The kick on the record you keep going back to is almost never one sample.
I spent an embarrassing amount of time not knowing that. Load a kick from a pack. EQ it. Compress it. Saturate it. Wonder why it still sounded like a demo next to the reference track. Load a different kick. Repeat until 2am.
The problem was never the processing. It was that I was asking one sample to do three completely different jobs, and most samples are only good at one or two of them.
So let's talk about the three jobs.
A kick has three jobs, and they barely overlap
Think about where a kick has to work. A club rig with subs you feel in your ribs. A car. A laptop speaker the size of a postage stamp. Someone's cheap earbuds on the bus.
Those are four completely different physics problems, and no single sample solves all of them.
Sub. The weight. The part you feel more than hear. This is the low fundamental, usually somewhere around 40 to 60 Hz depending on the key and the genre. On a proper system it's the thing that makes the room feel pressurized. On a phone it is completely, totally absent. Not quiet. Absent. The speaker physically cannot make that frequency.
Body. The punch. The thump. This lives in the low mids, roughly 80 to 200 Hz, and it's the part that actually makes people move. If sub is the weight, body is the impact. It's also where the character lives. When you describe a kick as "909," "wooden," "boxy," "distorted," "round," you are describing the body.
Click. The top. The transient. Anywhere from about 2 kHz up. Soloed, it sounds ridiculous, like someone tapping a pen on a desk, and the first time you hear it on its own you will assume you did something wrong. You didn't.
Here is why the click matters more than anyone tells you: it is the only part of your kick that survives earbuds, phone speakers, laptops, and a loud club where the sub is already saturated by everything else in the mix. On a phone, the click is the kick. That's the whole kick. That's it.
Three jobs. Three sample slots. That's the entire idea.
How to pick a sample for each role
The listening test is different for each one, and that's the part people skip.
Picking a sub.
You want something close to a sine. Minimal harmonic content, steady pitch, clean tail. Listen for a tail that wobbles or drifts in pitch, because that drift will fight your bassline later and you won't know why.
Also listen for whether the sub sample has a click of its own baked into the front. Most do. That built-in transient will fight the click layer you add later. You'll filter it out anyway, but know it's there.
Quick test: play the sub sample through your phone speaker. If you hear almost nothing, congratulations, you picked a good sub.
Picking a body.
This is the identity of the kick, so pick it like you're picking a lead sound. Play it at conversational volume on whatever's nearest. Does it have attitude? Does it sound like the genre you're making? Is there an audible pitch drop at the start, and do you like the shape of it?
Don't pick a body sample for its low end. Pick it for its character. The low end is somebody else's job.
Picking a click.
Short. Ruthlessly short. Anything with a tail on it is a hi-hat, not a click.
Two things to listen for. One, is there low end hiding underneath it? Usually yes, and you'll high-pass it out. Two, is it pitched? Some clicks have an audible tone, and a pitched click can clash with the key of the track in a way that's genuinely hard to diagnose.
A good source of clicks nobody talks about: the first 15 milliseconds of a snare, a rim, a clap, or a woodblock. Chop the front off and throw the rest away.
The order of operations: body first, always
There's a real argument about where to start, and I land firmly on one side.
Start with the body.
The body layer is the identity of the kick. It's the part a listener would actually describe if you asked them what the kick sounds like. Everything else is support.
If you start with the sub, you'll fall in love with a sub, and then spend forty minutes auditioning bodies that all sound slightly wrong against it. You're now solving a compatibility problem instead of making a creative decision.
Body first. Then sub, for weight. Then click, for translation.
There's a nice logic to that order, too: character, then power, then clarity. Each layer is fixing a specific complaint about the one before it.
Give every layer its own lane
This is the step that separates a layered kick that sounds huge from a layered kick that sounds like mud with extra steps.
If all three samples run full range, you now have three low ends stacked on top of each other. They interfere. Some frequencies add, some cancel, and the result is a kick that meters loud and feels weak. It's the single most common way layering goes wrong. I wrote more about the mechanics of that here: why your layered kick sounds weaker.
The fix is boring and it works. Filter each layer to its job.
- Sub: low-pass it, somewhere around 90 to 120 Hz. It has no business above that.
- Body: high-pass around 50 to 70 Hz so it stops competing with the sub. Low-pass it somewhere in the 800 Hz to 2 kHz range so it stops competing with the click.
- Click: high-pass it, 800 Hz to 2 kHz. Aggressively. Yes, it'll sound thin. It's supposed to.
Those numbers are starting points, not laws. Move them with your ears. But the principle holds: three layers, three lanes, minimal overlap.
Solo each layer after filtering. Each one should sound incomplete and slightly stupid on its own. That's how you know it's doing exactly one job.
Phase and tuning, briefly
Two more things, both worth their own article, both of which have one.
Phase. Any time two layers have low-frequency content, their alignment matters enormously. Nudging one layer by a single millisecond, or flipping its polarity, can change the perceived level of the whole kick by a surprising amount. Always check in mono. More here: why your layered kick sounds weaker.
Tuning. Your sub and body layers have a pitch whether you chose it or not. If that pitch disagrees with the key of the track, the kick and the bass will do a slow wobble against each other on every hit and the low end will feel unstable. Getting them in the same key is often the single biggest improvement you can make. How to figure out what key a sample is in: how to find the key of a kick sample.
Why three, and not five, or eight
Because phase problems compound, and they compound fast.
Every pair of layers has a phase relationship to manage. Three layers is three relationships. Five layers is ten. Eight layers is twenty-eight. You can do the math and see where that goes.
Meanwhile the audible return drops off a cliff. Layers four and five are almost never adding a new job. They're adding another opinion about a job that's already covered.
Three layers cover all three frequency zones with almost no redundancy. If your kick still isn't working with three, adding a fourth won't fix it. Something is wrong with one of the three you have.
Then commit. Render it to one sample.
Once it sounds right, print it. Render the whole stack down to a single audio file and use that.
Three reasons.
CPU. Three sample players plus filters plus phase alignment, on every kick in the project, adds up in a way that gets annoying at mix time.
Chopping. One sample is a thing you can reverse, pitch, gate, stutter, chop, and drag into the next project. A three-layer chain is a thing you have to open and negotiate with.
Your brain. This is the real one. An un-rendered kick is an open invitation to keep tweaking. I have ADHD, and I will happily spend forty minutes on a kick and zero minutes on the arrangement. Printing it to one file closes the door. The kick is done. Go write the track.
How the balance shifts by genre
Same three layers, different mix.
House. Body-forward. The bassline is doing real work down there, so the sub layer stays modest and short so the two aren't fighting for the same space. Warmer, rounder body. Click present but polite. You want the kick to feel like it's part of a groove, not standing in front of it.
Techno. Sub and body both louder and longer. The kick is often carrying the entire low end by itself, so it can afford to take up room. Click can be aggressive, distorted, even nasty. Longer tail. The kick is the track.
Tech house. Tight and short across the board. Click-forward, because so much of tech house gets heard on phones, in cars, and on club systems where the mids do the work. Short controlled sub. The body does the punching, the click does the talking.
Same recipe. Different amounts.
The short version
- Pick your body layer first. It's the character.
- Add a sub for weight. Low-pass it.
- Add a click for translation. High-pass it hard.
- Give each layer its own frequency lane.
- Check phase in mono.
- Tune the pitched layers to the track.
- Stop at three.
- Render it and move on.
That's it. That's the whole method. You can do this in any DAW with three sample players, three EQs, and some patience.
And if you'd rather not do it by hand
I built Kick Alchemist because I got tired of building that chain from scratch every single time.
Three slots, named for the jobs: BASE for sub, CORE for body, SPARK for the click. Per-layer tuning to the track key. Per-layer phase alignment. CARVE hands each layer its own frequency range so they stop fighting. COOK renders the whole stack down to one sample you drag straight into your DAW. And TRANSMUTE shuffles layer combinations when you're staring at a blank project and need somewhere to start.
It layers samples you already own. It isn't a synth and doesn't pretend to be.
Windows 10/11, VST3, $69, one-time, no subscription. There's a free demo with everything working for 25 minutes a session, which is enough to build a few kicks and see if you like it.
Or just do it by hand. The method's the same either way, and the method is the part that actually matters.