciding what we do on a given wall. Absorption is more likely to be beneficial on a front or rear wall, and if possible we’d rather have diffusion on the side walls. Absorption on the wall behind the speakers is usually conducive to a very deep soundstage, and if the room needs some absorption that is probably the best place to put it.

LB: Your line of speakers currently consists of six different models. How do they differ from one another, and for what type of listener is each intended?
DL: Well, on paper it’s seven models, but one of them hasn’t been promoted much at this point.
One of the main distinctions between my models is monopolar vs. bipolar configurations. The bipolars make more demands on available real estate in your living room, and in return give a more rich presentation. One might think of my bipolars as high-output, high-efficiency alternatives to a quality planar speaker, and in fact most of my bipolar owners are former planar owners.
The Jazz Modules (photo right) and Dream Makers are my original controlled-pattern designs, with the Jazz being the most popular probably because of its combination of pricing, size, style, and placement flexibility.
The Planetarium line takes advantage of the Swarm four-piece multi-sub system, which some listeners have said produces the best bass they’ve heard.

We have the monopole Alpha and bipolar Beta main modules, and take advantage of the Swarm’s coverage of the bottom two octaves by incorporating higher efficiency than either the Jazz Modules or Dream Makers.
My newest speakers are the Rhythm Prisms (left) and Cloud Chasers (below left), monopolar and bipolar variations that are my attempt to make my speakers more affordable without taking a step backwards in sound quality. The Rhythm Prisms are probably my most room-adaptable model to date, and the Cloud Chasers are pretty closer to the Dream Makers at less than half the price.
 

The model I haven’t promoted much is the Planetarium Gamma, which is essentially a Rhythm Prism in a compact stand-mount enclosure, intended to be used with a Swarm. Also I have done a few semi-custom speakers from time to time; as long as I can adapt one of my models without a crossover redesign, customized variations are often feasible.
Budget and room considerations are probably the main factors in deciding which speaker would work best for someone. All of them work well with a wide variety of music, though the separate subwoofer system will of course go lower than the one-box systems.

LB: Let’s turn now to bass. What is “SWARM,” and how does it work?
DL: The concept behind the Swarm arises from a brief conversation with Earl Geddes, as I was driving him to the airport after CES in 2006. I had been trying for years to come up with a subwoofer system that would match up well with Magneplanars and Quads, trying various enclosure types in a quest for very good “pitch definition” in the bass region, along with good impact (good planars excel at the former but not the latter,). Anyway, Earl mentioned that scattering multiple subs asymmetrically around the room resulted in a net smoothing of the in-room bass, as each sub would interact with the room differently so that the sum would be smoother than any one along. The lightbulb went off in my head, and I asked him for permission to use the idea. He said yes. By the time we got to the airport, I was already designing the Swarm in my head. Let me digress for a minute into acoustics and psychoacoustics. The ear/brain system tends to smooth out peaks and dips that are fairly close to one another, but if they are more than 1/3 octave or so apart then the peaks and dips are usually audible. Now we get room-induced peaks and dips all up and down the spectrum, but only in the bass region are they typically far enough apart (due to the wavelengths involved) that the ear cannot smooth them out. So in the bass region the problem is not too many peaks and dips – the problem is that they are too few and far between! Another factor is that it takes the ear a fair amount of time to hear bass frequencies. The ear cannot even detect the presence of bass energy from less than one full cycle, and it takes several cycles to detect the pitch. So considering the wavelengths and room dimensions, by the time we can hear bass tones the room’s effect is in full swing. Perceptually, in our home listening rooms there is no such thing as “direct sound” in the bass region; by the time we even begin to hear it, it’s all reverberant sound.
The Swarm consists of four fairly small subs and a single kilowatt shelf-mounted external amplifier. The subs are “voiced” to have a gentle rolloff over most of the bass region that is the approximate inverse of anticipated room gain (the vented version comes closer to this ideal than the low-Q sealed version does). The amplifier has a steep 24 dB per octave lowpass filter so that the subs can be scattered without betraying their locations by leaking lower midrange energy, along with a single band of parametric EQ in case the scattering alone doesn’t do the trick. To the best of my knowledge, none of my customers are using the parametric EQ because the in-room bass is sufficiently smoothed as it is. Not only does the Swarm result in a smoothing of the in-room peaks and dips, but the peaks and dips that remain are more numerous and closer together, so that the ear’s smoothing mechanism can be effective. Now at first glance it might seem that multiple bass sources with multiple arrival times results in loss of impact and/or mud, but that is not the case in practice. As explained above, we cannot hear the bass wavefront before the room has its say. The ear responds primarily to frequency response (rather than to time-domain behavior) in the bass region, so when we smooth the frequency response we are solving the biggest problem. Because the low fundamentals and first few overtones are present in proper proportion, the pitch definition is very good. The argument for a single large equalized sub can of course be made, and there are some exceptional examples on the market, but equalization is a local rather than a global solution. In other words, the room-induced peak-and-dip pattern varies so much from one location to another within a room that fixing the frequency response at one location via equalization will almost inevitably make it worse elsewhere. In contrast, the multisub approach smoothes the bass throughout the room (decreases the spatial variance), ironically making equalization (if needed) even more effective.

LB: System synergy exists at all levels, but probably nowhere is it more important than the amp-speaker interaction. What kind of amplifiers do you feel work best with your speakers? Was this a design goal?
DL: The combination of variable-tuning ports and smooth, fairly high impedances makes my speakers compatible with many high output impedance (low damping factor) tube amps. The lowest-powered SETs are probably not a good match except maybe for the higher efficiency Planetariums, but other than that I don’t think there are many constraints on amplifier type. One of my first customers is a manufacturer of SET amps, and he drives his Jazz Modules with five watts per channel. I’ve also shown with inexpensive solid state amps, and made on-site sales at audio shows based on the combination.
With a high output impedance tube amp, there is a window of opportunity that my designs can exploit. A high output impedance amp will often tend to over-emphasize the bass region, but by tuning the port a lot lower than normal we can move that emphasis downward in frequency. Using the Atma-Sphere S-30 OTL amp, for example, this technique can extend the bass approximately 1/3 octave deeper than what we’d get with a solid-state amp.

LB: Last but not least, let’s discuss your business model. It is no secret that the industry is migrating away from Brick and Mortar stores, though that is arguably still the most effective means for customers to learn about and audition a product. You currently sell your speakers directly. Is it your intention to stay with that system, or would you like to establish a dealer network?
DL: My business model is this: Find a way to do what I love and still pay the bills. Okay, maybe that’s more of a philosophy – credit to Ralph Karsten for articulating it for me. To answer your question, I will probably stay with the direct marketing model, though it is much better suited to products that don’t cost so darn much to ship. I have had several dealers inquire about carrying my speakers, but here are my reservations: I’d have to significantly raise prices to include a reasonable markup, and I’d lose the personal satisfaction of working directly with people who have a high enough opinion of (or at least sufficient faith in) what I’m doing to vote with their wallets.
Now I have thought about offering customers the option of functioning as a “demo site” of some sort in exchange for a small percentage of whatever sales they generate, but this muddies the waters if an enthusiastic customer wants to post on the internet. If he has a financial interest in generating activity and sales in his area, in effect he’s now a “dealer” and he should follow the etiquette that prohibits unsolicited online raves from dealers… else he risks the sudden fury of the internet forums when he’s eventually “outed.” As for the state of the industry, I’m not equipped to offer any significant insights. I think one positive trend is that the audio press and regional/consumer-oriented audio shows are now playing a larger role. While high-end audio contracted during the recession, it did not die out and is making something of a comeback I think.
One thing I would like to see in the future of audio is awareness out there in non-audiophile circles that listening to music well-reproduced has benefits beyond the obvious one of enjoyment. Think of what kind of people are found within your circle of audiophile friends. Are they not they an exceptionally fine collection of individuals? I think the difference is that they listen to a lot of music, and when we listen deeply something speaks deeply to our souls, and we emerge from the experience moved, sometimes drained, and changed for the better.

LB: Thank you Duke for an illuminating discussion....
DL: And thank you, Larry!

Audiokinesis website: www.audiokinesis.com
Audiokinesis discussion forum: http://www.audiocircle.com/index.php?board=135.0