Making growl presets has been my actual job since 2022, and I’m not proud of my first fifty attempts. Lots of buzz. Plenty of rasp. Zero talking. If your growls are stuck sounding like an angry sawtooth, here’s the serum 2 dubstep growl recipe I actually use on the Monosounds packs.
No single magic knob does it. What works is a chain: a wavetable with some vowel in it, the right warp mode, moving formants, distortion feeding a phaser, and a resampling loop that piles up character one layer at a time.
All of this got easier in Serum 2 than it ever was in Serum 1. The spectral oscillator on its own hands you growl material that simply didn’t exist before March 2025. So let’s build one from nothing.
Why a Serum 2 Dubstep Growl Actually Talks
Formants are the reason a growl sounds like speech. They’re the resonant peaks that separate an “oh” from an “ee”, and your ears are built to lock onto them. Push formants around in rhythm and nearly any tone starts sounding vocal.
Which means the entire job comes down to this: make something harmonically dense, then shove its formants around in time with the track. Everything below exists for that one purpose.
One rule before we start. Growls belong in the midrange, call it 150 Hz up to 2 kHz. Your sub stays a separate, untouched layer — the split gets its own treatment in my Serum 2 808 and sub bass guide. Whatever you do, keep the sub out of the growl’s distortion.
Pick a Vowel-ish Wavetable First
Don’t reach for the default saw. Go through the factory tables and drag the WT Position knob across each one, listening for anything that resembles a mouth opening and closing while you sweep. When a table sounds vocal with zero processing on it, that’s your foundation.
Even better: build the table yourself. Say “yah-woh-yoi” into your phone, chop the recording down, then drop the .wav straight onto oscillator A. Serum 2 turns it into a wavetable right there. There’s a full walkthrough in my custom wavetables guide, but the one-line version is that source audio with a steady pitch converts cleanest.
Where I begin: unison at 1 while I’m designing (a mono growl punches harder and spares the CPU), one octave down, level near 75%. Before touching anything else, route an LFO to WT Position. A table that doesn’t move will never talk.
Warp Modes: Bend and FM Carry the Sound
The warp section is what gets a frozen tone moving. For growls, two modes do most of the work.
Bend pushes the waveform toward one side. Somewhere around 30-50% it gives nearly any vowel table a nasal snarl. Take the LFO already sitting on WT Position and send it to warp amount too, at half the depth. When those two movements drift slightly against each other, the patch chews rather than wobbles.
FM controls how mean it gets. Put a plain sine on oscillator B one octave under A, flip A’s warp mode to FM from B, and turn it up. Around 15-25% you get a throaty rasp; go past 40% and it turns metallic and screamy, which works at a drop and grates everywhere else.
Reading about warp settings is slower than hearing them. Open a few patches from our Grumble Skrillex-inspired preset pack and look at where the warp amounts sit. FM warp shows up somewhere in nearly every patch.
One of those patches is below, so you can hear what those warp values sound like in practice.
Moving the Formant Filter
On to the filter. The filter menu in Serum 2 includes formant types that act like a vocal tract, so cutoff stops being a brightness knob and becomes a vowel selector.
Pick one, then put a slow LFO or an envelope on cutoff. Small range. Going from “oh” to “ah” is a tiny knob move; sweep the whole thing and you get a filter demo instead of a voice.
Set resonance somewhere in the 20-40% zone so the vowels stay defined. Any higher and the patch starts to whistle.
Now layer the motion: WT Position moving, warp moving, formant moving, each driven by a slightly different source. That’s three separate layers of vowel movement, and most of the time a serum 2 dubstep growl is already talking at this point, before any distortion gets involved.
Distortion Into Phaser: The Classic Chain
Exact settings matter less than the order here. Distortion goes first because it generates fresh harmonics; the phaser behind it cuts moving notches into them. Reverse them and all you get is fizz.
This is my default chain in the FX tab:
| Slot | Effect | Settings that matter |
|---|---|---|
| 1 | Distortion | Overdrive mode, drive 40-60% |
| 2 | Phaser | Rate 1/2 sync, feedback 60%, mix 50% |
| 3 | EQ | High-pass at 120 Hz, dip 3 dB near 400 Hz |
| 4 | Compressor | Multiband, tame the 2-5 kHz band |
| 5 | Hyper/Dimension | Mix 15-25% for stereo width |
Serum 2 now runs several FX buses, so the growl oscillators get their own bus and anything clean stays off it. You can also modulate individual effect parameters, something Serum 1 never allowed — an LFO on phaser feedback buys you an extra layer of talk for free. The full FX rundown lives in my Serum 2 effects guide.
LFO Shapes for Talking Rhythm
A plain sine LFO gives you a siren, not syllables. Draw the shape yourself.
My usual move: an LFO with three or four steps at different heights, curved falls between them, synced to 1/2. Most dubstep sits at 140-150 BPM in half-time, and 1/2-note movement lands right in that pocket.
Then add a second, quicker LFO at 1/8 with a small depth, aimed at the formant filter. That puts syllables inside each word. Flip it onto a triplet grid whenever you want the classic “yoi yoi” bounce.
Drag-and-drop modulation in Serum 2 makes the wiring fast: pick up the LFO header, drop it on WT Position, warp, and formant cutoff, then give each destination its own depth.
The Resampling Loop
Most tutorials never mention this step, and it happens to be where the thick, layered pro sound actually lives. A single pass through the synth doesn’t cut it. Every serum 2 dubstep growl we ship has been through this loop twice at minimum.
Print your growl to audio: a single held note, two to four bars, full FX chain baked in. Then feed it back. Serum 2 offers two routes: drag the .wav onto an oscillator and it gets sliced into a new wavetable, or hit Resample to Oscillator and the whole processed sound becomes a fresh table without ever leaving the plugin.
Your distorted, phased growl just became raw material again. Sweep the new table’s position and you’re scrubbing through your own processed audio. Now add a different warp mode, another formant filter, fresh FX. After two or three passes the sound takes on a density no single pass can reach.
Just one rule: switch the modulation rhythm on every pass. Repeat the same LFO shape twice in a row and the layers smear into mush.
The Spectral Oscillator: Growl Material Serum 1 Couldn’t Give You
The spectral engine in Serum 2 resynthesizes any sample into individual partials you can edit one at a time, plus frames you can morph across. For growl design it’s basically cheating.
Give it a short vocal phrase or one of your own resampled growls, then start breaking things: mute whole clusters of upper partials for a hollow, throaty tone, or slide partials against each other until the grit goes inharmonic. Morph between frames with an LFO and you get vowel motion without a formant filter anywhere in the chain.
Keep an eye on CPU, though. Spectral mode is pricier than wavetable mode. Run unison at 3-5 rather than 16, limit your polyphony, and design at 2x oversampling. If a patch still chokes your session, Xfer keeps official CPU guidelines on their support page.
A handful of bass patches in our free 250+ Serum 2 preset pack run the spectral engine as their mid-growl layer. Load them and see which partials got muted. Pulling apart finished patches teaches you faster than any article ever will.
Serum 2 Growl Questions I Get Asked
Why does my serum 2 dubstep growl sound thin?
Nine times out of ten the growl is fine and the sub layer is missing. High-pass the growl around 120-150 Hz and put a clean sine or 808 underneath it. If the mids genuinely feel weak, the fix is usually one more resample pass. Reference material helps as well; my best Serum 2 bass presets roundup is full of patches to compare against.
What LFO rate should I use for dubstep growls?
Main movement at 1/2 sync for 140-150 BPM tracks, 1/8 for the inner syllables, and 1/4 triplets for bounce. Push past 1/16 and it stops reading as speech and turns into texture.
Do my old Serum 1 growl presets still work?
They do. Serum 2 opens Serum 1 presets directly, and Serum 1 owners got the upgrade for free. Old growls also make prime resampling material: render them out and rebuild the midrange with the spectral engine.
Do I need paid presets to make good growls?
Nope. Every technique in this article runs on factory content alone. Think of presets as a learning shortcut rather than a requirement. Grab our free Serum 2 bass presets, study how the mod matrix is wired, then build your own from zero.
Go Make Something Ugly
So that’s the full recipe: a vowel table, Bend plus FM warp, formant movement, distortion into a phaser, an LFO you drew by hand, then resample and run it all again. The first serum 2 dubstep growl you build this way will be rough. By the tenth, it’ll talk.
Want finished patches to pull apart? Our free pack with 250+ Serum 2 presets covers bass, growls, leads, keys, all of it. Royalty-free, no catch. Break them open and study how they’re wired.
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 →Looking for the packs themselves? Here is all my Serum 2 preset packs — sorted by genre and by sound, with a demo on every one.
