1. Learn
    2. Saturation and distortion
    3. Saturation

    What is saturation?

    Saturation is gentle distortion. When you push a sound harder than tape or a valve circuit can cleanly handle, its peaks get rounded off and it picks up new frequencies called harmonics, at whole-number multiples of each note. A little makes a sound thicker and easier to hear; a lot turns it into obvious distortion.

    What it does
    Rounds off peaks and adds harmonics above each note
    The main controls
    Drive, character, tone, mix and output
    Used on
    Bass, drums, vocals and the mix bus
    The usual mistake
    Judging it by how much louder it makes things
    SAT·1Saturation lab
    Whole loopInput after driveOutput
    One peak, sample by sample
    Transfer curve
    The line bends off the diagonal near the top

    Tape

    Tape with 4.0 dB of drive. The tape curve is almost symmetrical, so most of what it adds is odd harmonics, led by the 3rd at −24 dB, and its slight lean adds 2 even ones. Tape also loses top end as you push it, so the loop comes out 0.5 dB darker above 3 kHz than the original. Auto gain is turning the result up by 2.8 dB to keep it as loud as the original.
    5Added, 2 even
    −0.5dB above 3k
    14.6Crest, was 17
    Harmonics on a 110 Hz sineOddEvenOther
    Spectrum of the loopOriginalOutput
    Drive+4.0 dB
    Character30%
    Bits8.0
    Rate6.0 kHz
    ToneOpen
    Mix100%
    Output0.0 dB

    Follow the steps. Each one sets the controls for you and tells you what to listen for.

    Step 1 of 8
    Start with a little tapePress play. The bass is going through the tape curve with 4 dB of drive, which only bends the loudest peaks. Press A/B to hear the original at the same loudness.
    Train your earsThe lab plays the current sound through one of three characters with the screens covered, at the same loudness as the original, and you name it. The bass and the chords make it easiest to hear.

    Loop from Flow, a Sounds Like You pack.

    Saturation in plain English

    Every piece of audio gear has a limit to how much signal it can pass cleanly. Below that limit, double the input and you get double the output. Push past it and the output can't keep up: the loudest parts of the wave get squashed while the quiet parts carry on as before. That squashing changes the shape of the waveform, and a changed shape always means new frequencies.

    The word comes from magnetic tape, which is saturated when it can't hold any more magnetism. Valves and transformers run out of room in a similar way, so the peaks run into a soft wall instead of a hard one. Recording a little too hot onto tape was never a disaster the way a digital over is. It sounded thicker, and engineers started doing it on purpose.

    A saturation plug-in copies that behaviour with a transfer curve, a rule that says what level comes out for every level that goes in. A straight diagonal line changes nothing. A line that bends over at the top is saturation, and the harder you drive the sound into the bend, the more it changes. Every mode in the lab above is a different curve, and the current one is drawn on the square screen.

    What are harmonics?

    Play the A below middle C and its fundamental, the frequency you hear as the pitch, is 220 Hz. The A an octave below that, the one the lab's sine sits on, is 110 Hz. Harmonics are frequencies at whole-number multiples of that frequency, so a 110 Hz note has them at 220 Hz, 330 Hz and on up. Real instruments already have them, which is why a piano and a flute playing the same note sound different. Saturation adds more.

    Because harmonics line up with the note, they tend to blend in. The 2nd is exactly an octave up and the 4th two octaves up, while the 3rd, an octave and a fifth above, is still a consonant interval. Further up the series the intervals get less friendly; the 7th harmonic lands a little flat of a minor seventh, two octaves above the note.

    The first harmonics of a 110 Hz note
    HarmonicFrequencyNearest noteInterval above the note
    1st110 HzA2The note itself
    2nd220 HzA3One octave
    3rd330 HzE4An octave and a fifth
    4th440 HzA4Two octaves
    5th550 HzClose to C♯5Two octaves and a major third
    6th660 HzE5Two octaves and a fifth
    7th770 HzA little flat of G5Two octaves and a flat minor seventh

    Odd and even harmonics

    A curve that treats the top and bottom of the wave the same way, called a symmetrical curve, only ever adds odd harmonics: the 3rd, 5th, 7th and so on. A curve that leans to one side, bending the top harder than the bottom, adds even harmonics too.

    Odd harmonics on their own tend to sound hollow or edgy. Even harmonics, led by the octave, sound fuller. Most real saturation has some of both, and the balance between them is a big part of why tape, valves and transistors sound different from each other. The lab's harmonic screen colours them apart, odd in pink and even in amber, so you can watch the balance shift as you change modes. There's more in even vs odd harmonics.

    The dashed line is one cycle of a sine wave and the solid line is the same cycle after each curve, scaled to the same height. Only the valve treats the top and bottom of the wave differently, and that lopsided shape is where even harmonics come from.

    Reading the lab

    The cards along the top pick the curve, and while the lab has focus, keys 1 to 7 switch between them. The big screen underneath shows the whole loop. Its pale shape is the sound after the drive control, before the curve touches it, so it grows as you turn the drive up; the coloured shape is what actually comes out. When the coloured shape stays the same size while the pale one keeps growing, the curve is flattening everything you push into it.

    The first small screen zooms into two milliseconds around the loudest peak in the loop, about 90 samples at 44.1 kHz, with a dot for every sample. That's where you see a peak get rounded, cut flat, folded over or turned into steps. Next to it is the transfer curve for the current mode, with dashed lines marking where the loop's loudest peaks land on it.

    Underneath, two screens show frequencies. The harmonic bars come from running a pure 110 Hz sine through exactly the settings you've chosen and measuring what comes out, so they show the added harmonics with nothing from the music in the way. The spectrum next to them shows the real loop. Its pale shape is the original and the coloured line is the result, and when you press play it follows the music live, so you can watch new frequencies appear above the bass notes as you turn the drive up.

    The panel on the right explains what your settings are doing, using numbers measured from those screens, and it talks about the control you moved last. Auto gain is on by default: the lab measures the loop and keeps the output at the same loudness as the original, so drive changes the tone and leaves the level alone. A/B flips to the original, also at matched loudness.

    Any sound works. Upload a file, or press Record and play something into your microphone for eight seconds, and it loops through the lab in place of ours. It stays on your device.

    The controls

    Saturation plug-ins vary more than compressors do, but nearly all of them have some version of these. Ableton Live's Saturator, for example, has Drive, a choice of curve shapes, Output and Dry/Wet.

    Drive

    Drive is input gain before the curve. It decides how much of the wave reaches the part of the curve that bends. On the lab's soft clipper, 12 dB of drive puts the 3rd harmonic of a sine about 12 dB below the note and the 5th about 20 dB below. Every bit of extra drive pushes more of each cycle into the bend, so the harmonics climb and more of them clear the −60 dB line on the harmonic screen.

    Drive also raises the level going out, which is why most saturators have an output control next to it. Switch Auto gain off and add 20 dB of drive to the tape mode, and the loop gets much louder. The louder version will win any quick comparison whether it sounds better or not.

    Character

    Character is how far the curve leans to one side, and so how many even harmonics it adds. At 0% the soft and hard clippers are symmetrical and produce odd harmonics only. Turn it to 60% on the soft clipper with 12 dB of drive and the 2nd harmonic comes up from nothing to about 22 dB below the note. Plug-ins often call it asymmetry or bias. On the lab's wavefolder it also moves where the folds land.

    Tone

    Saturation adds harmonics all the way up the spectrum, and the highest ones can sound fizzy, especially from a hard clipper. A tone control is a filter after the curve that rolls them off. In the lab it's a low-pass filter that cuts at 12 dB per octave above the frequency you set, so with it at 5 kHz, a harmonic at 10 kHz comes out about 12 dB quieter than it would with the control open. The top end figure in the panel shows how much brighter or darker the loop is than the original once the overall level is taken out of the comparison.

    Mix

    The mix control blends the saturated sound with the clean one. At 100% you hear only the saturated version. At 35%, the clean sound sits on top and the saturated copy fills in underneath, so you get the density without losing the original attack. That's parallel saturation, which has its own section further down.

    Output and Auto gain

    Output is plain volume after everything else. Use it to bring the saturated sound back to the level of the original. Auto gain in the lab does that for you by measuring the average level of the whole loop, and Output then trims on top of it.

    Tape, valve and transistor saturation

    The sound of saturation depends on the shape of the curve, and every kind of hardware bends it differently. Plug-ins that copy hardware are mostly copying those curves, along with the filtering that comes with them.

    Tape

    Tape saturates gradually. Its curve is close to symmetrical, so most of what it adds is odd harmonics, mainly the 3rd; the maximum output level printed on a tape's spec sheet is the level where that 3rd harmonic reaches 3% of the signal. Tape also struggles to hold high frequencies at high levels, so pushing it softens the top end a little. That's tape warmth. Many machines also have a small bass lift called head bump, which this lab doesn't model; the tape saturation lesson covers it.

    Valve

    A single valve stage, such as a triode in a preamp, has a curve that leans to one side. It adds a strong 2nd harmonic even at modest levels, and because that harmonic sits exactly an octave above the note, it thickens the sound without adding edge. Drive it harder and odd harmonics come in as well. Push-pull valve amplifiers, which use two valves working on opposite halves of the wave, cancel out much of the even content. In the lab, with the character at 60%, the valve mode stays led by the 2nd harmonic until around 20 dB of drive. Americans call valves tubes, so you'll find the same thing under tube saturation.

    Transistors and clipping

    Transistor and op-amp circuits usually stay clean for longer and then run out of room more abruptly. That's closer to the lab's soft and hard clippers: mostly odd harmonics, and the sharper the corner of the curve, the further up the spectrum they reach. A hard clipper cuts every peak to a flat line, and because a sharp corner can only be built from a long row of high harmonics, it sounds buzzy on sustained notes. Clipping has its own lesson, and soft clipping vs hard clipping vs limiting compares the two clippers with a limiter.

    Wavefolding, bitcrushing and downsampling

    The last three cards are digital effects that come from synthesisers and early samplers more than from studio gear. A wavefolder flips peaks back down instead of flattening them, so a sine turns into a shape with extra wiggles in every cycle and a hollow, metallic ring; Ableton's Saturator has a similar curve called Sinoid Fold. A bitcrusher forces the wave onto fewer levels. Every bit you remove halves the number of levels and costs about 6 dB of clean range, so at 5 bits the quiet parts of a loop turn into grainy noise. A downsampler holds each sample for longer, which leaves out-of-tune tones above the music. Ableton's Redux and Logic's Bitcrusher do both of those last two jobs. See wavefolding, bitcrushing and sample rate reduction.

    Saturation vs distortion

    Saturation and distortion are the same process. Both change the shape of the wave and add harmonics, and the names describe how far you've gone. Saturation usually means the gentle end, where the new harmonics sit low in level and blend with the sound, so you'd describe the result as thicker or warmer. Distortion means you can clearly hear that the sound has been changed, the way an overdriven guitar amp sounds nothing like a clean one.

    You can cross from one to the other on a single control. Put the lab on tape with the bass loop and sweep the drive from 0 to 36 dB: somewhere past 20 dB it stops sounding like a warmer bass and starts sounding like a distorted one. Where that line falls depends on the sound, so there's no fixed setting for either. Saturation vs distortion goes into the comparison in more detail.

    Parallel saturation

    Parallel saturation means mixing a heavily saturated copy of a sound underneath the clean original. The clean signal keeps the attack and the detail, and the saturated copy adds harmonics and density without the original being visibly squashed. It's the same idea as parallel compression, and it works especially well on drums, where heavy saturation at 100% would round off every transient you wanted to keep.

    If your saturator has a mix knob, start with the drive far higher than sounds good on its own, then bring the mix down to between 20% and 40%. Without a mix knob, put the saturator on a send or a duplicate track and fade it in. Check the result with A/B, because parallel saturation also makes things louder. The parallel saturation lesson covers routing it in a DAW.

    Where to use saturation

    Bass

    Bass is where saturation does the most good. A 55 Hz bass note is hard to hear on phone and laptop speakers, which can barely reproduce anything that low. Add harmonics at 110 Hz, 165 Hz and 220 Hz and the bass becomes audible on those speakers, because your ear works out the pitch of a note from its harmonics even when the note's lowest frequency is missing. Valve and tape modes are the usual choice. If the low end starts to sound woolly, keep the very bottom clean by saturating a copy, filtering its lows out and mixing it back in.

    Drums

    On drums, saturation rounds the tallest part of each hit, which lowers the crest factor and makes them feel denser and louder at the same peak level. Tape is a common choice on a drum bus, and a soft clipper on the snare gives it more aggression. Parallel saturation is the safest way to start, since the clean drums keep their snap on top.

    Vocals

    A small amount of valve or tape saturation helps a vocal sit forward in a mix, because the added harmonics land in the upper mids, where hearing is most sensitive. Keep it gentle and listen to the S and T sounds, which turn harsh quickly when saturated. A de-esser before the saturator, or the tone control pulled down to around 8 kHz, keeps them in check.

    The mix bus

    On a whole mix, saturation can make the parts feel like they were recorded together. Use a decibel or two of drive and a tape or valve curve. If you can clearly hear it working on an A/B at matched loudness, it's too much for a mix bus, because whatever it adds lands on every instrument at once.

    Starting settings for common sounds

    Drive amounts depend on the plug-in and on how loud the sound is going in, so treat these as starting points in this lab. Your ears and the A/B switch have the final say.

    Saturation starting points in the lab
    SourceModeDriveMixListen for
    BassValve or tape10–20 dB50–100%The bass still audible at low volume on small speakers
    KickTape or soft clip6–14 dB100%A rounder, heavier hit without a fuzzy tail
    Drum busValve or tape16–28 dB20–40%More weight with the snap still on top
    Lead vocalValve or tape4–10 dB30–60%Presence without harsh S sounds
    Synth or guitarHard clip or wavefolder12–30 dB100%Edge and sustain, with the tone control taming fizz
    Mix busTape2–6 dB100%A difference you only just notice on A/B

    How to add saturation, step by step

    1. Pick the curve for the jobTape or valve for warmth and weight, a soft clipper for density, and a hard clipper or wavefolder when you want the distortion to be heard.
    2. Turn the drive up past the point you wantGo until it clearly sounds like too much. What the saturation adds is easier to hear when there's plenty of it, and backing off from there is quicker than creeping up.
    3. Back the drive offBring it down until the sound is thicker without being obviously distorted, or to the point where the distortion sounds right if that's the goal.
    4. Shape the top endIf the result is fizzy, lower the tone control or put an EQ after the saturator and roll off above 6 to 10 kHz.
    5. Blend itOn drums, or anywhere you want to keep the attack, pull the mix down until the original transients come back.
    6. Match the level and compareSet the output so the saturated and bypassed versions are equally loud, then switch between them. Keep it only if you prefer it at the same level.

    Aliasing and oversampling

    Digital saturation can add harmonics that are higher than the sample rate can store. Anything above half the sample rate, which is 22.05 kHz at 44.1 kHz, folds back down to a new frequency that has nothing to do with the note, and these out-of-tune tones are called aliasing. You can see them in the lab as the grey bar on the harmonic screen: almost nothing on tape, a little on a hard clipper driven hard, and a lot on the downsampler, which lowers the point where folding starts. Many plug-ins offer oversampling, which runs the curve at two, four or more times the sample rate and filters the extra harmonics out before coming back down. Turn it on for hard clipping and heavy drive, especially on high notes. The aliasing and oversampling lesson shows it in detail.

    Common saturation mistakes

    Choosing it by loudness

    Saturation nearly always makes a sound louder, and louder sounds better at first. Switch Auto gain off in the lab and the coach panel tells you how many decibels the saturated loop has gained. In your DAW, match the output to the bypassed level before you decide anything.

    Saturating everything a little

    One saturator rarely causes problems. Ten of them, each adding a few harmonics in the same upper mid range, add up to a mix that sounds crowded and tiring. Pick the parts that need help to be heard and leave the rest clean.

    Losing the low end

    Heavy saturation on a sub-bass flattens its long waves into something closer to a square wave, and the result can sound fuzzy and less deep. If the bottom loses weight, use the filtered copy from the bass section above, or switch to a gentler mode.

    Mistaking aliasing for saturation

    When a saturated high synth line sounds grainy and slightly out of tune, that's often aliasing. Turn on the plug-in's oversampling, or use a softer curve.

    Training your ears to hear saturation

    The differences between kinds of saturation are easiest to hear on sustained notes. The quiz under the lab plays the current sound through tape, valve or a hard clipper at the same loudness as the original, with the screens covered, and asks which one it is. Try it on the bass first. Tape tends to sound the darkest, valve the fullest, and the hard clipper puts a buzz on every note that the other two don't. Eight rounds take a few minutes, and your best score is saved in your browser.

    Questions people ask

    What is the difference between saturation and distortion?

    They're the same process at different strengths. Saturation is the gentle end, where the peaks are rounded and the new harmonics blend with the sound. Distortion is the same thing pushed until the harmonics are obvious and the sound clearly changes character.

    Does saturation make things louder?

    It makes them sound louder at the same peak level, because rounding the peaks off lets the average level come up. Many saturation plug-ins also get louder as you add drive, so turn the output down to match before you judge it.

    Should saturation go before or after compression?

    Either works. Saturation before a compressor gives it a denser sound to work on and can sound more aggressive. After the compressor, it reacts to a more even level, so the amount of colour stays more consistent from note to note. Try both and compare at matched loudness.

    Is saturation the same as warmth?

    Warmth usually means extra low harmonics, the 2nd and 3rd, plus a slightly softer top end. Tape adds mostly the 3rd and rounds the top off as it's pushed, and a single valve stage adds a strong 2nd. A hard clipper adds lots of high harmonics instead and sounds brighter, so plenty of saturation is anything but warm.

    Can I saturate the whole mix?

    Yes, lightly. A tape or valve plug-in on the mix bus with a decibel or two of drive can make the parts feel more connected. If you can hear it working on a quick A/B at matched loudness, it's probably too much for a mix bus.

    Why does my saturation sound harsh?

    Usually too many high harmonics. Use a softer curve such as tape, lower the drive, or roll off the top with a tone control or an EQ after the saturator. Turning on oversampling can help too, because harshness from aliasing sounds out of tune.

    Keep going

    Zac Dobos

    Zac runs Sounds Like You, which makes sample packs, preset packs and the Murmur granular synth plug-in. The lessons in Learn are free, and every one has a working version of the effect in it.