“So you’ve got this person called Dave (from New Jersey) … He doesn’t know how to mix a song!” - Abdi Arif
Recently, I’ve had an abundance of time to make music. But even in the before-time when things were normal, I’d struggle to finish most of the things I started, and somehow this blogpost almost joined that pile.
Let’s skip the rambling oh-so-human introduction dripping with so much disingenuous sincerity you can barely make out what the person is saying above their SEO ranking. …But not before I shamelessly plug my backcatalog ahem:
All of my previous tracks were painstakingly sequenced and arranged manually one bar at a time.
Being a blockpusher left something to be desired and after seeing a bunch of “live-looping” electronic music acts (often solo artists at that), I developed a new 5 step process for making music that people might actually want to dance to (and maybe even sign to a label or even give me like 99 cents).
This first step is actually the same as blockpushing, at least for every producer I’ve ever seen.
Find some sounds you like (a sample, a melody, a bassline, a drum pattern, some weird ambient texture, a vocal, whatever) and build it up with things that go with it, making one giant loop.
If you can solo and mute individual elements and start to imagine what the song will be, here is the departure from blockpushing.
oh god does this blogpost need screenshots? just use your imagination please D:
Load up clips into scenes and start to arrange not in terms of “4 bars of this then 2 bars of that where are we in the timeline oh that should come first”
Instead, be like “ok this group of sounds is a section, and this one is another” and leave room for improvisation.
Have at least one instrument ready to play during recording and have lots of knobs mapped to be able to modulate the sounds.
I’d also recommend having one-shot effects/samples ready to go (and quantized to play in time when you trigger them!)
Now, of course, single-handedly performing a very complex song with lots of different instruments at once in one take is difficult. So, I’ve cheated by recording solo sections while prepping a song and looped them, cut them up as needed and inserted them into the scenes as if they were prefabricated samples (well I mean they are but I made them for that song but whatever).
Common advice is “mix as you go” but that’s step 3 and we’re not there yet. I do however get stuff sounding “good enough” in what I refer to as “sound design” or as Timbaland puts it “sonics”. Just get stuff sounding to where you can vibe with it, but don’t try to get stuff sounding perfect upfront because that’s just going to disrupt this part of the creative process.
And so when you’re ready, imagine yourself playing this song you’re about to make
As in, imagine playing the finished track that’s been pressed to a record or burned to a CD or even just the audio file but that isn’t as cool.
Don’t think of this as a live performance act that you gotta be able to reproduce consistently. Unless it turns out that way and that becomes your thing.
This is simply a creative method for writing music
Anyway yeah, imagine that this song is playing in a specific place and time for a specific audience. Whether that’s a place you’ve been to or a place you want to go one day; real or imaginary; present, past or future and whether it’s for your best friend, or enormous crowds of total strangers as long as it’s anyone more than just yourself it’ll be great.
Hit record and just play.
Admittedly, apart from the first track I did this with which was a purely magical single take, sometimes you try to perform it and it’s a disaster.
Usually this comes down to needing to organize the project file better (which I’m terrible at) and needing to have a more solid structure in mind.
Sometimes it comes down to needing to practice your keyboard playing or realizing that you can’t in fact turn 3 knobs simultaneously in real time.
But even a failed take can tell you whether this song idea is worth pursuing. And the more upset you are that you failed the better the song is (probably).
I mean if you realized it wasn’t a good idea for a track aren’t you glad you figured that out before painstakingly arranging it trying to make it work spending hours on the mixdown paying to get it mastered just to have it flop on the dancefloor? Nothing was lost.
On average, it’s been taking me about a half-dozen attempts before I get a good take.
But even the best take is gonna need some edits
Line stuff up in time, maybe cutdown the 2 minute intro to something reasonable
or not :^)
Fix/add automation, remove superfluous sections, quantize notes, re-record the synth that you didn’t realize was out of tune until you listened to it on speakers…
Re-perform sections as needed, you get the idea. We’re back in blockpusher territory, but if you were vibing with what you were playing, you at least know you have a solid arrangement in there somewhere.
Inevitably I start doing mix stuff during the edit process (“that should be louder oh let’s get some exciter on the bass why is this damn thing so quiet”) especially stereo effects, but your focus first and foremost should be the arrangement during this step.
To borrow from programming (but I’m too lazy to paraphrase it so just figure it out):
OK I haven’t actually gotten to this step yet because I started mixing the first track and it took me a month, I realized my hardware was terrible and bought an audio interface and headphones and listened to it for a month straight and I’m still not happy with it.
And I’m too excited about this process right now and making as many complete song ideas as I possibly can to stop and focus on getting the mix perfect every time.
It’s probably definitely worth it to outsource the mix if you have the money and an engineer you trust knows the sound you’re going for.
Actually if I focus and try really hard I’m usually pretty satisified with my mixes
whatever this isn’t a confessional it’s a how-to.
At this point you’re doing the mixdown by making sure the track sounds the way it should sound on a soundsystem and doesn’t have phase issues, clipping, unwanted distortion/noise, etc.
But also it should sound as good as it possibly can on any sound system by enhancing the stereo image, the clarity and definition, and other nebulous nerdy audio engineer jargon.
It should sound as good as possible but not as loud as possible, that’s the final step.
Yeah I’m definitely not at this step yet with anything.
It’s very important to mix your stuff as quiet as possible so you can pay someone $200 per second of audio to slam it through a $10,000 limiter at +1000000 dB.
You know, it’s true, I didn’t have any idea how to mix a song when I started making music. I’m slightly ashamed to say that I didn’t know about or I should say understand the -6 dB of headroom thing until this year.
I mean I don’t understand it since 32-bit audio can be scaled down without loss of resolution (in other words, 32-bit audio can even be clipping above digital 0dB and be brought down to whatever volume you need without loss of quality).
I used to and sometimes still put a compressor and limiter on during the entire song writing process and A/B it during the mixdown.
But anyway yeah when I started I didn’t understand the mixing/mastering thing I just would look at the waveforms of songs I liked and tried to make my stuff “look” like that.
Which meant hitting digital 0 or getting pretty darn close to it most of the time.
And some of that stuff is actually loud and :thumbs_up_emoji:
But I’ve yet to actually master anything properly and idk if it can truly be done without a proper setup.
Maybe it’s actually definitely worth it to pay to have the tracks mastered by a professional.
I just hope they’re not going to yell at me about -6dB headroom D:
Hopefully a label would take care of the mixing/mastering end of it for me (or tell me what to do, or refer me to someone).
Maybe that’s not how it works. I don’t know.
Maybe you gotta be a one-person powerhouse songwriter/producer/engineer that only produces 100% original no-samples-used music that somehow sounds exactly like everything out there right now (but not last week). And you record, mix, and master a track every morning before hopping on a plane to go DJ in Tokyo and never sleep.
Unless you’re famous in which case you literally have a team of producers and engineers ghostwriting for you and you don’t even need to beatmatch or pay for music anymore.
I really don’t know.
All I know is that this process has helped me produce songs that I would actually send to a label and that’s not something I ever felt confident about before so yay.