35 Years and Just Getting Started: The J. Gordon Holt Interview Page 8

Stone: Purists might argue that there's always some place that has the right balance so it can be done with just two microphones; you just have to find it.
Holt: Well, they're welcome to spend all the time they want looking for it! [laughs] I mean, there are certain situations where I've walked in, looked at the stage layout, and knew right off that I couldn't do it all with two mikes. That's when the mixer comes out.

Stone: Most of your recordings are done with an M-S microphone configuration.
Holt: Yeah. I've been experimenting with M-S because it showed promise of being free from some of the problems of ORTF, which I otherwise would be using.

Stone: For readers who don't know, what is M-S?
Holt: It's called "Middle-Side" recording. It's also called "Mono-Stereo," which is wrong! What it means is, you have one microphone that's picking up from just a single direction---usually aiming forward, or it might be an omnidirectional mike picking up from all directions. And then the other mike---the Side mike---is a side-facing figure-8, which is dead at the front and the back and sensitive only toward the left and the right. The two sides of that mike are out of phase with each other, and you orient the mike so its left-aiming side is in-phase with the M mike. When you add their signals together, you get a mix of front and left that behaves like a cardioid mike aimed about 45 degrees to left of center. To get the right channel, you first reverse the polarity of the S mike's signal and mix that with the M mike's signal. This time, the resulting polar pattern is directional at 45 degrees to the right of center.

Stone: What's the advantage of M-S over ORTF with a pair of cardioid microphones?
Holt: The main advantage is complete mono compatibility. If you're recording something in stereo and it's going to be broadcast over radio---you know, mono radio---or it's going on a mono soundtrack, you have the output from the M microphone to use as is. All you have to do is mix the two channels together, eliminating the difference signal from the S mike, and you end up with the output that came only from the M mike. So you don't have problems with flanging or interference.

The other advantage is that, since none of the instruments is way off the M mike's axis, there's no tendency for the sound to change in treble response according to direction. With ORTF and coincident cardioid, in order to get adequate separation, you have to angle the mikes so far apart that instruments that are way off-axis from each mike suffer from treble rolloff, and imaging specificity gets badly degraded.

Stone: Some feel that omnidirectional mikes are the only way to really capture full fidelity. How do you feel about that?
Holt: In some respects, they're right. Spaced-omni mikes have the most realistic low-end weight when reproduced through a stereo pair of speakers.

Stone: What are their disadvantages?
Holt: If you get them far enough apart to get real instrumental separation, the imaging stinks. Center images are too broad, and their directionality is vague. And if you use omnis for surround reproduction, which is the only way to reproduce three-dimensional space properly, the amount of bass you get is much too much. You end up being overwhelmed by bass.

Stone: There seem to be two schools in terms of recording orchestras.
Holt: I flunked them both! [laughter]

Stone: I didn't say tests, I said schools! One being that you go in and record what you're hearing, with the goal of just trying to get an accurate portrayal of what's currently there: literal recording. The other is trying to get as close as possible to what you might conceive the ideal sound of an orchestra to be. They seem to be diametrically opposed to each other.
Holt: I prefer the "documentary" or literal sound, because it's there. It exists in nature. Admittedly, no one in an audience seat is going to hear what your microphones hear when they're 15' in the air. But at least you're starting out with a real sound, and if you aim to reproduce that, you can end up with a recording that actually sounds very much like that through a good system. The idea of using the orchestra as simply a starting point, like raw material that is then shaped to fit someone's preconceived notion of beautiful sound, is slippery-slope material because the result depends entirely on the personal preferences of the recording engineer.

Stone: Doesn't that put the audiophile at a decided disadvantage when he's trying to evaluate his system, if most of his software is "bel canto"---engineered to sound beautiful rather than accurate?
Holt: Oh, absolutely. Unfortunately, there isn't much you can do. The hard fact of the matter is that the best performances are generally done by the best orchestras, and they're usually recorded by the companies that have the most money and the least concern for the music! They multimike everything because it gives them maximum "creative control," and then they can't keep their cotton-picking fingers off the knobs.

Stone: So you usually use just two microphones in M-S.
Holt: I'll tell you, if I could afford a really multichannel recording system, that's how I'd be recording everything. Ambience-recovery surround is fine as long as it's all I have, but discrete surround would be much more realistic. And a lot more fun to record, too.

Stone: Do you think surround-sound will offer a chance to do better quality documentary recording?
Holt: Well, sure. I mean, it just goes against all common sense to believe that you can reproduce sounds that normally come at you from all directions through just two speakers in front of you.

Stone: Some people argue that, if the phase is totally accurate, the ear re-creates all of those multidirectional cues.
Holt: The ear can re-create a semblance of them, yes. I have a recording---a test CD done by JVC way early in the CD period---that has a couple of tones on it; sinewaves, I believe. They start off being in synchrony with each other so the signal appears dead center in front. And then the tone repeats over and over, each time with a longer phase difference. If you sit in the middle of a room that is reasonably good for music listening, you'll hear the sound actually march all the way around the room, including behind you, from just two channels. But if you move your head off-center by an inch or so, it's gone. With surround speakers, you can turn and face the back of the room, and hear a real stereo "stage" behind you. If you ever hear real surround reproduction done properly, you'll never mistake the two again.

The thing about a really good multichannel system is that, if you're hearing something coming from behind, you can turn around and look toward it and it stays there. You still hear it coming from where it was. With a stereo system, you turn to look toward it and all of a sudden it's coming from one or the other of the stereo speakers. One of the big deals about discrete surround is that you can actually produce phantom images between the front and back speakers, so that they locate along the sides of the room. That rarely happens with derived surround.