There’s one reason layering fails: each layer is trying to cover the whole spectrum. Stack three full-range patches and the result isn’t triple the size; it’s mud wearing a chorus.
The cure is unglamorous and reliable: assign every layer a single job, then filter away everything outside it.
Give every layer one role
When I stack synths I think in three lanes:
- Sub (under roughly 100 Hz): the weight. Something sine-like, mono, zero effects.
- Body (100 Hz up to 2 kHz): the tone people actually recognize. Your main patch lives here.
- Top (2 kHz and above): sparkle and air. A bright pluck, a noise layer, or your main patch cloned an octave higher and filtered.
Often two layers do the job: body with sub, or body with top. Three is the ceiling. Five layers doesn’t buy you a bigger sound; it buys you five problems.
Think about octaves when choosing, too: body at the written pitch, top pushed one or two octaves higher. Most mud is born from three patches crammed into one octave, before mixing even enters the picture.
Big brass-style stacks make a useful benchmark: massive, yet the weight, the body and the bite each arrive from their own defined place.
EQ each layer into its lane
Filters are how the roles get enforced:
- Put a highpass on the body around 100–150 Hz so the sub gets left alone.
- Lowpass the sub at about 100–120 Hz. Nothing gets through above that.
- Highpass the top layer hard, 2 kHz minimum. Soloed it’ll sound ridiculous; inside the stack it’ll be exactly right.
Beware of solo listening. A layer that’s been properly cut into its lane sounds thin and broken by itself. The sum is the only thing worth judging.
Stacked vocal textures obey the same law: cut low, sitting high, they float over the top without crowding the mids. This is what that kind of stack sounds like.
Check phase before you blame the EQ
When adding a layer makes the combined sound thinner, you’re hearing phase cancellation, not a bad EQ. The usual culprit is two layers overlapping in the lows; a pair of saws on the same note can partly cancel depending on where their cycles happen to line up.
Fast ways to diagnose it:
- Invert polarity on one of the layers. A big tonal shift means the layers are tangled in the low end, so cut the lane filters tighter.
- Listen to the stack in mono. Wide layers that evaporate in mono are cancelling.
- Slide one layer’s start by a few milliseconds and compare. Occasionally that alone fixes everything.
Prefer to see the problem rather than hear it? Bounce both layers plus the sum and zoom into the waveforms. A summed low end that looks smaller than either layer on its own is your cancellation, caught red-handed.
The structural fix that never fails: exactly one layer gets to exist below 150 Hz. Phase trouble lives in the lows, so if no lows are shared, no trouble exists.
Keep unison off the bottom
Unison and detune go on the body and top, never on the sub. Detuned voices wander in and out of phase, which makes the low end pump at random. Stereo wideners are banned down there for the same reason. The bottom layer stays mono, single-voice, boring — and that boring stability is exactly what makes the wide upper layers read as huge rather than seasick.
When one good preset beats three layers
A confession from someone whose job is making presets: a properly designed patch is already a stack. One oscillator holds the body, a sub oscillator sits below, noise supplies the top — and it’s all phase-aligned at the source. If you’re piling three patches on top of one weak one to save it, the quicker fix is a stronger starting patch. In that situation, browse every Serum 2 pack and grab one built for the job rather than stacking three that weren’t.
My personal limit is twenty minutes. If the stack hasn’t come together by then, I quit stacking and replace the source patch instead. That single rule has rescued more sessions than any layering technique I know.
Try it on real material: download the free Serum 2 presets, find the fullest patch in the set, and ask whether it needs a layer at all before you build a stack around it. And if cross-genre layering is a regular thing for you, the Serum 2 preset bundles deliver the broadest palette of ready layers for the money.
Related reading: when a single patch sounds weak by itself, why synth sounds thin goes through the patch-level causes. If instead your stack piled up weight in the low mids, you’ve wandered into muddy bass territory.
FAQ
How many layers should a synth sound have?
Two or three, each with a distinct role. Past three, the usual truth is that none of them was the right patch in the first place.
Should layers play the same notes?
Same notes or octave copies. The sub doubles the line, usually an octave below. Give each layer its own harmony and you get mush in a hurry.
Why does my layered sound disappear in the mix?
Phase cancellation, most often, or midrange collision with other instruments. Flip polarity to check, listen in mono, and pull each layer’s lane EQ tighter.
Do I need different synths for each layer?
No. Two Serum 2 instances running different patches layer perfectly well. Mixing synths adds character variety, along with less predictable phase behavior in the lows.
One job per layer, filters holding the lines, a mono sub, and judge only the sum. That’s the whole discipline — everything else is choosing good source patches.
I run Monosounds.studio and make the packs myself. Ten-plus years of producing, 97+ releases, 16,000+ downloads. Every Serum 2 preset we ship passes a CPU check on a mid-range laptop and gets used in a real track before release. If it does not survive a mix, it does not go in the pack.
More about Maxim & the studio →
