John Tchilinguirian: Energetic Speakers Page 2
Deutsch: Give me an example of something that happened, perhaps during the development of the Veritas, where you found out something in the listening test and went back to analyze it technically.
Tchilinguirian: One was the diffraction effects from cabinets, and how they degraded the dispersion characteristics. A long time ago—way back with the Energy 22 Connoisseur—one of the things that made that speaker special was its spaciousness, transparency, and soundstaging. We used that speaker as a stepping stone to learn a lot of things. It was quite a speaker for its time; it still is for an 8" two-way.
We learned a lot from that speaker. One of the main things was diffraction of the cabinet. Any diffraction not only caused ripples in the on-axis response, but also degraded the off-axis performance. We looked at a lot of different enclosures [for the Veritas], and we found that the shape we have now, with the chamfered edges, gave us very good horizontal dispersion from top to bottom.
We looked at a number of different tweeters, again from the dispersion point of view, and wanted one that could handle a lot of power and had low distortion, good transient performance, and wide dispersion. We elected to go with a 1" dome, but developed it in a way that would give us the same dispersion as a ¾" dome. Just little things like the shape of the dome and the suspension, the loading, the front baffle.
Deutsch: All of this was tested in prototypes?
Tchilinguirian: A number of prototypes that we would test and listen to, test and listen to. Of course, we wanted a flat frequency response, but not all speakers with a flat frequency response sound good. You also have to look at the off-axis performance. What do the off-axis curves look like in comparison with the on-axis curves? They should look like near replicas, maybe down in amplitude, but smooth, without a lot of dips and peaks. Once we were happy with the tweeter...
Deutsch: You started with the tweeter?
Tchilinguirian: Yes. Energy has a lot of experience with tweeter design, going back to the 22's Dual Hyperdome. We like to cross the tweeter over fairly low; one of the reasons is the dispersion pattern. But in order to have a tweeter that crosses over very low, it must have very low distortion and a low fundamental resonance. So we've done a lot of work on how to achieve that: a cloth suspension, a bored-out core piece, damping in the core piece. The Veritas tweeter has a fundamental resonance at about 850Hz—quite low for a 1" tweeter.
Deutsch: What about the top-end resonance?
Tchilinguirian: The top end is 23kHz, but it's notched out in the crossover. And, because of the suspension system, its amplitude is not as high as it would be with other metal-dome tweeters—it damps out the oil-can resonance of the metal dome. And then we put in an electrical network to bring it down further and look at it in the time domain to make sure that it wasn't ringing.
Same with the midrange—actually, all the drive-units. One of the preliminary concepts was that the drivers alone needed to cover a wide range—well beyond what they would cover through the crossover network. The reason for that is that we wanted a nice phase relation between all drivers. So even though the drivers roll off quite steeply, we didn't want them to have trouble a couple of octaves or more beyond their range. So the [raw] woofers go up to 3kHz. The [raw] midrange covers a range from 170Hz to 9kHz. The [raw] tweeter goes down to 800Hz, and after we trap out the 23kHz resonance, it's flat up to 30kHz. We did a lot of trial and error on what the crossover frequencies should be. We'd listen to it, find it's not right, measure it, maybe this is why, and then choose another one—until we got what we wanted in terms of the addition of all the drivers when hooked up to the crossover.
Deutsch: Did you ever face a situation where the measured result pointed in one direction but the listening pointed in another?
Deutsch: So you chose what, according to measurements, was an inferior performance—but you preferred it?
Tchilinguirian: Yes, because it's a visual thing. When you're looking at curves, your whole thought process is visually based, and what you use as a reference may not be the right one. Once we get into the listening test, we look at the comments and try to decipher what people are saying...
Deutsch: ...to see if there's a common thread.
Tchilinguirian: Absolutely. And then go back to try and find out why that is. And then decide maybe it's in this area, let's lower the crossover point, maybe it's the trough in the 75 degrees response, pick up the dispersion in that area. Oh yes, that sounds better. It's very time-consuming, but it's something that needs to be done when you try to get up to the level where we want to be.