Why I Painstakingly Built My Own Analog Drum Machine

Editor's note: LA Priest, aka Sam Dust, is known for his singular take on cosmic, funky, psychedelic pop. While writing his latest album, he did something truly unusual. From scratch, he created a drum machine, which served as both the rhythmic center of the album's songs and its namesake. He named it Gene.

We asked him why, in a world of readymade drum machines, he felt compelled to make his own. What did he want from Gene that he couldn't get from other devices? And what did he learn in the years-long building process? He tells us in his own words below.

On an autumn day in 2015, I was wandering in the hills around my house in North Wales trying to imagine the perfect drum machine. Earlier that year the first LA Priest album came out. It took seven years on and off to make it, and I had written it mostly by just waiting for an idea to pop into my head, so I knew that I really needed to start writing faster in the future.

LA Priest. Photo by Isaac Eastgate. Used with permission.

I did sometimes have these rapid bursts of writing when I got a new instrument that I found inspiring. I even built a few things for Inji [released in 2015], but they would only really inspire one song each, because they would only do one sound really well. Some of them didn’t exist for long, because after I recorded a song, I would just tear them down to build something new.

I had made only one drum machine before Gene, as part of a modular synth that I used when I was producing the Xenoula album [released in 2017]. It was really basic, but you could do a lot with it if you combined it with lots of the other synth modules. I used to send these simple little drum tones through an [Analogue Systems] RS-240 frequency shifter and they would sound massive. That’s where I got the idea to add a frequency shifter (mine is actually a modified ring modulator) circuit into the Gene. I sort of recreated that giant modular synth inside the drum machine in miniature.

The first thing I ever tried to build actually was a tape-loop based drum machine. I spent two months building this giant mess of wires and switches without testing it once. When it was all done, I realised that if it didn’t work I would have no chance of knowing how to fix it, because I'd never studied anything about electronics. So I turned it on, and it didn’t make a sound. Years later I figured out that I hadn’t sent power to any of the chips. Anyway, after that I made lots of small synths and things and built up gradually.

I wanted a drum machine because I was getting back into playing guitar and needed rhythms to play over. I had bought and sold countless drum machines, but I found them all way too rigid and cold-sounding to really go with a guitar or be right for my songs.

The first Gene prototype, 2015.

So going back to that autumn day, I was wrestling with finding a way to make an analog drum machine with totally free timing, where you could put the beats anywhere you like. Then I had an idea, and I did that stereotypical idea-having thing of running back to my house and scribbling on papers. I worked out how to build it with timing “sliders” for each beat, so that you could mould a rhythm with your hands, almost like if you were making it out of clay.

I made the first prototype later that day: It had two drums that you could move around anywhere you liked and they would just loop round and round infinitely. So I sat listening to those two drums for hours! I felt mad to have made this and to imagine where it could go.

I would sometimes spend a whole day just, for example, playing a snare noise over and over again, changing one part in the circuit at a time, then record the sound—repeating that and comparing the recordings. I'd say I got pretty unhealthily obsessed at some points. It’s hard to know when to stop. It was very similar to making a piece of music, except at the end of each day I would only have a few milliseconds of sound and a bad back to show for it. But I did it for months. All so that I could hopefully make music really easily at the end of it.

The second Gene prototype in action.

The first truly musical version of the machine had 12 self-oscillating feedback loops all tuned to different frequencies, which would be routed to four rotary switches to choose the frequencies you wanted. Then four sliders would set the timing of each of those four channels, with a decay, volume, and noise control for each channel. The noise control introduced white noise into the feedback loops, which made these really strange, mechanical-sounding, whistling drums.

You can hear that sound on lots of bits of the album Gene (especially at the end of the song “Black Smoke”), because I kept that prototype working and built it into the next version of the drum machine that I ended up using all the way through making the album.

I made one more version in between recording sessions for the last songs to go on the record. I designed a way more elaborate sequencer where you dialed in the sound instead of the timing for each beat, but using the dial below that you could shift the beat over a range of +2 beats, meaning you could still make these insane time signatures, but you could instantly rearrange it into a standard quantised rhythm by turning the timing dials to the left.

The first finished Gene.

I put the most effort into that version. I wanted it to be the cover of the record so I gave it a square box and even got the titles of what I thought were going to be the singles engraved into it. I also added this feature where you can insert "sound cards" into the slots on the front panel, so that you could keep making new sounds for it. It could keep making new sounds forever in theory, maybe even be used on another album! (That would be a first for me though.)

After I finished the recording, I realised it was probably pretty rare for a musician to do what I had done, or at least to take it as far as I took it.

As far as I know, it’s the only drum machine that you can change the timing of each beat down to the millisecond. It also uses 100% analog circuitry, which is sort of rare, and each one is built by hand, actually by my hands. There are lots of other things that are completely unique in its design: Every part of its circuitry was trial-and-error, so it ended up with some really unusual things going on in there. It's pretty hard to explain how some of it even works.

I did get a lot of good feedback for the nearly finished machine when I joined the Muffwiggler forum. I put up a few sound clips and asked them if they thought it was any good, and I started getting a lot of messages every day. I wasn’t really prepared for it. I decided to sell them some circuit boards so that they could build and test out the design for me. I ended up selling about 30 of those but the questions would take me hours to respond to sometimes, because they wanted to know everything. It was exactly what I needed though. They spotted some design mistakes that I would never have noticed.

The second Gene prototype in action.

It became the main source of inspiration for my songs. I would sometimes even just set all the controls randomly on it and hear a song idea in all the chaos. Other times I would just have one drum repeating and build a song over that. Even though I expected it to work pretty well for me, since I chose every part of it, sometimes that doesn’t work out, because you can stick too rigidly to what’s familiar to you and end up limiting your possible outcomes. But the drum machine ended up having enough of a mind of its own for that not to be the case.

I did originally want it to have a full synth voice inside it, so you could have this wide tonal variation across the rhythm. I did get that a bit with the frequency shifter on the last version, which has an envelope generator so it changes pitch throughout the bar. But I’d like to maybe make another version of Gene that is just a couple of synthesizer voices that can be pitch- or volume-sequenced to make drums rather than triggering static sounds.

I think the realm of drums/percussion is still somewhere relatively unexplored. Our ears are really good at analysing the first milliseconds of a sound; I heard that we adapted that to decipher speech, but I think it’s why we can also tell millions of different drum sounds apart too. I don’t think I’ll ever get bored of drums. I'd like to go back to a few of my early plans that never got built though; electroacoustic keyboard things, some kind of Ondes Martenot–ish designs and definitely the tape-powered drum machine.

The miniGene.

The miniGene [which can be found for sale here] started out as an experiment to shove the best handful of features of Gene into a pocket-sized box. I went back to the prototype sequencer where you can use one dial to move a beat anywhere you like. It has two of those for each drum, but then a switch lets you repeat or not repeat them in various combinations.

I also adapted the drum sounds and everything because it runs on a 9V battery or power supply. I think the lower voltage actually makes some of the sounds better in a warmer way, which I didn’t expect at all.

Finally, I threw in a few new tricks for good measure, so the snare drum can make three different sounds in one rhythm; it can also be two different pitched toms. Then you can change the decay of all the drums and push them into this warm gooey distortion at high volume settings, so the sound can turn into something really unrecognisable.

It’s ended up being a lot of fun for only having a few controls. I think limitations always bring out the best weirdness.

comments powered by Disqus

Reverb Gives

Your purchases help youth music programs get the gear they need to make music.

Carbon-Offset Shipping

Your purchases also help protect forests, including trees traditionally used to make instruments.

Oops, looks like you forgot something. Please check the fields highlighted in red.