how to make your bass sound old school (by making it worse)

you load up a bass patch, play the line, and it's fine. clean, tight, perfectly in tune, no noise anywhere. and it sounds like nothing. then you put on a record from 1996 and the bass on it is smaller, duller and noisier than yours, and it hits ten times harder.

here's the uncomfortable bit: those old records didn't sound like that because anyone chose it. the gear was just worse. the samplers people made those tracks on threw away most of the detail before the sound ever reached tape, and thirty years later we're all still chasing what that throwing-away did. so if you want your bass to sound old, you don't add anything. you take things out, on purpose, in a specific order.

what the old gear was actually doing wrong

take the E-mu SP-1200, released in 1987 and still the reference point for gritty sampled drums and bass. according to Wikipedia's entry on it, the machine's character, "often described as 'warm,' 'dirty,' and 'gritty'", is put down to its "low 26.04 kHz sampling rate, 12-bit sampling resolution, drop-sample pitch-shifting, and analog SSM2044 filter chips". that low sample rate wasn't an artistic decision either: the same entry says it "was chosen early on" as "a compromise between bandwidth and sampling time". memory was expensive, so corners got cut.

one of those corners is worth knowing about. Wikipedia notes that on the SP-1200 "a reconstruction filter was deliberately omitted, resulting in a brighter sound due to imaging". a reconstruction filter is the bit of circuitry that's supposed to smooth a digital signal back into a clean analogue one. leaving it out let extra, unintended frequencies through, and that mess is part of why people still want the thing.

but don't overdo the picture in your head. MusicRadar's guide to old-school sampling is blunt about it: "the sound of a 12-bit sampler, for one, isn't always completely filthy, it generally just sounds a bit tougher and punchier." tougher and punchier. that's the target, not destruction.

step one: bit reduction

bit depth is how precisely each moment of a sound is measured. Attack Magazine's guide to replicating vintage samplers explains that the numbers used to store a sound set its dynamic range, and "the larger the size of number used, the greater the dynamic range", with 8-bit sampling delivering only 48 dB against the roughly 120 dB of human hearing.

in plain words: imagine a ladder. every instant of your bass has to land on one of the rungs. with a lot of rungs, it lands almost exactly where it should. take rungs away and every instant gets rounded to whichever one is nearest, so the sound comes back slightly wrong, all the time. that constant small wrongness is the grit. that's it. that's the whole effect.

MusicRadar's explainer puts the mechanism simply: "bitcrushing (or its closely related cousin, sample rate reduction) is simply the process of reducing the accuracy of a digital signal", and "the more information we discard, the more inaccurate the signal gets". so the knob you're turning is literally an accuracy knob, turned backwards.

• start around 12 bits. MusicTech's lo-fi sampler guide says to "drag the number of bits down to 12 or lower", which is SP-1200 territory and about where you get character without obvious damage.
• listen to the tail, not the attack. bit reduction shows up most in the quiet part at the end of a note, where the signal is small and the rounding is proportionally bigger. if your bass notes are very short, you'll need more of it before you hear anything.
• do it before anything else. crushing a sound that's already been filtered and driven just crushes the processing along with it. put it first in the chain.
• check it in mono once. bit reduction adds noise-like content on both channels, and that can read differently when left and right are summed together, so flip your monitoring to mono before you commit.

step two: take the top off

bit reduction adds an edge, and left alone that edge is fizzy and modern-sounding, which is the opposite of what you're after. so the second move is to throw the highest frequencies away.

this is the other half of what the old machines did to you. Attack Magazine's explanation of sampling is that you capture a sound by measuring it over and over, and "the speed with which we do this (the sample rate) determines the maximum frequency that can be captured and played back". a sampler running at 26 kHz simply cannot store the very top of your sound, so it doesn't. everything above a certain point is gone, and the result is warmer and rounder because there's nothing bright left to be bright.

you get the same result with a low-pass filter, which lets the low frequencies pass and cuts the highs. every DAW has one. Attack's own summary of the pair is that "as we reduce the sample rate and bit depth, we bring out more character and crunch in the sound", and MusicTech suggests pulling the sample rate down alongside the bit depth, "such as from 44.1kHz to 22kHz", for extra authenticity.

• a touch, not a shutdown. in the video it's "a touch of low pass", no more than that. you want the fizz gone and the note still clear, not a muffled thud.
• sweep it with the mute button. pull the filter down until the bass sounds dull, then back it off until it comes alive again. the useful spot is usually just before you'd have called it dull.
• leave the low end completely alone. this step is about the top. if you find yourself touching anything under 200Hz you've picked up the wrong tool.

step three: drive, just enough to bite

now the sound is gritty and dark, and it's also probably a bit limp, because you've been removing things for two whole steps. drive puts the energy back.

drive is deliberate distortion, and it does two things at once. MusicRadar's explainer separates the family: distortion is "any form of processing which changes (or perhaps damages) the signal", while saturation is the gentler, tape-style version that engineers tend to find pleasant. either way, the peaks of your waveform get squashed flat and new harmonics appear that weren't there before.

the second thing is the one people miss. Sound On Sound, writing about getting mixes loud, points out that this kind of gradual distortion "is like compression", because "it can reduce the dynamic range of the signal, and this has the effect of increasing its average level". dynamic range is the gap between the loudest and quietest part of a sound. squash the loud peak and everything underneath it can come up, so the note sounds fuller and more constant without your fader moving. that's why a driven bass feels like it's pushing against the speaker even when the meter says it isn't any louder.

• just enough to bite. turn it up until you can hear the edge appear, then come back a little. if the note stops sounding like a note, you've gone too far.
• watch the level. drive makes things louder, and louder always sounds better for about three seconds. match the volume before and after so you're judging the character, not the level.
• if it gets harsh, go back a step. harshness usually means the bit reduction was too aggressive and the drive is amplifying it. lower the crush, then re-drive.
• a little goes further on a busy track. solo'd, drive sounds obvious. in a full mix, the same amount can disappear, so check it against the drums before you commit.

why the order matters

crush, filter, drive. try it in any other order and you get a different, worse result, for reasons that are easy to follow once you see them. crushing first means you're degrading the actual bass, not a copy of it that's already been coloured. filtering second cleans up the harshest part of what the crush added before it gets amplified. driving last means the drive is reacting to the sound you've built, and its harmonics land on top of everything rather than underneath it.

run the drive first instead and you feed a loud, bright signal into the bit reduction, which then crushes the distortion itself. it sounds fizzy and cheap rather than old, and no amount of filtering afterwards gets it back.

common mistakes

• going straight to the extreme setting. MusicRadar's take on the pair is that they first introduce "a bit of mild grit before breaking up completely at extreme settings", so the useful range is usually near the bottom of the knob. you don't need much.
• doing it to the whole bass bus. if you have a sub layer under your bass, keep it clean. crush the layer you actually hear, leave the weight alone.
• judging it solo. lo-fi processing is a mix decision. it sounds dramatic on its own and much subtler in context, so audition it against your drums.
• forgetting it's a one-way trip. once you print it, that information is gone. keep the clean version on a muted track for when tomorrow-you disagrees.
• using it to fix a boring bassline. character processing makes an interesting part more interesting. it does not make a bad part good. if the line itself isn't moving, fix that first.

one knob instead of three

in the video that whole chain is a single control. bassBox has a knob called CHARACTER that ages the sound instead of distorting it: turn it up and the bit depth drops away, the top end rolls off, and a soft saturation glues what's left back together. that's the three steps above, in the right order, at settings that stay usable the whole way. in the video we take it from 1% to 27% to 63% and the same bassline goes from clean to properly old without ever falling apart.

there's a separate DRIVE knob for when you want the aggressive version, which pushes up to 6dB into the saturator and pulls some back out the other side, gentle through the bottom half of its range and properly mean at the top. and the FILTER knob cuts the top off or sweeps it open if you'd rather do that part by hand. the presets underneath are 141 basses multisampled from real hardware, a DX7II, a Jupiter X, a JV-1080 and an OB-6, with 21 zones recorded per preset so they track properly across the keyboard. we wrote about those four machines and why each one sounds the way it does in four machines, one low end.

and if you want to think about saturation more broadly, the same logic runs through drums: we covered how much is too much in stop compressing your drums to death.

so: crush it, dull it, drive it. same bassline, more character, and about thirty seconds of work. press play. make a belter.

this post started as one of our tiktoks. the whole chain is in there if you'd rather hear it than read about it.

sources

• E-mu SP-1200 (Wikipedia)
• Character & Crunch: Replicating The Sound Of Vintage Samplers (Attack Magazine)
• Distortion, saturation and bitcrushing explained (MusicRadar)
• 10 authentic old-school sampling tips (MusicRadar)
• How to achieve a lo-fi hardware sampler sound in your DAW (MusicTech)
• Crafting Loud Mixes That Sound Great (Sound On Sound)