Ever since Real Tone Records released What They Say in 2010, Maya Jane Coles has been in the spotlight. She was just 22 years old then and had already been making music for years. In the decade since, she's amassed a staggering body of work that includes original material under her own name and as Nocturnal Sunshine, much of which she releases herself via her label, I/AM/ME. Since Maya Jane Coles released her first record in 2008, she has put out four solo albums, over 30 singles and EPs and more than 100 remixes for artists including Little Dragon and The Orb. Then there's her work as a producer for indie artists, her time in the duo She Is Danger and the Nocturnal Sunshine LP she released this fall. And she's done it all with little formal education in music or audio engineering.

Like many aspiring producers, she had to figure it all out herself: how to make certain sounds, how to keep arrangements simple yet interesting, how to structure a track, how to apply musical concepts like scales and keys, and how to make it all sound pro. She's managed to build an impressive catalog using accessible methods and gear. Her home studio isn't full of pricey hardware. She works mostly with a computer.

Coles' success is a testament to a hard-learned lesson that runs counter to a prevailing logic in underground dance music: you don't need clever production methods, knowledge of music theory or nice hardware to make great music. In fact, if you spend too much time watching YouTube tutorials and collecting vintage hardware, an inspired 22-year-old with a cracked DAW and an old sampler might zoom right past you.

When did you start producing and what made you want to do it?

I played instruments growing up—cello, guitar, drums, keys, bass, saxophone—so I did have a bit of a musical background. But I started producing through hip-hop. When I was 14 or 15 I would record guitars and go through my dad's old records to chop up snares and hi-hats and record them onto a cassette or a minidisc player. We had a Mac with Adobe After Effects because my dad was a graphic designer. You could edit audio with that software, so when I started messing around on the home computer I discovered I could chop up sounds and loops.

Then I realised that you don't have to use the audio functions of video-editing software—there's dedicated audio software. I managed to get a cracked version of Cubase, and a year or two later I found Logic was a lot better so I switched to that.

How did you learn to use those DAWs?

At first I just taught myself, really. I was inspired by hip-hop producers that were all very sample-based, so at first I was sampling other people's stuff. Eventually I decided to try and make an original and sample myself. A lot of my current process comes from recording hooks and melodies—I'll resample the audio and then process it so that it sounds like I've sampled some other track, but it's just my own stuff.

Are you sampling elements of tracks that you made before or recording instruments you're playing live in the studio?

Both. I've made music now for 16 or 17 years, so I've got libraries and archives of old tracks that have never seen the light of day. A lot of them are just sketches and ideas. If I'm stuck for ideas when I'm starting a new track, I'll find a loop that I used in a track that I never released.

I feel like a lot of my stuff is quite hook-based. It'll have a main riff, melody or chord progression. The bassline could also be the hook. Occasionally the drums are the key element, but that's not really how I tend to work. It also depends, because I work on so many styles that everything comes together differently for me. But the compositional side often is the first thing that comes out of me. I build around that.

Melodies and riffs sound simple, but there are a lot of music concepts at work behind them. When you were learning to play instruments, did you also pick up basic music theory like chords, scales and staying in key?

When I was in primary school, maybe around eight years old, my mum sent me to theory classes. I never wanted to be there and I didn't really pay any attention, but because I was so young I picked up a lot anyway and it became ingrained in me. I couldn't tell you anything about it, but I do feel it played a big role in me understanding the structure of chords, sounds and melodies and being able to compose music in that way.

The technical stuff is all very secondary for me—or at least, it was. I find in the DJ world, a lot of producers come from the technical side first without necessarily having a musical background, because it's not like you particularly need it when it comes to DJing. There's a difference between coming at composition through trial-and-error and making deliberate choices. A lot of people might be able to tell if two loops don't go together or if a melody doesn't fit over a bassline, but they wouldn't necessarily be able to create or think of something that does without trying it out. It's more like, "Oh, this doesn't work. That doesn't work. Does this?"

Melodies just come to me. That's the easiest part of the process. Music is very emotive for me, so usually the first thing I'm trying to create is an element that hits you in that spot and makes you feel something. Sometimes I'll sit and play guitar or bass until I find a three-note melody or something, and then resample it in Kontakt or Battery. My main tool is the built-in sampler on Logic, which is the EXS24. Over the years I've created hundreds of libraries and presets of sounds that I've recorded or programmed and bounced. Even if I use a soft synth I'll often end up exporting the audio and then processing it again, because I work best with audio.

Tell me about the libraries and presets you've saved.

Most of the libraries are homemade percussion sounds or samples of claps and hums, vocal stabs and kits. It's taken me years to organise it. Before, I was so focused on making the music that I'd never label anything or categorise sounds or organise my libraries. In the last few years, I've started to sort things according to date, and that's nice because you can see exactly how productive you were during different periods. It makes a hell of a lot of difference. I’ve lost hours trying to find sounds that I'd worked on before.

How do you sort the sounds into different kits? Do you have kits sorted according to the type of instrument or the style of music it might go with?

No matter what genre I'm making, I tend to use a lot of the same formulas for constructing music. My core sounds tend to be pretty similar—I use very similar kinds of kick drums, filters, compressors and effects throughout everything I do. The core sounds are the same but I effect them in different ways for different styles. I feel like that's one of the key things that keeps my character within anything I do. Whether I made a hip-hop track or a more acoustic, folky one and then some house or techno, you'd still hear a common thread because my approach is really similar. Maybe I'll use the same sub that I built in the EXS24 throughout tracks that otherwise sound completely different.

I'm going for quite a specific sound with drums, no matter what genre I'm making. A lot of the time it's very filtered and low-end-heavy. When I first started making music, I had no idea about mixdowns and stuff, so I would just go at something until it sounded like how I wanted it to. I'd be doing everything "wrong." But it sounded how I wanted, so I didn't care.

Is that the case for records you ended up releasing?

Yeah. I hate listening to my very early releases because my mixdowns sound all over the place to me now. My first release was like, 12 years ago. I think I was 19. At that age, I wasn't in this mindset of, "I'm going to wait till I'm perfect at what I'm doing to release it." I was like, "I've got the opportunity to release something. Fuck it, I'll put it out." When you're young, you don't think about the fact that that's going to stay there forever. Still, I wouldn't ever change that now.

People would often tell me, "Most people don't mix down their tracks. You can get an engineer to do that too if you're not happy with it." And I'd be like, "But then how am I going to learn?" The only reason I mix down stuff way better now is because I've had ten or 15 years of listening back to stuff and thinking about what I'd do differently. I quite like the fact that I kind of didn't care about things sounding perfect back then.

I've been struggling to find a balance between learning and overlearning the "right" and "wrong" ways to do things. Sometimes I'll give a track to a friend and say, "Why can't I get the bass to be more present in this mixdown?" and they'll make it sound better by breaking all the rules I was trying to respect. I'll sit there nervously and say, "Uhh, wait, I think you're redlining," and they'll just say, "Pff, so what?" and keep going with what they're doing until it sounds the way I want.

I get that. I was self-taught, and I had dial-up internet on my home computer. It wasn't like you could just go online and there would be hundreds or thousands of production tutorials like you see now. But growing up in London, I was lucky to have a lot of the right facilities around me. I grew up within Camden Council, and they had loads of charity-based studios where kids can go during summer holidays or once a week after school. There were free courses that taught beginners how to navigate the basics of Ableton and Logic. This is how you create a new project, here's how to hook up a channel to record, these are some basic EQing principles and so on. I took those things and taught myself the rest.

I didn't learn anything about proper mixdowns, though. My ears eventually tuned into things that I hadn't picked up on before—I remember listening to an old track I made and thinking, "Wow, that snare sounds super harsh. Why didn't I hear that?"

I didn't start trying to learn things in a more traditional way until I started producing for other people, recording vocals for other artists and so on. I felt like I should be following the rules more because I was no longer just responsible for my own output. I can do whatever I want to make my own track sound how I want. But when I produce for someone else I feel like I shouldn't fuck up their song. I felt an urge to be more "professional."

I started thinking about it like, "OK, there are a bunch of rules I could be following—not necessarily have to be following. But it would technically make my productions better if I do follow these rules." Then I started to think too much about the rules and how I shouldn't have certain frequencies clashing and then making a creative decision—like "I should take this element out"—based on that.

I definitely felt that it slightly changed the way I approached making music. It removed a lot of the creativity and fun for me. So then I swerved back, and now I think about it way less. Still, having the knowledge makes a world of difference, of course. I'll follow the rules and realize that I actually don't really like how it sounds because I did something that I thought I wasn't supposed to do. So then I'll put it back how it was or I stop thinking about it.

It can take listening out of the process, and instead you end up following rules or mixing by looking at the frequency spectrum analyser rather than letting your ear guide you. I have a hunch that learning to mix down is learning to listen more than learning to apply specific concepts or use certain tools. What do you think?

That's why I find it practically impossible to work on one track at a time, especially when mixing down. Your ears get used to sounds. I like to work on ten things at the same time so I can flick back and forth between them. I'll fine-tune something and then switch to the next thing for 30 minutes and then spend 30 minutes on another. Taking 30 minutes away from a track and focusing on something else gives my ears enough of a break to then be able to go back to the first thing and spot everything that I want to change instantly. Otherwise I could have been going around in circles for hours or made something worse because I'm over-listening.

Is the mixdown something you do as you go along, or do you try to get the idea out and then revise it? How separate is that process from EQing?

I prefer to try and mix down as I go along. That makes the final mixdown process a lot easier. It depends on the scenario—obviously, if I'm working on the go, that's the smallest priority because I haven't got monitors or silence, and you can't hear certain frequencies on headphones. So if I'm travelling, I tend to work more on ideas: making sounds, chopping up sounds, building libraries. It's like building up ammo for my tracks so that when I come back to the studio I have all this stuff that I can use. I've spent hours in a hotel or on a flight messing around with existing ideas that I've got, and then I bring it back to the studio and I'm like, "Oh man, I've fucked it all up."

What are some common challenges you have with your mixdowns? Do you have an overall goal?

My first focus is making sure there isn't any sound in the track that sticks out like a sore thumb and then to try and get everything sounding as full and rich as possible. There has to be depth in the track.

I want it to sound the way I imagine it in my head while still being suitable to send to a mastering engineer without redlining. I tend to way overdo the lows, especially when I'm making club music or something that revolves around the bass. I try to balance everything out. Even when I'm making something that isn't dance music, there's that clubber in me that always tends to want to make everything revolve around the bass and the drums. I'll make pop music with a bassy, club-rooted formula. So I do approach the mixdowns in a very similar way no matter what style I'm working on.

What's your process with recording, EQing and processing vocals?

I just use a Rode NT100 that I've had for years. It's not a super expensive mic but it definitely does the job. I approach things differently if I'm working with someone else's vocals as opposed to my own. Usually if I'm working with another artist, it's a full song and they're top-lining one of my tracks. I wouldn't be used to EQing their voice if I hadn't worked with them before, so that's a whole new thing I figure out as I do it.

If it's my own voice, I always cut a lot of lows, time stretch and filter. A lot of the time, if I'm sampling my voice, I don't want it to sound like I'm singing. I want my voice to sound like another instrument, so I treat it in the same way as I'd treat a guitar. I first started using my own voice as an alternative to sampling other old records or other people's music. If I wanted a one-note hum or stab or something, I would just record it myself. It wasn't my intention at first, but I feel like now it gives my music this signature sound that no one else can replicate. If you pick the right sections and effect and process them in the right way, it can turn something from being a kind of monotonous loop into something unique.

How do you arrange a track?

I just go with my intuition. When there are enough elements to start arranging the track I just start expanding and lay it all out. Builds and breakdowns come in where they feel natural.

Do you have a different workflow when you're making Nocturnal Sunshine material versus stuff you'll put out as Maya Jane Coles?

I don't ever actually sit in the studio thinking, "This is a Nocturnal Sunshine track and this an MJC track" before I start doing anything. I started releasing under a different alias because I was making so much music and wanted to put out more of it than I could or should within a certain space of time under one name. It made sense to divide up what sounded different into another project and release under that name so that I could release an album and then a completely different EP or mini-album a month later. It's more of a marketing thing—I was often told I couldn’t release all the music I wanted to release at once because there was just too much of it and too many different styles. That kind of conversation happens when you're working with bigger labels, or multiple small ones, because they don't want the release you're doing with one to clash with another.

It just made sense to use an alias and have a different outlet where I can think about releasing stuff that has nothing to do with a current album release. I just kind of sit down without focusing on too much of an aim and see what naturally comes out. That later gets divided into my different projects.

Yet there's a cohesive and distinct sound to the Nocturnal Sunshine stuff.

I think it ends up sounding similar because I'll continuously be working on track after track, but I'll never finish it until I know what I'm going to do with it. I'll rarely even fully arrange a track until I know I need to mix it down to release or put on an album. I'll have rough skeletons of tracks that could potentially go either way, and once I have a bunch that could end up being a Nocturnal Sunshine album, then I'll decide which of the others could go on a Maya Jane Coles record. Once that structure's there, I'll finish the tracks.

It's also to do with how I use the same sounds or settings across different tracks. I tend to finish all the tracks on an album or EP together. Say I have a guitar note that I recorded and then filtered, then I put it in the EXS24 and cut the attack to make it into a pad sound. On another track I'll bring out the release and put effects on it so that it's no longer a guitar note but more of a synth sound. If I have the skeletons of three tracks that I think will go on an EP together, I might add similar little sounds to all of them or apply a similar distortion or filter to them. I'll save my effects channel strip settings from certain key sounds and use them on other tracks that end up on the same album so there's a certain familiarity that carries through.

Are there certain effects or plug-ins you prefer?

I use a couple of pedals for effects and delays with a Loop Station. I use a lot of presets in a guitar effects unit, which I'll then tweak. A lot of the time the units have about eight different effects—multiple delays and distortions, EQ, chorus and so on. I'll eliminate four and add four of my own things, then I'll save the preset and go back to using it all the time.

Do you use any other outboard gear?

I have a couple of old toys that I've had from when I was, like, 15. On some of my early productions I used the Proteus 2000, made by E-mu. It's a sample-based module. You can build endless presets but it has a core base of samples. You can build practically any sound you want from these core samples and save the sounds as your own and send signal MIDI out with audio.

It's never been considered a "serious" piece of equipment, even in its time. But the sounds, when it comes to basslines and stuff—it just has this slightly saturated sound that you can't really recreate. But I don't use a lot of outboard gear. I've built my sound without gear, and part of me is like, "Why distract myself?"

While I recognise the power and possibility of computers, sometimes I wonder if I will ever really be able to recreate specific sonic characteristics of '90s dance music without using '90s gear. Are there things you can't fully achieve just working in the box?

I think a lot of people use hardware to replicate or create a vintage sound. And sure, there are certain things that you're not going to necessarily be able to instantly create on a DAW alone. When I want a sound like that, I'm not replicating what analogue hardware can do. I have my own processes to make it sound in a sort of vintage way, like resampling a million times or using bad recordings or something recorded on a really shitty mic. Layering sounds can really create that vibe. Maybe I'll record a storm outside, make that into a percussive sound and layer that with a synthesized drum machine sample.

I often layer multiple different sounds that cover four frequency ranges, especially with bass. There'll be a very filtered low, low sub below 100 Hz that sounds very muted, then the more solid, mono mid-range sound. If I'm creating a wobble bassline, then I'll make the cutoff frequency of the filter vary to create some movement. On top of that I'll have a more high-frequency, low-cut bass sound, and I'll filter that in and out throughout the whole track. I'll record the automation from start to finish so that there's never a static part and it's always changing a little bit.

That's especially important when the track revolves around a bassline, as with a lot of Nocturnal Sunshine stuff. I'll try and make sure that there's never eight bars where the bass is all flat and the movement is the same. Say automation is controlling the filter of one of the layered parts of the bass. Even if all the other three layered parts stay the same, it creates movement within that sound. A listener wouldn't necessarily know that that sound's created with three or four layers. It just gives it a slightly fuller sound with the movement.

How did you figure out what types of variation should happen and how often they should happen?

I feel like you could be making music for 40 years and still figure these things out bit by bit. One of the rules I like to give myself is: I should try and have enough movement within eight bars to keep the part interesting. Say I have a skeleton of a track that I'm ready to arrange. Before I take anything out or add any additional element, the core structure should not bore me from start to finish. I should be able to listen to just those core elements for four minutes without getting bored. If you can manage that, then the track's going to be good no matter how you arrange it.

How many core elements are typically in a skeleton?

Do you mean how many individual channels are there? Or how many discrete elements that you hear as a separate entity? Sometimes the drums can take up 20 channels.

Right. Let's say the drums are all together one element, the bass is all one element…

In that case, usually four different key elements are enough to get something good going. But within those elements there's a lot going on. When I first started making music I never thought that way. I'd listen to a track and hear drums, a bassline, a chord progression and some vocal stab or a sample. Then I'd go try to make a track that has the same kind of richness and simplicity, and I'd end up thinking, "Well, that track only has drums, a bassline and one chord sound—why does mine sound so much flatter?" You think there's just four elements but to create each one you're putting together like, 50 different elements.

I'd figure these things out but I still wasn't really thinking about clashing frequencies, and how if you're going to layer so many things you really need to get the EQing spot-on. Sometimes I'd get to a point where I'd be adding more stuff to mask something, which never fixes the problem. My arrangements would be a mess, and mixing it down would take forever. I feel like only now am I starting to perfect stuff for me personally. Everyone has their own techniques, and there's no real right or wrong way. Someone could tell you all their tips and secrets, but it might not help you.

Earlier you mentioned that you launched different projects and aliases because you were making so much music. What was driving or enabling your super-heavy output at certain periods in your life?

What I release only makes up about 10 percent of the music I make. My number one passion is making music. It's never a chore. It's one of the things I enjoy most in my life. It's never been that I'm trying to catch up on making more material. It's always the opposite, that I'm working on so much music that I end up needing to catch up with everything else in my life. One of the main struggles I had early on in my career was learning to balance my time between touring and production, especially when things were first getting really big for me. I didn't feel like I was in the position to turn down gigs when they were first being offered. I felt like I had to tour in every country and city that wanted me for the first or second time.

During that time, I saw a massive decline in my creative output. I was unhappy because I felt like the touring was holding me back massively from what I wanted to do. I had to experience that to step back and figure out a structure that worked for me, where I was able to do both things without it completely draining and exhausting me. Now I'm in a happy place and I have a good balance with friends, family, relationship, social life, work, touring, studio. Everything is kind of in the right place. It feels good.

Still, there's so much more that I want to do with production that I can't do yet, like film scores and production of music that's completely outside of the electronic world. I want to write pop music and produce for pop artists. Actually, I'm working on more hip-hop stuff at the moment. I've done a full circle and really gone back to my roots and what got me into making music.