brettworks

You sort of know when a track is done, even though it’s never really finished. The music holds together, its sounds are cohesive, and most importantly, it has an aura that you enjoy coming back to. The track may or may not be your best, yet it does something in a unique way. There are, though, questions that persist, no matter how much work you’ve done. Are those the right sounds? Is there something microscopically annoying over here? Does that part belong? Even when there aren’t obvious problems, your ears seek options not pursued. I could just remix this and start afresh, you think. Should I take the beat out? Is this just step one of something new?

It’s these kinds of musings I had while finishing up a recent project of six pieces for keyboard, 8000 Studies. I was perusing my archive, wondering what was ready to release. Lots of stuff was sort of ready, but the prospect of finessing the music to completion wasn’t appealing, because tracks with multiple parts require a lot of time to sift through. Then I noticed a folder for a project from seven months ago whose tracks were created on the same instrument. This compositional constraint helps me find ways around an unfamiliar piece of software and inspires me to write a series of pieces around a single sound: If I do one piece, how about two or three? And having several of a similar kind of track collected in a folder is easier for future me to spot. That’s how I decided on my latest release, which comes out this month.

The first step of revision is to re-open a file and have a listen. The idea here is to respond to whatever can be improved, heightened, amplified, refined, toned down, and so on. The important thing is to patiently listen again, this time around with more seasoned ears, and notice what’s noticeable. This always feels like a daunting task, but once I’ve begun ideas always come to mind. A good proportion of my editing fixes are simply volume-related. As that old musician’s saying goes (I’m not sure if it’s a saying, but I think about it often): make it more dynamic. What I notice when re-listening to a mostly finished track from half a year ago is how the balance of parts and effects is off. It’s simple to fix this, and simple to wreck the track in the process, since I’m not sure I understand how my music—or any music—achieves its ends. With this in mind, I tinker away, following my ear, making changes that make sense today.

After volume, EQ is the second most important editing tool. The general idea is to use EQ to sculpt the frequency spectrum so that each sound has its own frequency “slot” and thus doesn’t mask its neighboring sounds. For example, kick drum and bass sounds require finessing so they don’t cancel each other out. But sometimes frequencies tell a story that perhaps shouldn’t be EQ’d into predictability, so be careful. I use EQ mostly as a contrast tool: boosting what sounds dull and cutting what sounds harsh. As you learn to listen with an ear to what’s going on frequency-wise in music, you get better at knowing when sounds are obfuscated for the wrong reasons and how to clarify them with EQ. Producer Sibel Koçer (aka JakoJako) explains what one notices: “As soon as you turn the track on you hear what’s wrong: which frequencies are annoying, where it’s clashing, what is trying to fight against some space.”

After adjusting volumes and EQ’ing I’ll put the entire mix through a mastering chain—usually an EQ, a multiband compressor, a second compressor, and a limiter. Each of these devices is doing almost nothing; I use them mostly to make broad tweaks in brightness or darkness, shape the stereo width, catch unruly sounds (peaks), and bring the audio up to a reasonably robust, streaming-friendly volume without squashing its dynamic range. Less is more here: the best mastering is undetectable.

I follow this volume-EQ-mastering workflow to nudge a track towards done, but that doesn’t mean the workflow doesn’t occasionally lead me astray. For example, I spent a while editing the first track for 8000 Studies. I was convinced it needed a smidge of resonance in the form of a plate reverb, so I added it. Then I got further in the weeds of revision, weaving the reverb in and around the melody lines and making this new resonance part just so. I thought I had made a simple fix, but when I listened to the revised track against the original it immediately sounded wrong: the aura was gone. The music didn’t sound like a single gesture anymore; it sounded like an artifact had been layered onto it. The original version had its issues, but made up for them with its vibe. So I knew what I had to do: trash the revision and keep the original. Even though I could have tried doing more things to make it sound more finished, this track felt right and so was done.