Electronic Percussion (Part 1)
Self confessed synth merchant Dave Stewart reveals how to do away with drummers.
It is alleged that on or about the beginning of this decade, you did wilfully, wantonly and with music aforethought, excise from all records of this parish, a real live drummer, and that you did frequent the abode of drum machines and partake of matters electronically percussive. His Lord Andrew Duncan shall now read your statement to the court.
It all started with the Claptrap. That was the first rhythm generating mechanism that I got, and that was because I knew Dave Simmons. He told me about this Claptrap that he was working on, I thought it sounded really bonkers, but when I went to have a look at it the sound was great. I took it home and started messing around with it — at this point I'd just left the Bruford band and was keen to get into ways of making my own rhythms. The little boxes I'd come across in the past weren't any good, but the Claptrap had a real vibe to it.
The first thing I did was put Chorus Echo on it so that one clap would create a second. By turning the echo on and off on every other beat I got clap—(clap)—space—clap, and using that rhythm I came up with the demo of 'What Becomes Of The Brokenhearted'. Humble beginnings, but just using that and no other rhythm was enough for me to be able to come up with a lot of interesting keyboard parts. If I'd had a drummer there, filling everything in, like they do sometimes, it would have been difficult to have had that kind of freedom.
Although that sounds very simplistic, it was a real breakthrough because in the past I'd always thought that if I was going to do a tune, I'd have to use a drummer. It was radical to think that just a clap was enough.
I was aware that Dave Simmons was working on electric drums and I asked him to let me know as soon as he came up with anything. When he called I went to see him at St Albans, and he had these strange hexagonal mechanisms set up. He whacked away at them, and these great sounds came out of the speakers. He sounded like a drummer. Not bad since he's a keyboard player.
He explained that any drum sound was composed of a skin-sound, which had a distinct note, and a stick-sound — a sort of 'baff' — and that they had come up with ways of simulating that electronically with different noise generators.
So I had them make me some pads which I could play without feeling like an idiot. They were four, semi-circular pads set in a briefcase-shaped flight case. I had no drumming technique, I couldn't use my feet, but what I did was to get the Claptrap to play tempo and bash along with it on the pads. Using this sophisticated method we did 'It's My Party'.
This all sounds very old hat now, because everyone's used Simmons kits. But this was back in 1980 when there weren't too many of them around, and at the time I felt that it was Another Great Breakthrough. At this point I evolved a small amount of drumming technique — just enough to make engineers turn white when I brought the pads into the room — but still limiting enough to hold me up.
About this time I heard about the Linn drum computer, and when I got hold of one in early 1982, I thought it would be good to get the Simmons going, programmed by the Linn, so that I could put down patterns without having to annoy everyone and strain muscles in the process.
Then came the problem that many groups have faced since: how to get the Linn to trigger the Simmons.
Theoretically, the Simmons needs a 15 volt spike, or the output from one of its pads, to produce a trigger. If you just take the Linn audio outputs, say the snare, and plug one into the Simmons snare input, you do indeed get a trigger, but it has a funny 'front' because it's an audio message. That isn't really the sort of signal which will lead to that typical Simmons sound.
So what we did was very painstaking. We found that if you took the clave sound on the Linn and turned it up to its highest pitch, it triggered in a similar fashion to a pad. So we would lay down a Linn code, and then record each drum, one at a time, using the clave to trigger the appropriate sound. So if we wanted to play a Simmons bass drum, we'd plug the clave into the Simmons bass drum, and set up the pattern. Then, using the Linn code to sync to, we'd do the snare drum, and so on. It took ages. If we did one drum track in a day, we thought we were doing quite well.
Apart from the complication of having to do each drum separately so that you never heard it all together until it was too late, we were also using different tempo tracks, and on a Linn the tempos aren't programmable. So if we wanted the chorus to go faster we'd have to stop and lay a different code on another track at a faster tempo. When you came back, you'd have to have that code driving the Linn, and then you'd have to drop in the drum sound at the appropriate moment. You can imagine how long that took.
What we really wanted at that stage, was to have a machine into which you could program the tempo, then that would trigger the Simmons. That way you could have all the sounds without needing to record them one at a time.
But we couldn't find any such machine, so our album took a long time to do. Most of it was recorded with the Linn driving the Simmons, and occasionally I'd bash along with the Claptrap which, by this stage, was considered old fashioned.
At this point I found there were hardly any Linn sounds that I liked. The only ones I did use were the claps and the tambourine. I used to use the hi-hat as a simulation of a real hi-hat, but I felt that the sound was very gritty — not great, but it had to do because there was nothing else you could actually program to play the open and shut sounds and do all those things that drummers could do.
I still didn't want to resort to using a real drummer because although I was aware that I was sacrificing a lot in terms of feel and the human element, I still had total control of it which I thought was very important. I was also able to put together a few patterns which wouldn't have sounded right played on a drum kit, but which actually sounded pretty good played on a Simmons. All of which struck me as being quite significant: the fact that you were coming up with something which was different from a drummer.
Still, the Linn had marked limitations. You couldn't program the tempo and you couldn't record your own sounds inside it without sending a tape to America where the cost-conscious digitising process inevitably changed your original sound.
There are now machines on the market on which you can program different tempos like the Oberheim DMX, but you still only have the numbers that they provide, like 110, 111, 112 beats per minute. You can't go in between them, and the difference between those numbers is quite a lot. You can't program a bar to speed up gradually, you have to cut from one number to another.
The Simmons' sequencer is the same in that you can enter different tempo values, but they don't even approximate to BPM, they just have arbitrary numbers like 35 which is a tempo of some description, but God knows what it is in BPM. You'd have to sit there with a stop watch to work it out. Again you can program it to go from, say, 35 to 36, and I suppose you could program one beat at 35 and another at 36, but you wouldn't get the feeling of a smooth transition. You should be able to write these things in with a curve to them.
Meanwhile, I still had to get round the problem of the Linn driving the Simmons. Enter Mike Kemp, an engineer and one of the designers of Spaceward Studios. He used an oscilloscope to look at the spike which came out of a pad and tried to simulate it by designing a circuit which would drive the Simmons in the same way. He then built seven in a box, one for each channel of the Simmons. This trigger box would take seven different inputs which would be seven different Linn outputs. So you could take all the Linn sounds, plug them into the seven trigger box inputs, take the seven outputs and plug them into the SDS5 and at last the Linn would trigger the SDS5.
He also built in a hold-off circuit which would ignore a spurious signal in the circuit for a certain length of time. Let's say you've got a snare drum on tape and it sounds like crap and you want to put in the Simmons snare instead. You plug the snare on tape into this box which will generate a pulse that drives the Simmons sound. If there's a bit of grot on the tape, like someone coughing, or a guitar solo, you can set the hold-off so that once it's heard the first beat, the hold-off activates for X seconds and goes off just in time to let the next beat through, so you don't get any double triggering.
If you've got a sound which goes (blows elaborate and lengthy raspberry) and you want it to go 'baff', again you can use the hold-off button to play the first beat, stop, wait the set amount of time, and come in with the next one. I couldn't see the point of it when he explained it to me initially, but it's turned out to be really useful.
Then I started thinking about triggering the Linn off tape. Mine is the first, the LM1, and it's got no trigger interface, so Mike then built a jack field which corresponded to its front panel. For every drum on the front panel there was a jack socket and a multicore link. Then he built an output on the trigger box carrying a 15 volt spike. If you plugged it into the clap socket on the Linn and coughed down a mike into the trigger box, the Linn would go 'Clap'. That made the Linn triggerable. You could take any sound and use it to make the Linn play.
This left one big problem. The Linn has a mode called 'high resolution'. Think of it as being a cogwheel. The beats you play always have to fall into one of the cogs. If they aren't close enough together, a beat gets moved into the nearest free cog. The trouble is, you don't play like that.
The way round it is to have the Linn running slowly, play the pattern you want very slowly, and then speed the whole thing up to the right tempo, so minimising the discrepancies.