Saturation on Vocals: Which Type for Which Problem
Vocal mixing · 4 min read
Pick the saturation by the problem. Tape for glue and a softer top, tube for warmth, transistor or diode for edge, transformer for weight in the low mids. Drive the input and pull the output back by the same amount: that changes the harmonics without changing the level.
Drive, then trim
+6 / -6 dB
Same level in and out · completely different harmonics
The 60-second answer
Put the saturator after the first compressor and before the presence EQ. Push the input 3–8 dB, pull the output down by the same amount, and set the mix knob to 20–50% on a lead. Tube for warmth, tape for glue, transistor for edge, transformer for weight. De-ess 1–2 dB after it.
What saturation actually adds
A nonlinear circuit generates harmonics that were never in the signal: multiples of every frequency the voice contains. Second-order harmonics sit an octave above the note, which the ear reads as musical — thicker, warmer, closer. Third-order harmonics sit an octave and a fifth up, which reads as edge. Which ones you get depends on the transfer curve: asymmetric single-ended circuits, the classic valve stage, lean even, while symmetric clipping (most solid-state and diode distortion) leans odd. Warmth is even, aggression is odd.
Why it makes a vocal louder with no compressor
Two things happen at once. The curve rounds off peaks, so you get roughly 1–3 dB of peak reduction before anything sounds distorted, which lets the average level come up. And the new harmonics land above the fundamental, mostly in the 1–6 kHz region where hearing is most sensitive. The meter barely moves and the vocal reads more present — the cheapest way to make a voice cut, and the easiest thing to overdo.
Which type for which problem
| # | Do this |
|---|---|
| 01 |
Tape — glue and a softer top Low-order harmonics plus real compression from the medium, softer above roughly 10 kHz, a gentle lift in the low end. For a harsh vocal detached from the beat. Drive 3–6 dB · mix 40–100% |
| 02 |
Tube — warmth and body Asymmetric valve stages favour second-order content, an octave up, reading as body and closeness. Use it on a thin voice before a 200 Hz boost. Drive 4–8 dB · mix 30–60% |
| 03 |
Transistor or diode — edge Symmetric clipping favours third-order content, an octave and a fifth up. Bite and aggression. Good on a rap lead or an adlib bus. Drive 2–5 dB · mix 15–35% |
| 04 |
Transformer — weight Iron saturates at low frequencies first, so it thickens 60–250 Hz and adds crossover-region grit at low levels. For a lead that sounds small on a phone. Drive 2–4 dB · mix 50–100% |
| 05 |
Trust the analyser, not the badge Feed a held note through and watch where the new peaks land. Two plugins with the same label can behave differently. Analyser on a held note |
| 06 |
Place it after the first compressor Drive depends on input level, so it needs a stable one. Keep it before the presence EQ too. Comp → sat → EQ |
| 07 |
Two small stages beat one big one 3 dB into a tube and 3 dB into a tape beats 6 dB into either: two harmonic sets, not more of one. 2 x 3 dB |
The trick: drive the input, trim the output, never chase it with the mix knob
Every analog-modelled saturator has an internal reference level, and how much harmonic content it makes is set by how hard you hit that reference, not by the wet-dry balance. Leave the input alone and raise the mix and all you get is more of a weak effect, which mostly reads as a slightly louder vocal. Push the input 6 dB and pull the output down 6 dB instead: the level out is identical, the meter does not move, and the harmonic content is a different animal. You are now comparing tone against tone with no loudness bias. So: drive, compensate on the output, A/B against bypass, then use the mix knob to dial back what you chose.
Stop guessing the chain.
The Vocal Chain Bible is 88 real vocal chains from Grammy-winning engineers — every plugin, every setting, in order, with free alternatives for each one. 5+ years, 100+ interviews, one PDF.
Get the Vocal Chain Bible — $48 →Mistakes that kill it
- Saturating a track that was never gain staged. Analog models respond to input level, so the same settings on a hotter track are not the same settings.
- Saturating before the corrective EQ. A resonance you have not removed yet gets harmonics of its own.
- 100% wet on the lead. Keep the mix at 20–50%. Save fully wet for a parallel bus, adlibs or a printed effect.
- Judging with the output louder. Saturation always raises perceived loudness. Match the level or you are picking the louder option.
Saturation is a tone decision made with a level control. Set the drive, then give the level back.
Keep going
FAQ
What saturation is best for vocals?
Tube or tape for most leads: warmth and glue without edge. Transistor-style clipping when the vocal has to cut through a dense beat.
Where does saturation go in a vocal chain?
After the first compressor, before the presence EQ. The compressor gives it a stable input; the EQ shapes the result.
How much saturation is too much on vocals?
If you hear it as distortion on a solo listen at normal volume, it is too much for a lead. Aim for the point where bypassing it makes the vocal smaller.
Who's behind this
Ivan Meshcheriakov · IvanFromRnD
5+ years ago I moved to LA with zero connections and started building from nothing — 14-hour days, meeting people and interning at studios. It took me 5+ years to meet and interview 100+ Grammy-winning engineers — then I turned it all into the Vocal Chain Bible.
Read my full story →