Was all of Jazz From Hell recorded on Synclavier?
No. There's one cut on there, a guitar solo, that was done with a band on the '82 tour. That's "St. Etienne." Everything else is 100 percent Synclavier.
The Synclavier has direct-to-hard-disk recording options that would let you, for example, have somebody come in and actually play saxophone, and it would still be recorded on Synclavier. Was any of that done, or was it all input directly?
No, it was all done with samples and synthesis. It was all typed in or performed in on the keyboard, or performed in using [Roland] Octapads.
Typing it in must be a fairly slow process.
Well, I worked for eight months on this album. So there's quite a bit of work in it.
It sounds like a real breakthrough album, with the vocabulary you used before, but distilled in a very new way.
Wait until you hear the stuff that's coming up. When I first started with the Synclavier, we didn't have a very advanced sampling system. We had mono sampling with not a lot of RAM. Then, at great expense, I picked up the rest of the new sampling gear. We were doing stereo samples here in the studio before Synclavier even had stereo sampling. We figured out a way to do it, and it changed a lot of ways that you could write for the instrument. So the compositions that are on Jazz From Hell already sound old to me, compared to what I'm doing now.
There are some places where we can hear that it's an acoustic guitar sample or a saxophone sample or something quite clearly. But in other places it's not so clear. On "Night School," for example, there's a sustained sound that has a piano attack and something else spliced onto it.
It's actually not spliced; it's simultaneous. It's a stereo sample, a combination of trumpet with pitch-bend and grand piano. The piano notes are not short. They attack, and then as they ring off, you get to hear an unusual noise, which is the acoustic piano playing bends. That's a real easy thing to do on the Synclavier.
So even though you're calling them stereo samples, they weren't always used to create a stereo field.
Well, when I say stereo sample, on the Synclavier you have four partials. You can have a different sound on each partial, which means that when you strike one note, you can have four completely different sounds come out, or you can have two stereo pairs. Or you can have a stereo pair and two other sounds at random. In the case of that particular sound, it is a mono piano and a mono trumpet sample. But the accompanying keyboard sounds are all stereo grand piano.
Have you ever sampled your own guitar and used that?
I've sampled a few notes. I've never plugged into the thing and said, "Now I'm going to sample myself." We extracted them from digital tapes of live performances. A couple of good feedback notes are plopped in. I haven't really gone hog-wild with guitar samples, but Dweezil [Zappa's son] did a whole guitar sampling session last year, and the stereo fuzz-tone samples are just now being trimmed and built into patches, so I'll have a whole assortment of characteristic heavy metal noises.
Do you ever use a guitar synthesizer controller with the Synclavier?
I've tried it, but because of the style I play and the way my hands land on the guitar, it has never felt comfortable to me. I've tried maybe three or four different systems, but none of them drove me crazy.
In general, do you like working with samples, or if you could get good synthetic reproductions that eliminated all of the associated problems, would you prefer that?
I couldn't imagine that any kind of a synthetic reproduction would be able to give you the type of nuance that you get out of a sample. For string pads and things like that, you could fake it pretty good. For bogus globulant brass ensemble stuff, that kind of orchestral cheese, you could get away with FM. And Minimoog bass sounds and things like that usually sound best if they're actually coming out of a Minimoog. What we've done to get those kinds of sounds is sample the Minimoog. You see, with samples, not only are you getting the sound of the instrument, you're getting the ability to capture the instrument in different types of air spaces. For example, we have both dry and ambient room sound percussion noises, and dry and ambient wind. Even with the classical guitar, different types of environments make a big difference.
What percentage of Jazz from Hell was input by typing, and why would you choose one type of input over another?
There are three different ways to type in. One is in a language called Script, which I don't know. I don't use that at all. Bob Rice can type Script. But that's more like writing a computer program, so it has no charm for me. Another way is with their Music Printing program. You can enter or delete notes with the cursor while looking at real music on staves. And if you want to write in tuplets -- if you have 7 over 3 or something like that -- it's real easy to do it that way. You just make a couple of marks and then redraw the screen. You now have edit blocks that correspond to a septuplet over three quarters, or whatever you want. It could be anything. Then you just enter the pitches. For that kind of stuff, that's the easiest way for me to do it. The third way to type is a facility called the G Page. The screen is split into three segments, and you can display three tracks of data at the same time on the screen. In each of those three units, you have three columns of information. The left-hand one tells you the start time of the note. In other words, the beginning of the piece would be beat 1, and all the subsequent beats have numbers. This data reads out either in seconds, beats, or SMPTE numbers; that's all selectable. The center column gives you the name of the pitch an a number which tells you the octave that the pitch lives in. And the right-hand column gives you the duration. All that is editable, so you can move the cursor around, add and delete notes, change start times, which changes the rhythm, and change the pitch and the octave and how long the note lasts. I divide my time between doing stuff on the G Page and doing stuff in the Music Printing.
Have you ever worked with any of the lower-end gear that has some similar functions?
Well, I had a [Yamaha] DX5 and a rack of [Yamaha] TX modules. I also use the [Axxess Systems] Mapper, which is MIDIed to the Synclavier. We've been able to get some truly frightening things out of that. I also have a [Yamaha] CS-80, I've got Electrocomps, Minimoogs, Synkeys. . . . All the heavy-duty hardware that a rock and roll touring band would use, I've purchased and supplied to whoever the keyboard guy is who does the tour. So I know basically what the consumer end of the synthesizer stuff is like, even though I'm not a keyboard player and never expect to be. I am a composer, and as a composer you deal with timbre and other technical matters, and it pays to know what's available so that you can write for it..
Has technology also made it possible to hear more things in your head than you were hearing before?
Well, let's say that a person had never heard a bassoon in his life. And the day that he hears one he's either going to say, "That's the ugliest thing I've ever heard," or "That's God's instrument." Or maybe something in between. But you're going to have a response to an instrument. Every composer has some image in his mind of what he wants his stuff to sound like -- not just the composition, but the overall tonal quality of what he's writing. In my head I have an audio image, not just of the notes, but of the way the notes will sound played in an idealized air space, which is something you can't get in the real world. The closest you can get to it is a digital recording with digital control over imaginary audio ambience. When you can design rooms to your own specifications with a Lexicon, and then place your music in that space, that's getting pretty close to what it's really all about. It's not just the notes on paper that matter, but what they turn into when you start making air molecules move. If it's on paper, it's roughly the equivalent of a recipe for something to eat. The ingredients may sound good on paper, but how do you know whether or not you're going to like it until you eat it? It also resembles the blueprint for a building. A good composition will take into account that you need to have toilets, you need doors going in and out, windows, ventilation. You need all the basic stuff, and then all the rest of it is interior and exterior decorating.
But as you get more involved in electronic instruments, do the things that you're hearing change?
I was going to get to that. Obviously, if you're dealing only with the instruments to which most composers normally have access -- in other words, the known instruments -- you will tend to think in terms of what to do with a known instrument. The moment you get your hands on a piece of equipment like this, where you can modify known instruments in ways that human beings just never do, such as add notes to the top and bottom of the range, or allow a piano to perform pitch-bends or vibrato, even basic things like that will cause you to rethink the existing musical universe. The other thing you get to do is invent sounds from scratch. Of course, that opens up a wide range.
One of the most intriguing things about working with a Synclavier is what it lets you do with rhythm. That's always been one of my favorite things to investigate. It's possible to get accurate performances of the most ridiculous rhythmic combinations. I'll give you an example. I've been working in large tuplets recently. If you're in 3/4, I'd put in a tuplet -- say, a bar of 3/4 that has a 75-tuplet or 35-tuplet in it. You can hear that there's a waltz going on, but when these things occur, it's like, "What is that? Where do these things come from? Why does it still have a groove to it?" It still relates mathematically to something else that's going on in the bar. With this system, you can pick a random number, then take any size bar of music and divide it up into those components. You're going to have an 88-tuplet or an 87-tuplet. Or you can take a composition that has, say, ten bars of 4/4. The first bar you start with an 88-tuplet, the next bar is 87, 86, 85, 84, 83, 82, 81, something like that. You could never hand that to a musician on a piece of paper and say, "Here, do this."
So electronic media have really freed you to get closer to your ideal, to what you're hearing.
It's really made that possible. The next question is whether anybody in the audience wants to hear it. That's the big problem, because the further out I get with these timbral combinations and the unusual rhythms, the further away it gets from any possibility of radio play. And without radio play or some kind of advertising for the album, nobody's even going to know it's there, let alone pick it up. Some people, when they hear it, they absolutely don't like it just on principle because it doesn't have that boom, boom, boom on the floor all the time. I'm delighted that I have the opportunity to go wandering around out in the zones of this thing. I would like it if I had some company out there.
You were talking about creating a specialized ambience with digital reverb, yet Jazz from Hell sounds relatively dry. What kind of processing, if any, did you use on that album?
There's a lot of real subtle processing. It is absolutely not dry. There are tricks to using echo. If you want something to really sound like it's echoing, then that's an obvious effect, like yelling into a cave, that kind of stuff. That tends to make things get soft around the edges. The way ambience is perceived in this album is, each composition has to exist in some sort of imaginary air space, and you don't want the air space to fight against the musical content. You don't just pick an echo program at random, then turn it on and say, "Now we've got air space." What we do is, for each piece, depending on how much transient information is in the piece or what style the piece is, we tailor at least three different rooms. In other words, we have a live echo chamber and two Lexicons. and that gives you the possibility of locating different types of orchestration in different types of imaginary rooms, and then combining those things to make the final stereo picture.
Excerpts from a Keyboard Magazine interview, February 1987"Not quite Jazz, not quite Hell. An electronic Telletubbian Utopia that will make you feel that pink has become blue - except track #7, which really shouldn't be here." Ken Kercheval
Released in 1986. UMRK Digital Remix in 1990 by Bob Stone.