Speed ills


Many EDM producers have heard about The ill Methodology that is taught and promoted by ill.Gates. Not the album of the same name, but the actual workflow methodology that ill.Gates teaches. Perhaps you’ve even attended his online “Breakthrough” or “Workflow” seminars, where he goes into more detail about specific aspects of his methodology. Perhaps you’ve even joined the Class of 808 or Weekly Download tiers of ill.Gate’s Producer Dojo, where his methodology is taught and practiced in detail.

In all these places, ill.Gates consistently promotes and explains the concept that “Through Quantity Comes Quality”, and then goes into great detail about workflow techniques that help you write songs fast. As in, literally 2 or 3 full songs in a single day. Or at the very least, a single, full song finished in one day, in one single “daytime session”.

Many EDM producers at this point (especially newer producers) might say “That’s impossible! I can spend weeks or months on a single song.” For that matter, even new Class of 808 members can often still puzzle about how, exactly, this is possible, until they ask the right questions of other 808 students and mentors. Or, alternatively, the occasional post will appear in producer-oriented groups stating that “so-and-so big name artist told me that I should take months to slowly improve a song, before releasing it”.

This post is meant to elaborate on ill.Gates’ assertion that song-writing speed is the key to producing high-quality releases. Not just the why but also the how. The clues are all there in his various workshops and Weekly Download episodes and recorded one-on-one lessons with his original Class of 808 students. But the “aha!” moment is never clearly described in a direct way, so that’s what I’m trying to do with this essay.

The best way to arrive at that “aha!” moment is to work backwards from the very end of the production process: your published releases.

The music industry has changed drastically in the past decade. Albums are effectively a dead, old-school practice. Yes, the big music publishers still cling to the old album release model, at a cadence of one new album every year or so for their stable of signed artists. And arguably, this model still works reasonably well for the really big name artists with millions of fans and huge promotional budgets from their major label.

But the EDM market is very different. It is highly democratized and independent, with tons of great artists who don’t even bother with trying to get signed to a major label, but who instead self-release and self-promote, either alone or through collectives such as Producer Dojo. Most of the major labels don’t even understand the EDM scene and market and how fans consume EDM music. There are indeed some niche labels in the EDM space who operate more like major labels, but still with important differences.

And the main difference, across the entire EDM market, is release cadence. In the EDM world, Single releases are where its at. Ideally at a cadence of one new single released literally every 3 or 4 weeks, at most. Occasionally, an artist will put out an EP, which might simply be remixes of a formerly-released single. Or maybe just a “rollup” of the last four singles. Or maybe an EP will indeed feature entirely new songs, but might comprise only TWO new songs. Albums? Almost never. Usually, an “album” put out by an EDM artist is just a large collection of all their best singles, like a “Greatest Hits” album in the old days. Or the niche labels in the EDM space will put out mostly the occasional “compilation” album showcasing one or two songs from the best artists on the label. And yes, no doubt, a prolific and well-established EDM artist will still put out the occasional album with a fair amount of entirely new material on it, but they can get away with that because they've already built a huge following and a lot of fan engagement. 

Why the focus on single releases at a super fast cadence of one new song dribbled out every 3 or 4 weeks? It all boils down to where and how EDM producers actually make money off their art. It’s not from streaming revenue (Spotify, Apple Music, YouTube, etc.). It’s not from download sales through iTunes/Amazon, or even Beatport. Instead, nearly all the revenue for EDM artists comes from performing as DJs at shows and festivals, and from merchandising, and from private teaching/mentoring. For 99% of EDM artists, there is effectively zero revenue from streaming platforms and download sales. Not an exaggeration. The market numbers simply aren’t there except for the very few biggest names in the genre. The numbers aren’t there in part because there is no real barrier to entry for new EDM artists, so the competition for ears and eyeballs and discovery by EDM fans is fierce. And the other part is that EDM songs are consumed in much the same way that social media is consumed. Maybe you stumble across a new song or artist in some playlist shared with you (or in an algorithmic “radio” playlist). The chances that you’ll actually stop and add that song to your library and to your own personal playlists? Pretty small. There’s just too much new EDM music coming out constantly and its easier to simply drink from the firehose. The chances that you’ll actually stop and go buy an EDM song you just heard for the first time as it scrolls across your playlist of the moment? Practically zero. The chances that you’ll actually look at your phone (or whatever) and take note of the artist name for that song that sounds cool as it rolls across your playlist? Still pretty small. Or if you do, you won’t recognize it and you won’t remember it later.

Except… When you keep repeatedly running across the same new artist name over a period of 3-6 months... When one of your friends or a reddit or an FB group repeatedly mentions some artist name they recently discovered… When it suddenly “clicks” that you’ve noticed a FEW cool songs roll across your playlists and you’ve seen the same unfamiliar artist name repeatedly attached to all of them… THAT is the moment you become engaged and interested in learning more about that artist. THAT is when you actually google them and stumble across their artist website, or their YouTube channel, or their artist page on Spotify and you press the follow button. THAT is when you start wondering whether they might be playing at a show or festival near you. THAT is when you start asking your local clubs and festival promoters "When you are bringing that new artist here to do a show?

See the pattern here? EDM artists attract potential fans by putting out a constant stream of new music every month. Not once per year. Not twice per year. Every 3-4 weeks. Like clockwork. This is how you become visible, and how potential fans come to trust that you will be a dependable stream of new entertainment that is worth becoming invested in. This is how you make fans irritated that they "missed" your last two great songs, and dammit, they better follow you on Spotify and give you their email address so they can be the first to hear your next song and tell their friends about it.

By contrast, If you work all year on one album release, you have a period of like two weeks out of that entire year when you might be able to catch some eyeballs and hearts and minds. You have only two weeks out of that entire year to be the subject of a conversation among EDM lovers. After that, you’re already old news and crowded out by all the new stuff that released right after you did. And you fade from everyone's awareness until that short two-week window next year. When they only think about you for two weeks out of an entire year, how are you going to motivate fans to ask their local clubs and festivals to bring you in so they can buy tickets to have you thrill them in person?

So repeat after me: Albums are dead. Even EPs are pretty much dead. It’s all about the next new single and how fast you can crank out the single releases and constantly pop up on peoples’ radar over and over.

By now, some producers might be panicking: “How the hell can I possibly put out a polished single every 3-4 weeks?” Trust me; it can be done. Easily. Handily. Even by a producer with "intermediate" skill. This is what ill.Gates teaches with his methodology. I watch at least 30-40 different artists in Class of 808 practice doing this every single month with our internal Cypher challenges, which result in a new Cypher mixtape from the Dojo pretty much once per month. These artists are doing it at the same time they’re ALSO working on songs for other Dojo releases. And at the same time as they’re just writing new songs for their release backlog!

So how does this work? Here’s the missing link that is easy to overlook when ill.Gates first presents his methodology concepts to you. The missing link, the “aha!” moment revolves around three simple ideas:

  1. You have to change your “definition of done” for “a song”.
  2. You have to change your definition of when “writing a song” starts and ends.
  3. You have to understand the concept—and value—of what it means to “fail fast”.

I’m going to start with point #3. It’s a concept familiar to software developers, and ill.Gates has recently starting trying to promote and explain that concept in his workshops. The simple truth is that out of every 10 songs you write, a few of them will be “good”. Maybe one of them will be “great”. And the other 6 or so will be “meh” or “weak”. And here’s the root of what’s wrong with the “perfectionist” approach to music production, where you ALLOW yourself to spend waaaaayyyyyy too much time trying to polish a turd. Where you spend months pulling out the same song over and over, trying to “rescue” it because it seems promising and has some cool things about it, but it’s just not quite good enough yet. Where you beat your head against the wall and struggle to make a song finally work and sound like something you’d be proud to release...

…If you're doing any of this, you’re doing it wrong. If you are really struggling with a song and need to walk away and try again later, you should simply put it out of its misery. And strip out the few really tasty bits it might have, to add to your library of “sparks” and “robot” sounds for reuse in some new song some other day.

The great songs practically write themselves. They just fall out of your head and everything just works. The “face” and the hooks and the earworms just magically appear. What you want are to stumble upon those great songs as fast as you can, as often as you can. Which means being willing to fail fast and simply recognize when a song isn’t good enough, and to move on to the next attempt.

Now, those of you who pay attention to another point ill.Gates teaches might be saying, “But he also teaches that you should finish every song, because it’s only by going through the entire process that you get better and faster at making all the required decisions.” Which is also very true. And this brings us to points #1 and #2: what, exactly, is a “song”, and when does the “song writing” part of production begin and end?

The key here is to realize that “writing a song” comprises only 2 phases out of a 6 phase process. The full 6 phase process can be thought of as “producing and releasing a finished song”. And ideally, you should be doing all 6 phases and “releasing a finished song” every 3 to 4 weeks. But you should also be doing those other 2 phases many times, “writing MANY songs”, within that same 3 to 4 weeks. You want an ever-growing stack or “backlog” of new songs that you’ve written. And every 3 to 4 weeks, you pull the BEST song off the top of that stack and you put it through the phases that come after the “writing a song” part. So the full production cadence is really split into two different production workflows that overlap with each other:

  • Writing many new songs in a given 3-4 week cycle, PLUS
  • Finishing the best song from your backlog, and releasing and promoting it, in a given 3-4 week cycle.

The 6 phases of “releasing a finished song” can be broken down as follows:

Phase 0 – Assembling your palettes (or “ingredients) for a song
Phase 1 – Writing your superloop and finding the “face” for the song
Phase 2 – Duplicating your superloop and then doing subtractive arrangement to arrive at full “journey” from intro through outro
Phase 3 – Flattening and detailing the arrangement
Phase 4 – Mixing
Phase 5 – sending it out for mastering (or self-mastering it)

Of these six phases, only phases 1 and 2 comprise “writing the song”. At the end of phase 2, you have a fully done song. The rhythmic, harmonic, and melodic elements are all there. The “face” of the song is there. The “emotional journey” from intro to outro is there. The quick and dirty leveling to make it all fit together reasonably well is there. You can play it back for yourself or others and perceive it as “95% there, just rough around the edges and needing some polish”.

Those phases 1 and 2 can be done really fast. Like, easily in 2-4 hours or less once you learn some basic workflow techniques and build a little self-discipline so you don’t spin out into unnecessary side-trips and distractions. And if you practice this song writing process with timers and linear recipes, you can get the process down to more like 1-2 hours per song. Seriously.

There are two key points here:

First, you don’t bother putting any song through the longer, more complex process of flattening, detailing, mixing, and mastering (phases 3-5) unless the song is your current “best of best”. This is the “fail fast” part. You simply don’t bother with all these steps unless the song is GREAT. Unless its sitting at the very top of your backlog. All those other songs that never rise to the top of your backlog? They’re the failures—even if some of them are actually pretty good too! Maybe some day one of them will rise to the top of your backlog, but until they do, they are FAIL, lol. And the ones down in the bottom half of the backlog are useful only as scrap to be broken apart and raided for a few interesting sounds that you might be able to use in some future song.

Second, that phase 0 part of “assembling your palettes (ingredients)” is REUSABLE across many phase 1 and 2 songwriting attempts. Sure, at first you’re spending a LOT of time creating those palettes and making them nice and modular and reusable over and over, but once a palette is built and refined across several song writing sessions, it’s FOREVER. Next time you need sounds and textures from that palette? It’s ready to simply grab and go! Over time, you’ll have a library of ready-to-use palettes to reach for over and over again as you write each new song.

Now, I don’t want to digress too much into the concept of “palettes”. I know it will be an alien—possibly meaningless—concept unless you’ve watched some of ill.Gates’ more recent workshops. Broadly speaking, though, it’s a concept of modularizing your sound design, and sound types, into different chunks that you reach for and they’re ready to go. The idea is to prevent you from doing sound design while you are in the middle of your song writing. During phase 1 and 2, you should be making decisions only about sound selection and sound arrangement. That’s it. You should NOT be stopping to spend an hour or more sound designing the perfect sub bass or perfect mid bass noise for your drop. You’re going to write a dubstep song? Then you are going to need a handful of ready-to-use palettes for each major type of sound that appears in a dubstep song. For example, a palette of aggressive dubstep kicks and snares—LOTS of them—that you can literally twiddle a knob and quickly pick the ones that sit best in your mix as you build up the superloop and finalize the arrangement. An instrument rack of sub bass sounds, with processing effects already in place, all of which provide a fast, punchy sub bass specifically for dubstep lines, and which has a few macro knobs you can twiddle to dial in a sound. Maybe your idea for a particular song writing session is to create a dubstep song with a “sci fi” feel, so you have 128s full of science-fiction-y one shot samples ready to grab and go. Twiddle some knobs to pick the one shots that sound best as you superloop and arrange the song. Similarly, you might have a folder of Serum user presets that have some especially “science fiction” sounding presets. So instead of auditioning literally thousands of Serum presets while writing that song, you quickly audition 10 of them. At most. And so on. This is the general idea behind “palettes”. And as you can see, the next time you want to write a song with a science fiction feel, you’ll probably reach for some of those same palettes again.

And finally, for the benefit of those who have never seen one of ill.Gates’ workshops, another key concept here is to do all of your sound design and library management and palette-building in a completely different session from doing song writing. You never mix the two. He calls these broadly “daytime sessions and nighttime sessions”. If you have 2 to 8 hours to sit down and really focus on just songwriting, that’s a “daytime” session and you only do phase 1 and 2 work. You write a song. Nothing else. No sidetracking into preset surfing or sound design or checking your social feeds. You just grab your spark idea, your palettes that you think will fit that spark idea, and you superloop until the face is clear and all the important parts of a song are there (tension, payoff, and climax), and then you quickly lay out a subtractive arrangement to tell the story and provide an emotional journey.

If you have time left over in the session after “writing a song”, maybe you work on some phase 3, 4, and 5 work for that “best of best” song that you are currently preparing for release. Or maybe you write another song! Or you do BOTH. Phase 3, 4, and 5 work does not require long periods of creative focus. It can be done in short piecemeal stages where you listen, make a list of everything you hear that needs to be fixed/added/changed, and then set timers and work through that list. 20 minutes here, 40 minutes there. Work down your list one at a time and stop where you need to. All this phase 3, 4, and 5 work can be done piecemeal in bits and pieces across many days, because it’s technical, iterative work, not creative work.

Okay, so if you’ve stuck with me for this long, hopefully you have a better understanding of why “Speed ills”. And why being a perfectionist and having the mindset that you can and should sit on a song for months and slowly chip away at it is not a very strong approach. Ask yourself what happens when you try and try and spend months on a song and finally release it and… It’s just not that good. When in the same time, you could have cranked out literally dozens of songs and picked 4 really great “happy accidents” and put out a constant stream of single releases to keep your fans engaged and constantly watching for your next song to come out?   

Heather Greywalker