1996-07 Music & Computers

From This Might Be A Wiki
< Archived Interviews & Articles
Giant Step-Time—Computers in the Life of They Might Be Giants
By John Poultney, Music & Computers, July 1996
Archived from: https://www.proquest.com/docview/1535818?sourcetype=Magazines

Anybody who thinks pop songs are all cranked out to the same old formula has never listened to They Might Be Giants. TMBG, as they sometimes call themselves, don't just spice up radio playlists; they're one of the world's most critically acclaimed bands as well. They deftly combine the best elements of pop, jazz, rock, Spaghetti Western, and other styles with lyricals that teeter between absurd and brilliant. Noted musicologist Ira A. Robbins, author of The Trouser Press Guide to Records , call the group "almost beyond belief" because of their inventiveness and song-crafting prowess. TMBG write incredibly catchy songs that make you want to sing along at the top of your lungs to lyrics like, "Everybody wants prosthetic foreheads on their real heads."

The group's core is the songwriting duo of accordionist/saxophonist/clarinetist/keyboardist John Linnell and guitarist John Flansburgh. The two have long used computers in the songwriting, pre-production, and recording processes, and have only recently turned to recording primarily with live musicians instead of sequences. John Linnell took some time out from recording the band's sixth full-length album to chat with Music & Computers about technology, music, educational songs, influential composers, and self-propelled tribute albums.

So you're working on a new album?

Yes, we are. The as-yet-unnamed record is being recorded at River Sound in New York. This week I'm in my home studio listening to rough mixes, and I've been cutting up and pasting little bits of music to try to create a bridge for one song.

When you use a computer for songwriting, at what stage does the computer enter the process? Do you write at a piano?

I use the sequencer to construct the form of the song. That's its main role as far as composing. I'll take sections and move them around and figure out what the complete form of the song is. The sections I come up with just playing a keyboard and singing, or just singing, or playing the accordion and singing. But when I start assembling it in [Mark of the Unicorn] Performer, it's time to copy and paste sections. That's one of my favorite parts of writing, actually, coming up with the form.

Has the computer changed your songwriting?

Oh, sure. It's definitely made it easier. The thing about writing is, I'm always curious to know if I can go from one section to any other particular section in the song, whether that's going to make musical sense or not. With the sequencer it's really easy to experiment.

What are some of the songs on the new album?

We've got a whole bunch of new songs -- one called "S-E-X-X-Y," one called "[[James K. Polk]]" -- you know, the American president, which was formerly a B-side.

I've always liked how you guys bring up obscure characters like John Henry. I guess James K. Polk isn't that obscure ...

I'd call James K. Polk pretty obscure. I mean, yes, he was the president, but I didn't know anything about him when we wrote the song. My friend Matt and I came up with the song idea, based on these educational songs that are in the back of people's minds. Maybe no people of the younger generation, but certainly people in their 30s and older remember songs like "The Battle of New Orleans."

The thing about those songs is that they're kind of inane. They're not entirely accurate historically, but they do make you remember stuff that you never would otherwise. So we were musing on that fact, thinking about really obscure people. We thought, let's pick a president who was unimportant, you know, like a caretaker president, and write a song that elevates him to a higher status. The thing about James K. Polk is, we found out that he was actually extremely important. He presided over the last really big expansion of the West. He began the war with Mexico that ended in California, New Mexico, and Arizona being ceded to the U.S., and he also bought the Oregon Territories, which is now Oregon and Washington.

What else is on the record?

We've got a new song called "They Got Lost," which is about TMBG. It's based on real events that happened to us, slightly fictionalized. It's a real show-stopper. Then we have "You're Older Than You've Ever Been." And, let's see, Flansburgh has just written a song that I think is a complete powerhouse. It's called "How Can I Sing Like a Girl?" I predict that this song will put our kids through college, or put us through college, or something. Or it'll make people want to go to college.

Do you write songs together or independently?

We pretty much each write entire songs, but there are a few exceptions. Like another new song called "Rat Patrol," which is my music and his words. Do you remember that TV show? It was this war show with these guys in North Africa driving around in a jeep.

How did you guys get started working with computers in your music? As I understand it, it was a practicality at first. You didn't have a lot of gigs, so you needed to hold down the number of musicians you rehearsed and performed with.

That's correct. We did have gigs, but we had decided to work as a duo. But since neither one of us was an expert drummer, we had to have some way to perform. So we had a show with tapes playing back the rhythmic tracks. At first we had a cassette machine, then a reel-to-reel, and then at the very end of that period we had a 1/2-inch 8-track machine. You know, one of those big Otari units, which is a real bitch to lug around. And the silliest part is, we had a backup, which we never actually used because the main one never broke down.

You lugged two Otari 8-track machines around with you to gigs?

Yep. We had two. That was hard work. Nowadays we wouldn't have to do that, because 8-tracks will fit in the passenger seat of your car. In fact we both have TAS-CAM DA-88s now. It's like the [Alesis] ADAT machine, but we think it's a better product. It should become the standard for tiny little digital 8-track machines, but the problem is that the name "DA-88" is so hard to pronounce that immediately the ADAT has a tremendous advantage.

So when you performed as a duo at first, you would use the tapes just for rhythm tracks, playing along with them live?

Back when we were young? Yeah. We had some instruments at our houses, and we'd have a drummer come by and record him. And the bass parts were mostly a Micromoog [synthesizer].

So on "I Hope That I Get Old Before I Die," from the first album, that country-sounding bass was a Micromoog?

That's what that was. But by the time we were making our first record, we were using a lot of other things, like a Casio CZ-101. Great, great instrument. That was the portamento noise at the beginning of "Chess Piece Face." We were pretty heavily into the Casio scene at that time.

I wish Casio hadn't stopped making professional instruments. They had some really cool things, like their MIDI guitars.

They made this great thing, which we bought two of,'cause we figured they wouldn't keep making them -- these MIDI horns. They were unloading them at Sam Ash for like $39! We used one recently in fact, on an XTC cover.

Which song?

"25 O'Clock," by Dukes of Stratosphear. [ Ed. Note: This group was XTC's psychedelic alter ego, circa 1987. Check out Chips from the Chocolate Fireball, their compilation on Geffen Records, if you can find it .] It's on the upcoming XTC tribute album.

"John Flansburgh steps up to the mic with his Martin, as seen from a Connectix digital camera."

I hadn't even heard of that one. This "tribute album" business is getting out of hand.

I know ... it's crazy. Our next album after this is gonna be a They Might Be Giants tribute record. We're gonna record all our own songs.

Really?

No, just kidding. I think at the point where Kiss was supervising their own tribute record, it just seemed like the spirit of the whole thing got confused. There's some breach of protocol when you supervise your own tribute.

But getting back to technology, the point at which we really started getting into computers was when we did our second album, Lincoln , at the insistence of our producer Bill Krauss. It was something that we were attracted to as soon as we found out about it, but we had to be led. Bill had a Mac Plus, with Performer version 2 point something, and as soon as we sorted it out, the advantage was immediately obvious. For a while, I would go over to his apartment and do sequencing on his computer, with a tiny little keyboard. So some of the pre-production on Lincoln was done that way. Eventually John and I got computers and started doing it ourselves.

Have you continued using Performer this whole time?

Yes, we've stuck with it. It's interesting, really -- we got to know Performer inside and out up through version 4.2, which is now a couple of years old. Then they came out with version 5.0, and I got the literature and listened to what people were saying, and it occurred to me that it would probably make my computer run a lot slower, and the features that they'd added were not particularly useful for me. So I'm still using 4.2.

Actually, John just bought version 5.0. I think he was having problems with 4.2. He said my fears about 5.0 were confirmed, because there's this proprietary nonsense, like you have to use their own little MIDI system thing...

FreeMIDI?

Yeah, FreeMIDi. The other thing he said is that formerly you could just send MIDI channels to tracks, but now you have to assign devices. That's exactly what John doesn't want to do. The way we use Performer is that we take sequences into the studio and sometimes use them on different gear. You know, we're not just working in one studio all the time, so it seems that by making you define your setup like that, Performer has lost some of its flexibility in keeping things simple, and not having to know your entire studio in order to work. But I imagine there's a workaround.

You talked about touring in the early days with tape machines for a backup band. Was there ever a period when you toured with a computer sequencer?

No, we've never done that. There was a point where we considered it. We were using tapes in the live show and we had started using sequencers in the studio, and we thought, "Maybe if we travel around with a computer, we'd have more flexibility. We could change the show around while we're on the road." We started to think about what it would be like, and it occurred to us that the computer we were using was so liable to crash, particularly in the hectic environment of the live show, that it was just a completely crazy idea. We've all had our share of computer mishaps.

We had a computer go down on us just because of the heat in the control room when we were mixing our second album. We were using a Mac 512, which was the state of the art at that time, and one day there was this kind of horrible urine smell in the air, which nobody could account for. We were involved in some kind of heated discussion about the mix, and we looked over and there was a thick column of white smoke pouring out of the top of the computer. We looked at the screen, and there was something on it that nobody could describe later on. Some kind of demonic shape had appeared on the screen in place of the friendly Performer interface.

What type of computer do you use currently?

Macintosh Ilci's, both of us. Although in the studio this week I've been using my PowerBook for some sequencing.

Maybe Performer seems slow because you're running it on a Ilci.

Yeah, but the features still seemed not useful. I don't need color, I just need the thing to run properly and not take forever to do things.

So your approach is to use MIDI sequencing to compose and rehearse music before actually working with the other musicians?

Well, yes, but in most cases the sequences actually became the songs. The rhythm tracks on Lincoln, Flood, Apollo 18 , and Miscellaneous T were all Performer tracks.

I couldn't tell if a lot of those songs were using live musicians or not, especially on Flood . I could've sworn there were live drums on there, on "Lucky Ball and Chain," for example. Very lifelike flams.

Well, that's nice--it worked! Occasionally throughout those four records there are some live musicians, but all the bass and keyboards, and most of the drums, are sequences. On Flood, we did have a real drummer, Alan Bezozi, come in and play some MIDI drum pads [into Performer for] some of "Birdhouse in Your Soul," but playback was straight from the computer. He also played a live snare drum on "Particle Man."

Some of the engineers and producers we worked with, before we met them, had listened to Lincoln and said, "Wow, they got a new drum machine." But the accordion, guitar, clarinet, and most of the horns are all live throughout our recordings.

What's going on now?

Computer-wise? Well, the last record we made, John Henry, was a complete departure from those methods. We only did two songs with sequencing: "Destination Moon," in the middle section there's a brief moment of synth stuff that's not humanly playable, and then in "[[Snail Shell]]," it's part drum machine and part human being. The rest of the record we actually had the whole band in the big room at Bearsville Studios. It was very different from anything we'd ever done before.

But you still wrote the songs on the sequencer?

Yes. We sequenced the songs so we could show the parts to the other musicians. We gave them their parts on cassette, but at that point we stopped using computers and went with the live recording. We're still writing on computer, though.

It seems to me it's often best to record with live musicians, but to use computers for the things they can do best. That's why Frank Zappa preferred to record with the Synclavier rather than real musicians a lot of the time -- because it could play anything he could think of, no matter how fast it was.

Right, but that's a very specific criterion for what's good music. I like complicated music, you know. John and I definitely listened to Frank Zappa when we were teenagers, and there were a lot of things that were appealing. But I think we probably learned a lot from his mistakes. In some ways, it seems he was just obviously too much of a control freak, and the best thing about what he was doing was the human aspect of it.

And also the humor that he had, and you guys have too. You can't really program that.

Yeah, although I never cared for his scatological humor. It's funny -- I really liked him when I was a kid, Weasels Ripped My Flesh and all that, but there was a point where I really got turned off by his attitude. It was kind of condescending after a while. It makes sense to me that he got so involved in composing on the Synclavier...

He was probably difficult to deal with, if you weren't a musician of his caliber.

Exactly. When you think about stuff that's entirely controlled by the composer, it seems like the elements that are appealing about it are where the materials, the instruments, are pushed as far as they can be. The guy I'm thinking of is Conlon Nancarrow, who composed piano roll music for player pianos. It's in the category of American avant-garde, you know, but the thing that's appealing is that he chose a medium that no one else was really interested in, and really wrung his own thing out of it. He wrote for these supercharged player pianos. Most player pianos aren't able to play all the notes at once, but he ... I don't know how he modified them, but he made it so they could do these thunderstorms of music. Extremely dissonant, a lot of swipes up and down across the keyboard and stuff like that, but ultimately it's about the relationship between his ideas and the materials he was working with. He was limited by what the player piano can do, but he used that limitation to his best advantage. The thing about Zappa is, he always seemed like he wanted to play eight bazillion 128th-notes, which you could say is kind of interesting in itself, but it was definitely better when he had real musicians playing it.

Tell me a little about your equipment, both computer and non-computer.

Let's see. We have an endless pile of MIDI gear. Like I said, the early records were heavy on the Casios, and at a certain point John and I both got Casio FZ-1's and started getting into sampling with those, and eventually traded up to Akais, which we use now. We both have [Akai] 1000KB keyboard samplers, 'cause we got a deal on them, and I just got an Akai S2000, which is the same thing but with more memory. They sound great, just fantastic.

You also said you're working with a notebook?

I've got a PowerBook 145B that I've been bringing into the studio and sequencing with a very minimal setup. It's actually kind of appealing--in my little bag I can carry the PowerBook and this Yamaha MU5 [General MIDI module]. They're really great, and it works as an interface too, so I can bring that along and I've got the whole setup. That and this little Opcode Time Code Machine, just about half the size of a videocassette. It takes SMPTE and turns it into MIDI code. It's what I use at home, too: I just bring it with me.

Were any of the songs on the new album written using just the PowerBook and the MU5?

I did some writing last summer using that setup, but it was instrumental stuff, it wasn't TMBG stuff. I haven't done any real pro sequencing with that setup. I listen to my sequences with it when I'm at my house, not in my studio. And it's useful when I'm making charts, because the notation software plays MIDI as well. I think we used a tambourine sound right out of the MU5 for the version of "Spiraling Shape" that's going to be on the record.

Digital self-portrait of John Linnell at New York's River Sound. The Apple Powerbook 145b is running MOTU performer and synced to the multitrack tape recorder. In front of him is the "sensational" Casio CZ-101 used on many of TMBG's recordings.

What notation software are you using?

I use [an old version of] Mosaic, which evolved out of Mark of the Unicorn's original Composer software. I started using that a long time ago. I have a copy of [Coda] Finale, but I tried to learn how to use it and it was just too complicated for what I needed. I have a friend who uses it, and it's obviously better, but it was going to take me too long to learn. So I'm using this much simpler program, which is kind of buggy.

What about your live setup?

Live, I use a Roland A-30 controller and this wooden stand that a friend of mine built, with Danish modern table legs. It's very nice. Built into the stand are four rack spaces, so currently I have a Proteus FX [E-mu synthesizer] in one space, and a backup in another. I got the Proteus because it basically does everything I need.

I've noticed quite a number of vintage keyboard sounds on your records, like on "Twisting in the Wind" [from Flood]. I thought it was a Farfisa.

Oh yeah! That's actually an old favorite of mine--it's the four-operator version of the Yamaha DX7, the TX81Z. That was a single-rack thing with some really cheesy sounds. It does sound like a Farfisa [electric organ], though. I used to own one of those but it died.

Any other MIDI devices?

Well, at home I use the Roland U-220, a rack-mount synth. Also--this is one of the last things Casio made--a VZ-10m, which can actually make CZ-101 type sounds. It's very programmable. The sounds that come with it aren't so good, but it's really easy to make up sounds. Oh, and I still have my old Micromoog, which we haul out now and then, but it's not MIDled. We used it on Apollo 18 , and probably on Flood , I can't remember.

What do you think is better done by either computers or musicians?

Well, computers are good for any technical thing. Here's an example: This week we did a song where the whole band was set up, and the keyboard was playing into the sequencer while we were recording. So we were recording both the band and the sequence, and the guys could hear me while I was playing. Later we can mess around with the sequence if we want to, but the important thing is that we can all hear each other while we're recording.

This band really listens to each other. I'm kinda proud of that -- we're at a point where people know each other's thing, and interact well. Our drummer and bass player are sensitive to little "feel" things that happen. So when I'm playing the keyboard and the tempo breaks down, they can hear that really well.

I've noticed a lot of breaks like that in your songs, where the band stops and the keyboard continues underneath, like in "End of the Tour," on John Henry.

In our live show right now we do that, and also a lot of improvisational things. There's a song called "Spy," for example, where during the last part of the song, the band improvises while I conduct. There's a couple of different cues, like a number of fingers for hits, a signal for one person to start doing something, and then everyone else will join in once that's established.

Who are you playing with now?

The bassist is Graham Maby [from Joe Jackson's old band] and Brian Doherty on drums--he played on the last record--and Eric Schermerhorn on guitar, whose main gig is with Iggy Pop. We haven't figured out what do about horns yet.

What about Dial-A-Song [TMBG's phone line that plays unreleased songs]?

Dial-A-Song is still going, same number [718-387-6962].

Speaking of music and computers, Dial-a-Song is now running on a voicemail system on a Macintosh -- I think it's a Mac Classic -- in John Flansburgh's kitchen. We've had quite a few headaches, actually, with the voicemail software. The company abandoned the product before it was really finished. It's been an unbelievable nightmare. The computer kept crashing. The thing is, there was only this one company that would allow us to have a three-minute outgoing message. And originally there was no company that supported an outgoing sample rate of more than 7kHz or so, which was so horrible. It sounded worse than the phone answering machines that we used before. We thought we were going high-tech, but it turned out to be a step down. But we've overcome that, we're now up to a 22kHz rate, and there's about 30 songs on the hard drive. They change automatically every hour or so.

How often is it updated with new songs?

Well, we've got a whole batch of new songs on there, with the new record coming up. The policy of Dial-A-Song is that we've always used unreleased material, so once the record comes out it'll go back to the handful that were on there before.

What about the weird edits on your songs? Are those tightly planned, or do they come about as a result of studio experimentation?

Mostly from experimenting in the studio, but it's fairly easy to do stuff like that, so we'll often have complete demos done at home that already have all that crazy stuff on them. I did a demo at home of "Fingertips" [from Apollo 18], which is a collection of little songs that are all about ten seconds long. I sequenced each little one individually, recorded them onto a regular open-reel 2-track, and then spliced the tape together with a razor blade the old-fashioned way.

How do you get some of your guitar effects, like the noiseless, gated sound on "Ana Ng" and the tremolo on "Mr. Me" [both from Lincoln ] that stays perfectly in time with the song tempo?

Oh, yeah! As soon as we realized what we could do with the sequencer, we thought, here's this thing that is very mechanical and precise, and we can use it to control this noise gate. So we set up the sequencer to drive this synth that made a beeping sound that turned the noise gate on and off at a very exact interval. So John's guitar signal was essentially turning on and off as he strummed. He didn't actually play those stops like that.

It was the same thing on "Mr. Me." We just set the gate so it wasn't shutting off all the way, so it sounds like a tremolo. It's doing the same thing a tremolo would do, but the timing is controllable. Geez, I'd forgotten all about that. I guess we were just really noise-gate-happy at the time. In those days we were much more prone to seeing a piece of gear and thinking of the most crazy thing we could possibly do with it.

I was puzzled by that effect, having messed around with vibrato and tremolo and knowing how difficult it could be.

I think at the time we were sort of influenced by that Smiths song, "How Soon is Now?" I think they just recorded the tremolo guitar on the original track and then the band played along with it. That's a cool recording. I love the Smiths, although sometimes I get the feeling that I have a different response to them than most people. The record didn't make me cry or anything. I just thought they were really witty and interesting, but I met people who felt some deep emotional response to what they were doing.

What about the song "Turn Around," from Apollo 18? There appear to be some sped-up vocals towards the end.

That was supposed to sound like the Modernaires, the vocal group that sang with the Glenn Miller Orchestra, like in "Chattanooga Choo Choo." Not like the Chipmunks. It's all me singing; we sped up the tape for half of them, for the notes I couldn't reach. It's funny that most of the more innovative stuff we did with computers was with the second record we ever made [ Lincoln ]. We were obviously much more excited about exploiting it.

What about the distorted vocals that you do sometimes, like on "Stomp Box" [from John Henry]?

Do you ever listen to Indian film music? For some reason they always do this routine where they distort the vocals. It's particularly obvious now that the backing music is done with a lot of MIDI equipment. It's very clear, but the vocals still get all distorted. I think it's fascinating. There's this woman named Lada Mangeshkar, who's like 80 years old, and she sings on almost every film score. She has a really high, strong voice and it always sounds like she's overdriving the microphone. We weren't influenced by that specifically, but we think it's a cool sound. At first the recording engineers flinched when we wanted to do that, but now it's very accepted.

Have you ever thought about writing a song that depicted the impact of the computer on society?

The most interesting and important things I can think of to say about the computer would be pretty negative. But I don't know if that's enough for the kind of work we do. That's not really what we generally talk about. We're not exactly into social criticism, though with Flansburgh's songs there's a certain amount of that.

For me personally, the computer is really central. It's the way I work and have been working almost for the whole recording history of the band. I don't think it's indispensable: It'd be fun to write and record a whole album without using any computers. But even the first time we did an album with the band, which was the last album, John Henry, while we weren't really using the computer in the studio that much, we still did all of our demos using Performer. We gave those cassettes to the band, and they learned the parts that we had originally recorded using the computer. So we're oriented that way. I think the computer appeals to control freaks like me and John, in spite of the fact that you sometimes feel that the tail is wagging the dog a little bit, that the computer is dictating how you're going to work. Also, the problems that you have with the computer tend to occupy a lot of your day. My Ilci is really crashy, and I still haven't sorted out what's wrong with it.

One more thing. Does it ever seem like everyone's named John? I mean, me, you guys, your manager....

One tour we did, almost everyone was named John. Every club we went to, everyone was like, "Aaaah! You're all named John!"

Sometimes in big crowds people yell out "John!" and I think it's for me but it isn't.

Never?

Sometimes it is.

I'm glad to hear that.

Recording With Computer & Tape

When John Linnell says, "We were recording both the band and the sequence,"" here's what he means. The acoustic instruments (including electric guitar and so on) are recorded to tape as conventional audio tracks. The MIDI keyboard parts, however, are recorded to tape as audio and at the same time to the computer sequencer.

For this to work, the sequencer must be synchronized to the tape using some form of SMPTE time code or MTC (MIDI Time Code, a "MIDI-ized" version of the SMPTE signal). When the two machines are synchronized, they will always start playback at the same point in the music. Using a protocol called MIDI Machine Control, it's even possible to control the tape deck's transport buttons from the start/stop/record buttons on the computer screen.

After the recording is finished, the keyboard parts can be freely edited in the sequencer to clean up wrong notes, fix bad timing, or completely alter the part that was played so as to fit better with what the other instruments played. Before mixdown, the new keyboard part can be recorded to tape, either as a "punch-in," replacing the old keyboard part, or to a new tape track. Alternatively, the keys can be retained as a "virtual track" until mixdown, at which time the sequencer will play back the MIDI parts while the tape plays back the other instrument parts. The engineer can bring the MIDI parts up on some of the mixer channels and the tape tracks up on the other channels, and treat them as if they were all coming from the tape.

With some of the newer computer-based recording systems, which are tapeless, both the audio tracks and the MIDI tracks live in the same program and can be edited in much the same way. Thus the need to wrestle with sync signals and formats is eliminated.