Tech I by I 06.10.18

Mark Fell on his love of FM synthesis and algorithmic composition

Signal Path is a new series that delves into the creative process of our favorite producers and musicians. In this interview, Scott Wilson meets Rotherham-based producer Mark Fell, who talks about a lifetime of experimentation with FM synthesis and algorithmic composition using Max/MSP.

“We were just naive and stupid,” Mark Fell tells me from his home in Rotherham, recalling how a catalog number for the SND project with Mat Steel inadvertently become their name. “We were just making music and it never occurred to us that we needed a project name. If you go round someone’s house and cook a meal, you don’t say ‘and now I am unveiling this meal and it is called this.’ You just cook a meal and eat it.”

Fell’s wry, matter-of-fact approach to navigating the music industry is much the same now as it was in 1998, when SND’s Tplay redefined minimal club music with a stripped-back combination of lush house chords and crisp 2-step beats. And whether he’s using software like Max/MSP to create algorithmic composition processes, or exploring specific FM synth presets, his music is made on his own terms. Out of these occasionally obtuse concepts, he’s made some of the most inviting and accessible experimental music of the past 20 years and carved out a wildly varied multidisciplinary career; he’s just as likely to be found making installations as he is playing club shows.

Although Fell has been busy with different projects over the past decade – most notably the deep house-oriented Sensate Focus alias and collaborations with Errorsmith and Gabor Lazar – this year he released Intra, his first proper solo album since 2010. Written for Portugal’s Drumming Grupo De Percussão, who performed it on a metallophone system devised by Greek avant-garde composer Iannis Xenakis, it’s a multitimbral, ASMR-inducing delight that does strange things to your brain.

Despite an impressive CV that includes academic writing on the compositional process, Fell has no formal training. “I still don’t know anything about the theory of music,” he says. “If you show me a musical score I have no clue what it would means. I don’t think making music should be the domain of five hyper-nerds. It should be something that everyone can do. Ultimately music making is just about getting together and having a good time for a lot of people.”

Fell’s inclusive attitude to music extends to his curation of FACT’s stage at this month’s No Bounds Festival in Sheffield, which features Ugandan artist Sounds of Sisso, Hyperdub’s Klein, Sarah Davachi, Theo Burt and WANDA GROUP.

“In the age of Brexit and Trump, I think it’s really important that we champion how brilliant multicultural society is,” Fell says. “Some of the best music in the world comes from Britain – especially dance music – and I think that’s because we have huge amount of different cultures in the country. It’s an opportunity to make that statement as big as possible to counter the kind of hostility towards minority groups that’s going around the press at the moment.”

What’s your earliest memory of electronic music?

Mark Fell: The first sort of music that I really, really got into when I was a 13-year-old teenager was synth pop – the Human League, Soft Cell, Depeche Mode, things like that. So it was electronic music that was typically dance music. That’s what I initially got into and obviously being from Sheffield, at that time there was a lot going on – Cabaret Voltaire and The Human League were from there and a lot of other stuff besides that was going on in the city.

After that I managed to borrow a synthesizer off my parents’ next door neighbor. He was a technician at the university in the electronics department and he had a mono synth. So I borrowed it and that was it. I was like, “okay, this is amazing.” I was just really into synthetic sounds and synthesizers from then onwards.

House music, as a theme, has run through your music since the start. You were there when all of this kicked off, right?

Fell: Around 1985/’86 is when I first became aware of house music, then slightly after that, techno arrived. And it was like, “wow, this is amazing, this fulfills everything that I’ve been wanting for the past five years.” So about ’87/’88, when house and techno started to sweep through everything I was really into that world. But then just gradually I just wasn’t really into what was going on in Britain. I think because I’d come from this post-Throbbing Gristle industrial music background, when techno music arrived in Europe, there were a lot of producers who were trying to make it sort of weird but I thought it was sort of rubbish. Because if you’ve been through stuff like Throbbing Gristle and Coil, the weirder end of techno around 1990 was not particularly weird. It was like, “oh, we’ll just add some effects to it or something.” It all seemed a bit lost to me.

Then, in 1990/’91 met this guy called Callum Wordsworth who lived in Sheffield, and he was the person that introduced me to New York house music. I became aware that it was a thing in its own right and it was different from techno. What I liked about New York house music was the rhythmic structures and the chord structures and how those two things fit together. And for me it was the most interesting kind of music I’d heard for a long, long time just because of the way it carefully put those elements together. Even today that’s the kind of music I really like listening to.

Would you say you gravitated towards New York house for the same reason you liked The Human League in that they both have a pop sensibility?

Fell: Yeah, It’s got that clarity of production and craft to it. I guess there’s what you might call a commercial style of production behind it and yeah, I think it was that I probably liked.

Do you remember what the first synth you borrowed was?

Fell: It was a Powertran Transcendent 2000. I think it’s one that Joy Division had actually. It was a kit that you could buy and spend God knows how many months making it and that’s what you got. So I had that and then I had a drum machine, a Boss DR-55. My basic setup was a drum machine and a mono synth. And I’d just connect them together and see what they could do.

In just having these two things, would you agree that not much has changed for you throughout your career, because it’s always been based around two key elements – drums and chords?

Fell: Yeah. I have to stress that that kind of setup was the thing that I had around ’84/’85-ish and what I’d do throughout the ’80s was save money up, sell something, buy something a little bit different or better, save some more money, sell that. So I’ve progressed through just one machine at a time. But what you say is correct, that essentially the kind of structure of having something that creates pitched sequences and a drum machine is essentially the format that went into SND and probably most of my electronic music from that point as well.

Did you never think “I’ll buy another synth to do bass or melodies” or something to add a third element to the music?

Fell: I could never really afford it. I had no money basically. I didn’t work – I was either a student or unemployed for a long, long time. Although it became a horrible thing, I was very happy to be unemployed because it meant I could just do what I wanted to do and get an amount of money to live on each week. There were no jobs anyway because there was 100 people going for five jobs. I thought I might as well get out of the race.

But that just became my way of working. I think at some point I tried doing making music on more than two units but it just didn’t seem to work. Like having too many ingredients in a recipe. So I stuck to two machines. Probably 90% of the first SND stuff was all done on one sampler and the thought of adding to that equation just always seemed wrong. Even now I don’t have much equipment, but I have things that tend to be in storage and then I’ll just get them out to use for one thing and then put them back in storage.

I can’t understand why anyone would have a studio full of lots and lots of stuff. I see photos of other people’s studios and there are all these racks of equipment and stacks and keyboards and things and it just doesn’t compute. I was a kid in my bedroom for three years, with one piece of equipment, and I just explored in micro detail what that equipment could do, so I understood every little quirk of it.

For example, I had a Yamaha TX81Z, which uses four-operator FM synthesis and had a few quirks to it. There was this mode, which rather than being multitimbral was like a multitimbral setup, but every time you press the note to progress to the next sound in the series of sounds that you specify. So every time you played a note it could produce not just a series of notes but a series of sound changes as well. So I just did loads of work with that, setting up simple sequences of sound changes and note changes that would go in and out of phase and things. That became one of the kinds of techniques that I still use today.

Can you elaborate on how you use that kind of technique now?

Fell: I tend to be quite limited in the processes or technologies that I will use during the production of a record. So I won’t think “I can add this and I can add that and maybe I can put this over the top.” For me it’s always about one or two elements that you keep re-explaining in detail, rather than adding another layer on top. It’s about I look at the process and the technology and trying to just work with that, not confusing things by bringing too many ingredients to the production.

What is it specifically about FM synthesis that you like?

Fell: It’s a really flexible way of producing sound. One of the reasons why I was attracted to FM synthesis and why it’s still the thing that I use most today is the kind of tools it gives you to move harmonics or inharmonics around. It’s just a really interesting way of making sound. I got really bored of just like a low pass filter going “mwouw” and I wanted something else. FM synthesis gave me that.

FM synths from that era are notoriously difficult to program. Were you making your own patches, or modifying existing presets back then?

Fell: I was designing my own sounds, but often I’d start with the preset and then modify it. There was one cymbal sound on the TX81Z – a metallic, percussion thing with a nice decay on – that I remember just doing endless variations of. So often I’d start with a preset and then modify it or start from scratch. But that doesn’t mean I have anything against music productions that just use presets. One of my favorite producers from that time is Marc Kinchen, who did some really amazing work that sounds like it’s done with FM synthesis presets like Jazz Org and made great things out of them. There’s always been this negativity around using presets but I have no problem with them.

Isn’t your Sensate Focus material an exploration of what you can do with presets, or the ‘recognisable’ sounds that repeat themselves throughout club music history to a certain extent?

Fell: Sensate Focus was really about my difficulty with timeline editing environments. Most of what I do is not done in DAWs like Logic or Pro Tools where you have time going along an axis and you draw notes in. I can’t actually work with those environments. I think it’s some kind of neurological disorder that I’ve got, a bit like dyslexia where I just cannot function in that environment. I find it painful and horrible. So the Sensate Focus stuff was me deciding to do something in that environment that is the most structurally convoluted stuff you could possibly do.

For example, the Sensate Focus material doesn’t adhere to a 4/4 grid. So, just to loop things and work out where things start and end is an actual nightmare. I was doing that in order to think about the narrative structures that are present in house music, but also the sounds as well. It’s not just about sound though, it’s equally about doing a musical analysis of those structures, without trying to sound too sort of grandiose. Ultimately it’s just meant to sound good.

How exactly were you working in the DAW?

Fell: All those records are done completely in a timeline, just with the pencil tool drawing individual notes in and then cutting and pasting notes. So there are no Max/MSP-based automated processes, there’s no playing in real time. It’s all just a pencil tool drawing a note in. Like I said, each loop wasn’t a 4/4 structure that was divided into 16 equal bits. So there are all these unusual loop ends and timing divisions within the loops, which made it quite torturous to edit, but I liked the idea of what would come out of the music. What would the music end up being like if the process is as unpleasant as possible and as cognitively difficult as possible? Some of them are not as successful as others, but there’s one or two moments in that series that I think worked really well.

You mentioned Max/MSP, which has been your alternative to using a DAW for many years. How did you start making music with it and experimenting with mathematical algorithms for composition?

Fell: I think even my early experiments – messing around with the mono synth and the drum machine – were primitive kinds of systems that were sort of algorithmic. So the drum machine would send triggers to the mono synth in such a way that it would create evolving patterns. For example, imagine a loop with three events in, but the mono synth is playing a sequence of five events. I didn’t know it at the time, but from the very beginning I was working in that sort of way, trying out different ideas about constructing musical patterns.

I got a copy of Cubase when it first came out in 1987 with an Atari ST and that’s when I realized I had a problem with that kind of environment, which persevered for four or five years because I couldn’t find an alternative. I tried to implement algorithmic composition systems within Cubase by recording sequences and playing them back in ways that it wasn’t really intended to be used for, but the big change came in 1995 when Native Instruments released Generator, the precursor to Reaktor. I was working at a university at the time as a technician in the sound studios and computers were getting powerful enough to do real time DSP, so Generator was the first time that I was able to use a kind of computer-based real time DSP tool to explore things in more detail.

Maybe a year or two later, Max introduced MSP, and that was really the beginning for me. But at the same time, this was at the very beginning of working with Mat as SND, so I was still working on that project, which was still entirely based around Cubase and Logic. So the first couple of SND albums were all done in Cubase or Logic and constructed quite carefully, not using any kind of algorithmic processes at all. But because of those two albums, we start to get asked to do a lot of live work and that was the time that I thought “right, I’m really going to look at how to use Max/MSP in a live context.” And, and that was when I really got into it. The third album by SND, which was Tenderlove on Mille Plateaux, pretty much all of it was algorithmic.

How do your algorithmic composition processes work? What changes are they triggering?

Fell: The earliest stuff I did, I didn’t want to have to type different notes in, I just wanted to say “here’s a drum pattern and I want to have a few parameters that will change the structure of the pattern in real time”. That’s what my early intention was. So imagine that there’s a parameter called “relaxed” on one side and on the other side of the parameter it’s “intense” or whatever. I wanted to have that kind of control, or one that makes the pattern less or more dense. Or even just simple things like scrolling different layers in different directions as the track played.

I don’t know if you’d call it algorithmic, but it’s definitely a systematic way of working with musical structures that isn’t just about playing them. So there wasn’t ever a one to one relationship – here’s my finger, hitting a button and here’s the sound happening. There was always a process in the middle. The early algorithmic SND stuff was extended drum machine systems with a lot of real time control, while the stuff that I was doing outside SND was processes that generate patterns that are used to trigger sounds and modify the sounds.

But the basic reason for doing it was that it was more fun than just using Logic or Cubase and the stuff I was making just sounded loads better. So that’s why I was drawn to it. Y’know, I actually started making music because I thought it could be fun (laughs). I didn’t realize that actually a lot of it is really brutal hard work.

How does your new album, Intra, fit into the picture? Sonically, it sounds completely different from anything you’ve done before.

Fell: It’s actually very similar to a lot of my earlier work in terms of how the music is structured and produced. It’s a set of systems built in Max/MSP and the percussionists follow those systems structurally. The music is very similar to my earlier stuff in terms of its patterns and how the music evolves over the duration of the piece. The bit that’s different is that it’s performed by human beings and it’s an acoustic sound source.

About five years ago I got more and more interested in working with non-electronic music. The way I think about acoustic instruments is just as if they were synthesizers – they are just things that make sound. You can work with the digitally implemented model of a string or you can work with a string. And to be honest, I think the string sounds better than the digitally implemented model. The only problem with the real world is there’s not enough treble I think (laughs). If there was a god, he should have just added more treble on his mixing desk in heaven.

I saw Beatrice Dillon play a few weeks ago and what I was most struck by was how much her music focuses on the higher end of the spectrum.

Fell: I agree, I’ve noticed that about Beatrice as well. I think she’s a great producer. I saw her set at MUTEK last year and the clarity of the sound was really well done. It’s so easy just to produce these kind of smushed, mushy kinds of textures of noise and reverberation, so it’s nice when someone comes along and actually has an interest in the clarity and texture of sound. There’s so much music that is really badly produced and I don’t really know why it is. I feel like stopping people and saying “have you actually listened to what you’re doing?”

Are you actively using certain techniques to enhance the treble in your music or do you leave the tones as they are?

Fell: On FM synthesis models for example, if I want more activity on the top end, then I just put that into the sound itself – the algorithm will produce that. I also do boost the top end a little bit, but not much. I think I just really like clarity in sound and the kind of fizziness of the top end for me is really nice. But you do need good speakers to produce that in a club. If the sound system is bad, it will sound really bad irrespective of what you put through it, so when you play on something good, it makes such a massive difference. All I really want out of a performance is a good sound system.

Mark Fell is appearing at this year’s No Bounds Festival in Sheffield, UK, which takes place from October 12-14. For more information and tickets head to the No Bounds site.

Scott Wilson is FACT’s tech editor. Find him on Twitter

Read next: Sarah Davachi on the beauty of instruments, from analog synthesizers to pipe organs



Share Tweet