John Bau: Interstellar Overdrive Page 4
The only thing that ended up being good for both drive-units was having lots of area around the tweeter but getting rid of as much material on the baffle around the woofer as possible so that the woofer was almost baffle-free. The tweeter needs to see a lot of baffle because it needs lots of wavefront support down to a fairly low frequency. But with the woofer, you're basically playing with its radiation pattern; you want to get the dimensions so that they correspond to wavelengths that are within its beaming range. The strange baffle shape was the only way we could achieve both things. Once we saw what it was going to look like, we then started to wonder if it was going to fly, if it was going to be accepted. We were up against the same thing we had on the TC, where we were stretching the limits of aesthetic acceptability.
We don't sit and design things to look like that; it's just that functionally it had to be that way. So we made a cardboard mockup, and said, "Hmm, this doesn't look too bad. I wouldn't mind having one of those if it sounded good." By the way, we don't do any listening during this period. I just work with pieces of wood and stuff to change the dimensions of the baffle; I don't do any listening at all during the design period. It's not until I feel like I'm done that I listen.
Atkinson: In a sense, this stage of your design process is realizing as something physical what you've already modeled on the computer.
Bau: The modeling process anticipates the shortcomings of the drivers. We don't assume that we've got a bass/midrange unit that extends to DC. We have an idea that we're going to be able to give the woofer somewhere between two and two and a half cubic feet. That's the sort of space it likes for an optimally damped cabinet. I aim for a sealed-box bass alignment with a Q of 0.5, which I highly prefer over a ported box for transient cleanliness. That's what ends up coming out to be the best.
Atkinson: Yet hardly any designers opt for this low a Q. It seems that a Q between 0.7 and 1 has been fashionable for a long time.
Bau: I think that's the best compromise, if one's aim is volumetric efficiency rather than transient purity. But transient purity is what we're about. And there's a law that I became aware of with the help of Dick Heyser, which is that whenever you have sharp edges in one domain, it introduces problems in the other. Let's say you've got a system with a frequency response that extends out really flat but then dies like a rock; that will inevitably introduce disturbances in the time domain, in the timing relationships. So if transient purity is what you want, what you need are smooth roll-off characteristics at either end of the spectrum. This is a natural law you can't get around.
We accepted early on, therefore, that our systems weren't going to have low-end "hit" when compared with ported box designs, but they were certainly going to be clearer and better defined. It's just something that we accepted, because there was no other way of achieving the transient purity. You can take the same system and make it into a ported design and it would be inferior because you're changing the phase terms way up into the midrange when you play around with low-end stuff. We're talking three to four octaves above the frequency at which you're making the alterations, and there are still phase changes.
Atkinson: And yet you may have noted the continuing debate in Stereophile about the amount of bass which should be expected from a loudspeaker. It seems to be a subject which gets people very angry.
Bau: Yeah, people have pretty strong opinions about it. I'm not sure what to think about it. I have to say that Americans have been weaned on hyped bass, the sort of Hollywood kick where there's an 80-100Hz hump. It sounds magnificent, with lots of hit and sock, but when you actually measure it, it doesn't go that deep. I also have a hunch that whatever room someone tends to listen in influences their opinions a whole bunch. I don't know about you, but I live in a pretty solidly constructed house. In fact, half of my music room is below ground, so I don't have a whole lot of energy loss. I don't even have a pair of Angeluses at home yet, I have a pair of TCs, but to me the low-frequency response in that room is much more satisfying than in the house that I lived in before, which was a fairly cheaply constructed apartment. Because in that kind of room, most of your low end leaves the room. Your walls can't contain it.
The other thing that I've heard from reflex systems is that there's a monotonic quality to the bass timbre. I was a bass guitar player—still am—and changes in timbre in the low end are very important to me. Even when I was playing professionally, I found that I preferred sealed systems because I could hear subtle differences and I could create subtle differences better than with ported systems where there was always this indistinct sameness to the sound. Unless that was the sameness that I wanted, it was really something you couldn't correct for with an equalizer. It's always there. So that, coupled with the far better timing relationship response of the sealed system, sold me on the sealed system.
I didn't look at alternatives, although early on I was working with transmission line stuff and maybe someday it'll result in a product. Again, though, we have to work within the bounds of aesthetic acceptability, trying to sell a product that people have to put in the room that they generally want to look the nicest. If you're talking about a transmission line system, then how do you get the longest line with the least reflection problems in as small a space as possible? We have a solution to that, but it hasn't resulted in a product yet. We don't know how to make it—I know how to make one or two, but it's a pain in the ass.
Atkinson: Why did you dedicate the Angelus to Richard Heyser's memory? I assume that he had a strong influence on your thinking.