Distortion is an audio effect that reshapes a waveform through a nonlinear transfer function, so the output is no longer a scaled copy of the input. In audio effects, the term covers the deliberate use of clipping, saturation and waveshaping to add new harmonic content, perceived loudness and texture to guitars, bass, drums, vocals and synthesizers.
The same physical process is a defect when it happens by accident in a recording chain, which is why engineers distinguish intentional distortion as a creative effect from unwanted distortion as a measured impairment. This page treats distortion as an effect, while noting where the two meanings meet.
How distortion reshapes a waveform
A linear device multiplies every sample by the same gain, so a sine wave stays a sine wave. A distortion stage applies a transfer curve that bends at higher levels, flattening or folding the peaks. Any change in the shape of a periodic wave adds energy at integer multiples of the fundamental, the harmonics that the ear hears as brightness, edge or fuzz.
The symmetry of the curve decides which harmonics appear. A curve that treats positive and negative halves identically produces mainly odd harmonics (third, fifth, seventh), which sound hollow and aggressive. An asymmetric curve, typical of a single-ended tube stage or a biased diode, adds even harmonics (second, fourth), which sound fuller and are often described as warm.
With more than one note, a nonlinear stage also creates intermodulation products at sum and difference frequencies. These are not harmonically related to the notes, which is why a heavily distorted major chord sounds muddy while power chords of root and fifth stay clear: their difference tones reinforce the root.
The table below lists the main curve shapes used in distortion processors and the character each one produces.
| Curve shape | Behavior at the threshold | Dominant harmonics | Typical character |
|---|---|---|---|
| Hard clipping | Abrupt flat ceiling | Strong odd series | Buzzy, aggressive fuzz |
| Soft clipping | Gradual rounding of peaks | Lower odd series | Smooth overdrive |
| Asymmetric clipping | Halves clip at different levels | Even and odd | Warm, tube-like |
| Wavefolding | Peaks fold back downward | Dense upper series | Metallic, synthetic |
| Bit and rate reduction | Stepped amplitude or time | Inharmonic aliasing | Lo-fi, digital grit |
Types of distortion effects
Practitioners group distortion effects by how much gain they apply before the nonlinearity and by the circuit or algorithm they imitate. The names are informal, but they describe audibly different results that musicians and mixing engineers choose between deliberately.
| Effect type | Drive amount | Mechanism | Usual sources |
|---|---|---|---|
| Saturation | Light, a few dB into the curve | Tape, transformer or tube modeling | Mix bus, vocals, drums |
| Overdrive | Moderate | Soft clipping, often with a mid boost | Electric guitar, organ |
| Distortion pedal | High | Hard clipping after heavy gain | Rock and metal guitar |
| Fuzz | Very high | Transistor stage driven into square wave | Guitar, bass, synths |
| Bitcrusher | Not gain based | Reduced bit depth and sample rate | Electronic drums, effects |
Saturation sits at the subtle end: total harmonic distortion of a few percent or less, often felt as density rather than heard as grit. Fuzz sits at the other extreme, turning almost any input into a near square wave with heavy sustain. Between them, overdrive and distortion pedals shape the tone with pre- and post-clipping filters that decide which frequencies are driven hardest.
Key parameters and how they interact
Most distortion processors expose the same core controls, and the result depends on how they are balanced against each other rather than on any single setting. The most important relationship is between input drive and output level, because distortion raises perceived loudness and can mislead a listener comparing settings.
| Parameter | What it changes | Typical range |
|---|---|---|
| Drive or gain | How far the signal is pushed into the curve | 0 to 40 dB or more |
| Tone or pre-filter | Which frequencies are clipped hardest | Tilt or band around 500 Hz to 3 kHz |
| Output level | Final level after clipping | Set to match bypass loudness |
| Mix or blend | Ratio of distorted to clean signal | 0 to 100 percent |
| Oversampling | Internal sample rate multiple | 2x to 16x |
| Bias or asymmetry | Balance of even and odd harmonics | Model dependent |
In digital processors, oversampling is essential. Clipping generates harmonics far above the audible band, and without a higher internal sample rate they fold back as aliasing, producing tones unrelated to the notes being played. Oversampling of 4x to 8x removes most audible aliasing at moderate drive, with higher factors reserved for extreme settings.
Distortion also acts as a fast compressor. Once the signal reaches the ceiling, louder input produces almost no louder output, so crest factor drops and sustain rises. This is why a heavily driven guitar seems to hold notes longer than a clean one.
Distortion compared with related effects
Distortion is the only common effect whose core purpose is to change the harmonic content of a sound. Its neighbors in audio effects work mostly with time or pitch, which makes it a natural partner for them rather than a substitute.
| Effect | Domain it works in | Adds new frequencies | Main purpose |
|---|---|---|---|
| Distortion | Amplitude, nonlinear | Yes, harmonics and intermodulation | Texture, sustain, loudness |
| Compression | Amplitude, slow linear gain | Only slightly at fast settings | Dynamic control |
| Chorus and flanger | Time, modulated delay | No, creates comb filtering | Width and motion |
| Pitch shifting | Frequency, transposition | Moves existing partials | Harmony, correction |
| Reverb and delay | Time, linear echoes | No | Space and depth |
Order in the chain matters. Distortion placed before delay and reverb keeps echoes clean and articulate. Placed after them, the echoes are distorted together with the dry signal and intermodulate, which produces a dense, washed-out wall of sound that some genres want and most mixes do not.
Common mistakes and how to avoid them
Distortion is easy to overuse because louder almost always sounds better in a quick comparison. Most problems come from too much drive, poor filtering around the clipping stage, or digital artifacts.
- Loudness bias: match output level to bypass before judging whether the effect improves the sound.
- Muddy low end: cut lows before the clipping stage, because bass drives intermodulation hardest.
- Fizzy highs: low-pass the output, often between 5 and 8 kHz for guitar tones.
- Aliasing: enable oversampling when the processor offers it, especially on bright sources.
- Lost dynamics: use the blend control for parallel distortion to keep transients from the clean path.
- Accidental clipping: unintended converter or bus clipping is a fault, not an effect, and should be fixed at the gain stage.
The simple explanation
Imagine pushing a sound so hard that it no longer fits through a narrow doorway. The top and bottom of the sound wave get squashed or cut off, and the sound that comes out the other side is a different shape. That change in shape adds extra tones above the original note, and those tones are what people hear as grit, buzz or warmth.
A little of it makes a voice or a drum feel thicker and more present. Heavy amounts of it turn a clean guitar into the roaring sound of rock music. The key idea is that distortion adds something new to a sound, while most other effects only move, repeat or blend what is already there.
How it is used in practice
A mixing engineer is working on a pop song whose bass guitar disappears on small speakers, because most of its energy sits below 100 Hz, where phones and laptops reproduce almost nothing. Instead of boosting the low end, the engineer duplicates the bass track and sends the copy through a distortion plugin.
The copy is high-pass filtered at about 300 Hz, driven with moderate soft clipping and 4x oversampling, then blended in roughly 10 dB below the clean bass. The distortion generates harmonics of each bass note in the 300 Hz to 2 kHz range, which small speakers can play. The listener’s ear infers the missing fundamental from these harmonics.
The result is a bass line that stays audible on a phone while the clean track still carries the deep low end on full-range monitors, without adding any extra low-frequency energy to the mix.