High-End audio's not just dying - we're killing it.
Jim, let's get straight to the point about this Audio Police business. You've hinted at it repeatedly, in full-color and obviously very expensive ads. It strikes me as rather curious. Some of the very publications that ran them could in fact have been directly targeted by the implied criticisms. After all, these magazines may epitomize some of the trends you regard as having gone awry from what's truly important. Do you equate the entrenched establishment with this Audio Police?


Let's also talk about our shared latter-day fears - that High-End audio as we know and love it might soon have to be listed on the endangered species act. From our previous phone conversations, I sense you have some strong feelings on these matters. If this were the stock exchange and not just two guys shooting the breeze, I'd say that the floor was open for trading now - horseplay, sour grapes, witticisms and all.

You mean I'm "live" now? (Laughs.) Tape's rolling and the orchestra's out of tune? How 'bout we start with colorations in loudspeakers? It seems to me that the present focus on frequency response, on harmonic distortion, on mechanical resonances and so forth overlooks something far more sinister and deleterious. Take dynamic compression in multi-driver inefficient speakers, for example.


As you increase input levels, their drivers drift apart. Their dynamic tracking goes haywire. An increase of, say 250% of peak signal input measured against the median doesn't equate to the same rise in intensity across all drivers and all frequencies. Typically the tweeter won't scale properly. Most people are used to that. They won't notice until they hear our designs. But dynamic compression in the bass is very obvious even to the casual listener. It's why outboard subwoofers enjoy such popularity even with so-called full-range speakers.


Simply put, were you to transfer the idealized frequency response graph -- flat -- onto the equivalent dynamic response curve, the results would be anything but flat. They'd fluctuate wildly depending on frequency and amplitude. They'd show up as rather haphazard but grave deviations from anything even remotely resembling linearity.
When people categorically refer to hornspeakers as "colored", what coloration are they really talking about - in contrast to conventional designs?


I don't want to be hyping the Avantgardes, but one of the advantages of a hornspeaker in general -- which is different from a high-efficiency speaker, they're not the same thing -- is accurate dynamic tracking of the input signal. It should largerly remain the same as you go up and down in level.

A lot of inefficient speakers -- and I've been in that camp for aeons before I crossed over; you could call me a rehabilitated expert of sorts -- these low-sensitivity designs require a specific amount of input voltage to come alive. Fall below that and they're dead. Go above that and they become irritating quickly. But to feel the music, you've got to turn up the volume. So you do. Then you loose the battle with irritation. So you turn the music off. You probably feel relieved, too. You're worn out instead of energized. Fatigued instead of nourished. The whole thing is upside down. In some interview I read by Herb Reichert somewhere, he called this the "you just wanna go out and get some air" syndrome. (Cracks up laughing.)


That's the story, your honor. The real truth of it. In this whole sordid chapter on compression and coloration, the audiophile community's put the cart before the horse. The most insidious distortions -- deleterious to our listening enjoyment and clearly noted by our senses unlike other measurable parameters that often don't mean what they say -- the most harmful colorations center on broad-dispersion reflections (off the rear and sidewalls) and temperature-related dynamic compression.


Think about it. Voice coils heat up. Drivers change impedance and resistance. Dynamic tracking across the frequency spectrum falls apart. Things don't scale in proper proportion. Disparate drivers are cobbled together into one design but exhibit different thermal recovery times, different thermal tolerances. That's a huge coloration. But nobody talks about it - especially designers and devotees of speakers that require 50 watts just to get going.


If a voice coil has 50 watts across it for any extended period of time, its fundamental characteristics change. That's not an 'if' or 'maybe'. That you can take to the bank. Resistance increases, the amplitude of the response goes down. It's simple math. In extreme cases, even the crossover point is affected. Now let me ask you: Would you hold a 60-watt bulb in your hand after it's been on for about an hour? (I must have involuntarily shaken my head because Jim points a finger and wags it.)


I didn't think so. Now do you want to put your hand on that voice coil instead, see how hot it's gotten? Of course not. It's crazy. But nobody talks about this stuff. It's just brushed under the rug entirely.
Your carpet looks pretty flat to me, Jim. From what I can tell in this light at least. Or is that a mole hill in yonder corner?
Don't get me started. (Chuckles.) You don't want to see the skeletons in my closet. I used to work for Magnepan, remember?


In our present case, the prevailing myth-information used to be the honking coloration. In the three years since I've been doing this, people who've come here to Atlanta never have heard that. They forget all about that nonsense. But - the correct scaling of the dynamics really is very important.
It's why a great speaker can be satisfying when played softly. You still get the proper dynamic balance of the music. The relationship of soft and intense passages, of small inflections in delivery, what audiophiles call microdynamics, tiny little bursts of rapid miniature peaks, all that remains intact and unchanged. I remember...


When my partner first got his Avantgardes, he had four massive Goldmund amps driving a huge $90,000 modular dynamic speaker. He's someone very gifted and bright. He needs very little sleep. It's not uncommon for him to come down at 3:00 o'clock in the morning to get ready for the day. He goes into the music room, prepares projects for the day, perhaps hammers out some long-range plans or crunches numbers - all while the music's playing softly in the background.


We got him the UNOs first. This was before we had the speakers revoiced, before we had the new woofers. We put them in front of his reference speakers, plainly the wrong spot for serious results. After about an hour, we realized that this was way more fun. We broke the big speakers down and put the UNOs there.


It wasn't a week before he called me. "Jimmy, we've got a problem. I'm just not getting any work done." (Laughs in the retelling.) "I go downstairs as you know I always do. I put on a record or CD and play it softly. That hasn't changed. But now I get hooked. I'm hooked. Next thing you know, it's 7:00. I have to go to work. I've got nothing accomplished. There's maybe three little scratch marks on one lone sheet of paper. Nothing's happened."


And that's what I tell people. It's true. Avantgardes have a very serious drawback. They're not good for background listening. (Cracks up again.) But to grab you while things are playing softly means you need linear dynamic range. We all know that human hearing drops off a bit in the bass and treble when sounds get too quiet. But still. If you keep the general proportions of loud and quiet passages in proper perspective; if you maintain proper loudness of bass versus midrange versus treble; if the overall gestalt remains recognizable, just scaled down in intensity - then the music should still reach out and grab you. Even at subdued levels.


If you don't get that in your system, you're clearly suffering dynamic compression. In fact, most people do. Now, many high-efficiency speakers manage quite nicely -- and I'm terribly biased here -- but we do try to separate the notion of high-sensitivity speakers in general from horn-loaded designs.
How would you describe the difference as it relates to listening?
I wouldn't. Who cares about listening anymore. Right? It's the specs and measurements that matter. Remember? (Grins mysteriously.)


Trouble is, we measure the wrong things. Frequency response at 1W/1m. 2.83 volts. As though music was 1 watt of steady-state signal. When one speaker outputs 103dB, another 83dB for the same input signal. Don't throw me off the subject! Where were we?
Adrift on one of the 6 moons deep in outer space, rewriting the rules of what to listen for. High-sensitivity versus horns.
Indeed. Well, we do have high-efficiency drivers behind our horns but their spherical shapes act as acoustic transformers. They amplify the signal so that the drivers barely have to move. They always retain very tight linear performance conditions. In a high-sensitivity speaker without horn-loading -- single-driver or not -- you're asking the driver to undergo some pretty substantial excursions.


Is it possible to move the driver outwards a long ways, then get it back and have it stop at zero, instantaneously, without overshoot or settle-down? To the extent that it doesn't stop instantaneously, you have chaos.


A very famous large speaker that I'm intimately familiar with depended on a large diaphragm stretched very tightly. It started exceedingly fast but continued to vibrate a little at the end of each "rebound". It introduced a certain grunge factor. Things continued to go on after the actual music signal had stopped. But it's amazing what you can get used to and start missing when it's taken away.


People come here and I play something with guitar in it. And I try to prepare them. One of the specialties of our speakers -- most noticeable on vocals first, plucked instruments second -- is their ability to precisely start and stop. The natural tendency for people is a very surprised reaction. "How can that guitar ring out like that? I've got that CD, I've never heard it decay like that. Is that the horn thing going on?" You know, as though they'd discovered some new ringing coloration, had somehow spotted what the rumors about horns always talk about.


Well no. What you are hearing is the body of the instrument after the string has been plucked. That whole event is no longer masked like you're used to. It's far fuller and richer now. The driver has shut up reproducing the initial attack. It now let's you hear the next part of that note, the resonance of the instrument's body following the initial attack. It isn't ringing. It's the opposite of ringing! (Slaps his knee for emphasis.)


And you know, speakers that might claim to be low-mass and low ringing may in fact have more ringing. They start readily but won't stop on a dime, especially if the drive units don't sport overbuilt motors behind them. It always tickles me when someone points at a speaker as being very quick because it has low mass.


Take two dragstrip racers on the track. One's got a wimpy 180 hp Turbo Miata engine, the other is loaded down with a nitrous-oxide testosterone monstrosity. Which one will cross the winnig line first? Never mind the weight -- and ultimately, it is a combination of both weight and horsepower -- clearly the more powerful engine will outrace the smaller one. With speakers, low-mass lightness in a cone creates greater potential for deformation, flexing and various breakup modes. Again, nobody talks about that. We get part of it right but don't look at the whole picture. It's like promoting low-mass tone arm design over all the other contributing factors that enter the picture. A tonearm needs correct ratio of inertia from the weight to the pivot. It's the same kind of thinking. But we're again getting off-track. No pun intended (yeah, right - I just knew Smith was gonna get witty on me) ...