Tape saturation in plain English
Studio tape is a long strip of plastic coated with tiny magnetic particles. The record head turns the audio into a changing magnetic field, and as the tape slides past, the particles are left magnetised in a pattern that copies the sound. The playback head reads that pattern back.
At normal levels the pattern follows the audio closely. Turn the level up and there comes a point where the particles can't take much more magnetism, so each extra decibel going in gives you a little less coming out. That's magnetic saturation: the tape is getting close to holding as much magnetism as it can.
What matters for the sound is that tape gets there gradually. Digital audio has a hard ceiling at 0 dBFS, and anything above it is chopped off flat, which is clipping. Tape has no single line like that. The loudest moments get rounded off a little more as you push harder, so a level that would be a harsh digital over comes back from tape as a thicker, slightly softer version of the same sound. Engineers found that recording a little too hot often sounded better, and started doing it on purpose.
This page goes further into tape than the saturation lesson, which covers every type of saturation side by side.
What does tape saturation sound like?
People call it warm, or say it glues a mix together. Those words describe several separate changes that happen at once, and tape plug-ins usually give you a control for each.
The peaks get rounded
A transient is the short spike at the start of a drum hit or a plucked note. It's the loudest part of the sound, so it goes furthest into the curve and gets rounded off the most, while the quieter body of the note behind it passes through almost untouched. The gap between the peak and the average level, called the crest factor, gets smaller.
That's why tape is often described as a kind of compression. Strictly it isn't one, since nothing in it measures the level and turns it down over time; the curve changes the shape of each wave as it happens. The effect on a drum bus is similar, though: the hits feel denser and the kit sounds more like one instrument. Saturation vs compression covers where the two overlap.
Mostly odd harmonics
Rounding a wave changes its shape. Building that new shape takes new frequencies, called harmonics, which sit at whole-number multiples of the note itself, the fundamental. So a 110 Hz note picks up harmonics at 220, 330, 440 Hz and on up.
Tape's curve is close to symmetrical, bending the top and bottom of the wave by the same amount. A symmetrical curve only makes odd harmonics, at 3, 5 and 7 times the note, so the 3rd harmonic is the main thing tape adds. A single valve stage usually leans to one side and adds a strong 2nd harmonic as well, which is the main reason the two sound different. The even vs odd harmonics lesson explains why the shape of the curve decides this.
In practice a tape machine adds a little of the 2nd too, from its electronics and from imperfect bias, so most plug-ins aren't perfectly symmetrical. The lab's Character knob does the same job. Keep it under about 50% if you want something that behaves like tape.
A softer top end
Tape struggles more with high frequencies than with low ones. A tape recorder boosts the highs on the way onto tape and cuts them again on playback, which keeps hiss down, but it also means the high frequencies are the first to hit the limit. Push a tape machine hard and the cymbals and the S sounds in a vocal saturate first and come back a little duller than they went in.
This is the other half of what people mean by warmth. Saturation on its own adds harmonics and tends to make a sound brighter. Tape adds harmonics and takes some top end away at the same time, so it ends up thicker without getting much harsher. The lab's tape mode copies this: its top end starts to roll off at 16 kHz, and the roll-off creeps down towards 7 kHz as you add drive. That filter is gentle, so on the drums the new harmonics still win and the loop comes out slightly brighter, while the bass in step 6 comes out about a decibel darker.
Head bump
A tape machine's playback head doesn't read very long wavelengths evenly, so the frequency response at the bottom end has a small lift, called head bump, with a few smaller ripples below it. The wavelength on tape depends on how fast the tape moves, which means the bump moves with tape speed: run the tape twice as fast and it sits about an octave higher. It's usually a lift of a few decibels at most, somewhere in the bass, and it's part of why people say tape makes the low end sound bigger.
Head bump is an EQ change, so it happens at every level, even when the tape isn't saturating. The lab doesn't model it. If a tape plug-in makes your kick or bass feel louder, try switching its saturation off and leaving its EQ on if it has that option, and you'll often find the bump is doing most of the work. You can get a similar lift with a low shelf or a wide bell on an EQ.
Wow, flutter and hiss
No tape transport runs at an exactly steady speed. Slow drifts in speed are called wow and faster wobbles are called flutter, and both make the pitch move by a tiny amount. On a well-maintained studio machine it's very small, but it softens sustained notes in a way some people like. Tape also adds hiss, the noise of the particles themselves, which is why noise reduction systems such as Dolby A and Dolby SR were made.
Neither of these is saturation, and the lab models neither. Most tape plug-ins include them with their own controls. On your first few mixes, turn them off or down and add them back only if you want the sound of an old recording.
Reading the lab
The lab on this page starts on the tape curve. Three other curves sit next to it for comparison: a soft clipper, a valve curve and a hard clipper. Press 1 to 4 to switch between them while the loop plays.
The big screen shows the whole loop. The pale shape is the sound after the drive knob, before the curve touches it, and the coloured shape is what comes out. Under it, the close-up shows two milliseconds around the loudest hit, and the square screen is the transfer curve, with input level along the bottom and output up the side. For tape it's a gentle S-shape that bends further the more drive you add.
The bar graph runs a 110 Hz sine through the same settings and measures every harmonic that comes out, with odd ones in pink and even ones in amber. The spectrum next to it shows the loop itself, before and after, and it's the best place to watch the top end change. The Right now panel describes what your settings are doing, using numbers measured on the sound that's playing.
Auto gain is on, so the lab matches the loudness of the tape version to the original and you're comparing tone with tone. The tape mode here models two parts of a real machine, the soft curve and the loss of top end as you drive it. It leaves out head bump, wow and flutter and hiss, and its drive knob isn't calibrated to a real machine's levels, so treat the numbers as a guide to how tape behaves; they won't line up with the meters on a tape plug-in.
How hard can you hit tape?
Tape manufacturers publish a figure called maximum output level, or MOL. It's the level at which the 3rd harmonic reaches 3% of the signal, measured on a test tone. Three percent sounds tiny, and on a pure tone you can hear it, but on drums or a full mix you often can't.
Harmonic levels are usually written in decibels below the note, so it helps to know how that scale lines up with percentages.
| Below the note | Percentage | What you'd typically hear |
|---|---|---|
| −60 dB | 0.1% | Nothing on a full mix |
| −40 dB | 1% | A slight thickening on sustained notes |
| −30 dB | 3.2% | About a tape's maximum output level |
| −20 dB | 10% | Clear saturation on bass and vocals |
| −10 dB | 32% | Obvious distortion |
Now compare that with the lab. Put the sine through the tape mode with no drive at all and the 3rd harmonic comes out at about −30 dB, roughly 3%, because the test sine is close to full scale and the curve is already bending. At 12 dB of drive it's −16 dB, roughly 16%, well past anything a tape maker would call clean, and the 5th and 7th harmonics are climbing behind it at −26 and −35 dB.
Engineers used that headroom in different ways. Some set their machines to hotter operating levels so the tape was always working a little, and some recorded drums and bass into the curve on purpose while keeping vocals cleaner. For plug-ins, the lesson is that there's a range where tape adds weight without sounding distorted, and the top of that range depends on the sound. A drum loop can take far more than a held vocal note or a pad.
The controls on a tape plug-in
Tape plug-ins vary a lot. Some, such as Universal Audio's Studer A800 and Ampex ATR-102, copy specific machines closely, while Softube Tape and u-he Satin give you a broader set of controls. CHOW Tape Model is free and open source. The names change between them, but most have some version of these.
| Control | What it changes | Where to start |
|---|---|---|
| Input, drive or record level | How hard the sound hits the tape, so how much saturation you get | Low, then raise it until you hear the change on A/B |
| Output | Plain volume after the tape | Set it so bypassing doesn't change the level |
| Tape speed | Head bump frequency, top-end extension and hiss | 15 ips for more low end, 30 ips for a cleaner top end and less hiss |
| Tape type or formulation | How early the tape saturates and how it sounds when it does | The plug-in's default |
| Bias | Less bias is brighter with more distortion; more bias is darker and cleaner | The calibrated default, often marked 0 |
| Wow and flutter | Small, slow and fast pitch wobble | Off, unless you want an old or worn sound |
| Noise or hiss | Tape noise added under the signal | Off for mixing |
Bias is the strangest of these. Tape on its own records quiet sounds badly, so a tape machine mixes a very high, inaudible tone into the recording to get the particles moving, around 100 kHz or above on studio machines. How much of that tone you use moves the balance between distortion and top end. Engineers set it by recording a high test tone and adjusting until they reach the figure the tape manufacturer recommends; on a plug-in, leave it alone until everything else is set.
Tape speed is measured in inches per second (ips), and studio machines usually run at 15 or 30 ips, some at 7.5. Faster tape holds high frequencies better and has less hiss.
How to use tape saturation, step by step
- Start with the extras offTurn off hiss and wow and flutter, and leave bias and tape type on their defaults, so the drive is the only thing you're judging.
- Raise the drive slowlyBring it up until you can just hear a change when you bypass the plug-in. On drums, listen to the front of each hit. On bass, turn your monitors down and listen to whether you can still follow the low notes.
- Match the levelUse the output control so the plug-in is no louder than the bypassed signal. Almost every saturator gets louder as you drive it.
- Choose the speed by the low endSwitch between 15 and 30 ips and listen to the kick and bass. Pick the one where the low end sits better against the rest of the mix.
- Check the top endLook at a spectrum analyser or listen to the cymbals. If they've gone dull, back off the drive or try a faster speed before reaching for an EQ.
- Compare in contextBypass it with the whole mix playing. If you can't say what improved, turn the drive down or take it out.
Where to use tape saturation
Drums are the classic use. Tape rounds off the tallest part of each hit, so the kit sounds denser and you can raise its level without the peaks poking out. On a drum bus it often does some of the job you'd otherwise give a bus compressor. If the snare starts to lose its crack, blend the tape in parallel so the clean hit stays on top, which parallel saturation covers in full.
On bass, the 3rd and 5th harmonics of a 55 Hz note sit at 165 and 275 Hz, which small speakers can play even though they can't reproduce 55 Hz itself. Your ear rebuilds the pitch from the harmonics, so the bass line stays audible on small speakers, and tape does it without the brightness a clipper would add.
Vocals take less. A little tape thickens a vocal and lowers its peaks, but the S and T sounds saturate early, because they're high frequencies and tape runs out of room for those first. If the sibilance turns spitty, put a de-esser before the tape plug-in or lower the drive.
On a mix bus, use tape lightly. With a few decibels of drive the loudest drum hits come down a little and everything else is barely touched, so the mix feels more connected. Put it before the bus compressor and the limiter, and keep it subtle enough that you have to A/B to be sure it's there.
| Source | Drive | Mix | Listen for |
|---|---|---|---|
| Drum bus | 10–20 dB | 50–100% | Denser hits that still have a clear front |
| Kick | 8–16 dB | 100% | A rounder, heavier hit without fuzz in the tail |
| Bass | 10–18 dB | 100% | Low notes you can still follow at quiet levels |
| Lead vocal | 3–8 dB | 100% | More presence with S sounds that stay smooth |
| Mix bus | 2–6 dB | 100% | A change you only just notice on A/B |
Tape compared with other saturation
Against a soft clipper, tape's curve is similar in shape. The difference on a full sound comes from the top end: a soft clipper adds harmonics and leaves the highs alone, so it comes out brighter, while tape takes some of the highs away as it goes. In the lab the gap is small, a fraction of a decibel above 3 kHz on the drums, because its tape mode rolls the top off gently. On a real machine or a plug-in, slower tape speeds lose more top end.
Against a valve, the difference is mostly in the harmonics. A single valve stage leans to one side and adds a strong 2nd harmonic, an octave above the note, which sounds full and close; tape adds mostly the 3rd, an octave and a fifth up, which sounds firmer. Tape vs valve saturation compares the two in detail, and valve saturation covers the valve on its own.
A hard clipper is much further away. Its flat tops add a long row of high harmonics and some aliasing, and the sound turns buzzy, which is the opposite of what people reach for tape to do.
Common tape saturation mistakes
Tape on every track
Each instance takes a little top end off. One on the drums is fine, but twenty of them add up and the whole mix ends up dull and cloudy. Choose the few tracks that need it, usually drums, bass and the mix bus, and leave the rest clean.
Judging it louder
Drive raises the level, and a louder version nearly always sounds better for the first few seconds. Match the output to the bypassed level every time you change the drive.
Leaving hiss and wobble on by default
Some plug-ins load with noise and wow and flutter switched on, and on one track it's hard to hear. Across a whole session the hiss stacks up, and the wobble smears the tuning of sustained parts.
Blaming the saturation for the low end
When a tape plug-in makes the bass feel bigger, it's often the head bump, which is EQ. If the low end gets boomy, try the other tape speed, which moves the bump, before you turn down the drive.
Training your ears to hear tape
Tape is subtle at sensible settings, so it helps to learn its sound at stronger ones first. The quiz under the lab plays the current loop through tape, a soft clipper or a hard clipper with the screens covered, at the same loudness as the original, and asks which one it was. Tape and the soft clipper are the hardest pair to tell apart; listen to the cymbals and the snare, which come out a little darker through tape. The tape round also rolls the top end off from 5 kHz, a stronger loss than the lab's tape has on its own, so that the darkness shows on every sound.
Questions people ask
Does tape saturation add even or odd harmonics?
Mostly odd. Tape's curve is close to symmetrical, so the 3rd harmonic leads, followed by the 5th. A tape machine's electronics and the way the tape is biased can add a small amount of the 2nd, which is why some plug-ins let you dial in a little even content.
Is tape saturation the same as tape delay?
No. Tape delay uses a loop of tape with a record head and one or more playback heads to make echoes, and the repeats pick up saturation and wobble on the way. Tape saturation is the distortion on its own, with no echo.
Do I need a real tape machine to get the sound?
No. Good tape plug-ins model the curve, the top-end loss and head bump, and most add hiss and wobble that you can switch off. A real machine needs regular cleaning and calibration, and reels of studio tape are expensive.
Where should tape saturation go in the chain?
On a single track, put it after any EQ cuts and before the compressor, or at the end of the chain if you only want a touch of colour. On a mix bus it usually goes early, before the bus compressor and well before the limiter.
Why does my tape plug-in make things duller?
Because tape loses high frequencies, and loses more of them as you drive it harder. Try a faster tape speed, less drive or a different tape type, or add a gentle high shelf after the plug-in. Check you aren't stacking the same plug-in on every track, because the dulling adds up.
What does 15 ips or 30 ips mean?
Inches per second, the speed the tape moves past the heads. At 30 ips the machine records high frequencies more cleanly with less hiss, and head bump moves up in frequency. At 15 ips the top end rolls off sooner and the bump sits lower, so more of the bass gets the lift.