Tutorials Audio

Saturation and Distortion Explained

By Infinite Creation · Intermediate · ~20 min

A 1960s mid-century comic book style editorial illustration showing two audio engineers at a mixing desk, with an audio waveform display, reels of magnetic tape, stacked storage drives visible in the frame.

Overview

"Warmth" "grit" "glue", saturation gets described in every vague adjective except what it actually does: add harmonic content by gently (or aggressively) reshaping a waveform. This guide explains the mechanism plainly and gives a clear point in your mixing chain to actually use it.

What You Need

Any saturation plugin. Most DAWs include a basic tape or tube saturation stock plugin
A mixed track or bus to experiment on

Steps

Understand what saturation does to a waveform

Saturation gently clips or reshapes the peaks of a waveform, generating new harmonic overtones that weren't in the original signal. This is what creates perceived "warmth" and density. Heavier distortion does the same thing more aggressively and audibly.

Choose a type for the effect you want

Tape saturation rounds transients and adds gentle compression-like glue. Tube saturation adds mostly even-order harmonics for a warm, musical thickening. Hard clipping/distortion adds odd-order harmonics for an aggressive, edgier character. Pick based on genre and how obvious you want the effect to be.

Apply it after EQ and compression

Saturation reacts to whatever signal reaches it, apply it after your corrective EQ and compression so it's shaping an already-controlled signal, not amplifying problems that should have been fixed earlier in the chain.

Drive it gently and check in context

Start with the lightest drive setting and increase gradually while listening in the full mix, not solo, saturation that sounds appealingly "warm" solo often reads as harsh or fatiguing once every other element is also playing.

Use it to fix, not just to flavour

Saturation can make a thin digital recording sound fuller, help a quiet element cut through a dense mix without more volume, or add perceived loudness without a limiter. Treat it as a functional tool, not only a stylistic garnish.

Pro Tips

A little saturation on the master bus (very subtle) can help a mix feel more "finished" before it even reaches mastering, but keep it light enough to be nearly unnoticeable in isolation.
Saturation and compression often substitute for each other to some degree. A track that needs heavy compression to feel "there" sometimes responds just as well to lighter compression plus a touch of saturation.
Bypass frequently while dialling in an amount, saturation's effect is easy to over-apply once your ears adjust to it.

Tested: which curves add which harmonics, and what drive and loudness do

We tested the page's account of saturation on 2026-09-29 with FFmpeg 8.1.2 and numpy. A 1 kHz sine went through waveshaping curves, and the harmonics were read from an FFT as decibels against the 1 kHz tone. The symmetric curves are FFmpeg's own asoftclip filter, and its tanh matched numpy's to better than one part in a hundred thousand. The asymmetric curve is our own model, a bias added before the curve and the DC taken out afterwards, which is what a bias does in a tube stage. It is a model, not a plugin. Nothing was listened to.

Curve	2nd harmonic (dB)	3rd (dB)	4th (dB)	5th (dB)	Total harmonic distortion (%)
Hard clip (FFmpeg asoftclip)	none	-10	none	-16	36.8
tanh (asoftclip)	none	-18	none	-34	12.2
cubic (asoftclip)	none	-19	none	-98	11.1
sin (asoftclip)	none	-19	none	-50	10.9
tanh with a bias, DC removed (numpy model of a tube stage)	-19	-19	-31	-38	15.9
tanh, sine peak 0.1 into a curve that clips at 1	none	-62	none	-122	0.1
tanh, sine peak 0.3 into a curve that clips at 1	none	-43	none	-84	0.7
tanh, sine peak 0.6 into a curve that clips at 1	none	-31	none	-61	2.8
tanh, sine peak 1 into a curve that clips at 1	none	-24	none	-45	6.7
tanh, sine peak 2 into a curve that clips at 1	none	-15	none	-28	17.3

The even and odd claim holds, and it is clean. Every symmetric curve, hard clipping included, put no measurable energy in the 2nd or 4th harmonic, and a real 3rd and 5th. The biased curve put its strongest new harmonic at the 2nd, -19 dB down, and still added odd ones. So the page's split is right about the mechanism: the symmetry of the curve decides which orders appear, not whether the effect is called saturation or distortion.

Distortion and saturation are one mechanism at different amounts, as the FAQ says, and drive is the amount. With tanh, a sine peaking at 0.1 into a curve that clips at 1 gave 0.08 percent distortion, 0.3 gave 0.7, 0.6 gave 2.8, 1.0 gave 6.7, and 2.0 gave 17.3. The 3rd harmonic rose with every step. That supports "start with the lightest drive": distortion rose about 200 times across the range, and the top of it is far past subtle. The curve matters as well as the amount. On the same sine, hard clipping put its 5th harmonic 19 dB above tanh's, which is the edge the page describes.

The loudness claim holds in its measurable half. We ran a chord and kick signal through tanh at rising drive and scaled each result back to the same -1 dBFS peak. Integrated loudness went from -17.0 LUFS with no saturation to -15.4 at the lightest drive tried and -5.9 at the heaviest, a rise of 11.1 LU at the same peak, while the peak to loudness ratio fell from 14.3 to 3.4 dB. The heaviest drive here was hard enough to flatten the signal almost to a square wave. So saturation does raise loudness without raising the peak, by squashing peaks, which makes it a form of limiting. Whether it sounds bigger at the same LUFS, the page's other claim, was not tested.

"Adds harmonic content" leaves something out. Two tones, 1000 and 1370 Hz, through tanh, produced not only harmonics but intermodulation products at 630 and 1740 Hz, -21 and -21 dB against the tones, and nothing at the sum of the tones, since a symmetric curve makes only odd order products. The 3rd harmonic of the lower tone was -34 dB, so the intermodulation products were stronger than it. Those extra products are not harmonics of either tone, which is one reason a chord or a full mix is not the same case as a single note. The page's consonance point is about the harmonics of one note and does not cover these products, and we did not test how they sound.

Soft clip a file with FFmpeg's tanh curve, driving it 3 times harder first: ffmpeg -i in.wav -af volume=3,asoftclip=type=tanh out.wav

Validated. Measured 2026-09-29 with FFmpeg 8.1.2 (asoftclip, ebur128) and numpy. Steady sines and a synthetic chord and kick were used, and real material behaves less regularly. The biased curve is our own model of a tube stage. We did not test tape (no hysteresis, no head bump), any plugin, the claim that even harmonics sound warmer or that saturation sounds louder at the same LUFS, or the source by source advice. The command is the filter that was measured.

Knowledge Base

Even-Order vs. Odd-Order Harmonics

Even-order harmonics (added by tube/tape-style saturation) tend to sound warm and musically pleasant because they're consonant with the original note. Odd-order harmonics (added by hard clipping and most distortion) sound edgier and more aggressive because they're less consonant. This is the core reason tape/tube saturation and distortion pedals feel so different despite both being forms of the same underlying process.

Saturation as Perceived Loudness

Adding harmonics increases a signal's perceived loudness and density without necessarily raising its peak level. This is part of why a saturated mix can sound "bigger" at the same measured LUFS as an unsaturated one, and why light saturation is a common mastering-stage trick.

Where It Fits in a Full Mix

Common uses: on individual drum tracks for punch and character, on a vocal for presence and warmth without more volume, on a bass to help it translate on small speakers, and lightly on the master bus for overall cohesion.

Where This Fits

Saturation follows EQ and compression in the chain Mixing a Full Song lays out from rough levels to bounce.

FAQ

Q: Is saturation the same as distortion?
A: They're the same underlying mechanism (waveform reshaping) at different intensities, saturation is the subtle, often barely-audible end. Distortion is the obvious, aggressive end.

Q: Can saturation replace EQ?
A: No. EQ shapes existing frequency content. Saturation adds new harmonic content. They solve different problems and work well together, not as substitutes.

Vocabulary

Saturation. The gentle reshaping of a waveform's peaks that adds new harmonic content not present in the original signal. Perceived as warmth or density rather than as distortion, though the two are the same mechanism at different intensities.

Harmonic. A frequency that is a whole-number multiple of a signal's fundamental pitch. Saturation and distortion both work by adding harmonics that were not in the source, which is what makes a saturated signal sound thicker without literally being louder.

Odd-order harmonic. A harmonic at an odd multiple of the fundamental, such as three or five times its frequency. Hard clipping and most distortion pedals produce mostly odd-order harmonics, which reads as edgier and less consonant with the original note.

Even-order harmonic. A harmonic at an even multiple of the fundamental, such as two or four times its frequency. Tape and tube-style saturation produce mostly even-order harmonics, which reads as warm because they are musically consonant with what is already there.

Drive. The control that sets how hard a signal pushes into a saturation or distortion effect. Low drive adds a little harmonic content, high drive pushes toward audible distortion. The same source runs cleaner at a lower drive on the same setting.

Where to start, by source

A drive setting that works on one source can be far too much on another, so a starting point saves the trial and error of finding it by ear.

On drums, start light: enough tape or tube saturation to add punch to a snare or a kick without the transient audibly softening. A few percent of drive is usually plenty, since drum transients already carry their own harmonic content and a heavy hand flattens exactly what made the hit punchy.

On a vocal, tube-style saturation at a low to moderate drive adds presence and a sense of proximity without the vocal sounding processed. Push past that point and sibilance and plosives start to distort along with everything else, since saturation reacts to whatever level reaches it.

On a bass, saturation is often doing real work rather than decoration: adding upper harmonics that let a bass line translate on small speakers that cannot reproduce its fundamental. A moderate drive on a bass bus is common precisely because the audible benefit, not the coloration, is the point.

On the master bus, the amount should be barely perceptible in isolation. Bypass the effect while listening: if you can hear a clear difference with the mix soloed, it is very likely too much once every other element is also playing.