Short answer: autotune, also called pitch correction, moves a sung or played note to the nearest correct pitch in a key and scale you choose. Nudge it gently and it fixes small tuning slips so a take still sounds like a person singing. Set it to snap instantly and you get the hard, robotic effect instead.
The controls are broadly the same from one tool to the next: the key and scale it corrects to, the retune speed that sets how fast it snaps, and the amount or strength. It works best on one note at a time, and it cannot rescue a poor performance. None of this is tied to one product.
Autotune is a pitch-correction tool. It moves a note that is a little sharp or flat onto the nearest correct pitch in a key and scale you set, so the tuning lands where the music wants it. The word started as a product name and has become the common term for pitch correction in general, and most tools do the same core job.
There are two very different ways people use it. The first is transparent correction: you set it to work gently, so it only nudges the small tuning errors in a take and leaves the natural movement of the voice alone. Done well, a listener never notices it is there. The second is the hard, robotic effect, where the pitch snaps from note to note with no glide in between. That is a deliberate creative sound, not a fault, and we cover how it is made further down.
It works best on one note at a time, such as a lead vocal or a single-line instrument, because it has a single pitch to track and correct. On a full chord or a whole mix it has several notes at once to make sense of, which is a harder job and outside what most everyday pitch correction is meant for.
Pitch correction can only move a note to a pitch that is already in your chosen scale. It does not write the melody, choose the notes or add the feeling. It tidies the tuning of a performance you have already sung or played, which is why the take underneath still matters most.
It listens to the incoming note, works out its pitch, compares that to the notes allowed by your chosen key and scale, and moves it to the nearest allowed note. The retune speed then decides how quickly it makes that move, from a slow glide to an instant snap. Get the key and scale right first, because that is the map it corrects against.
Three controls do most of the work, and they are broadly the same whatever tool you open.
| Control | What it does | A place to start |
|---|---|---|
| Key and scale | The set of notes the tool is allowed to snap to. Set it to the key of your song so a note is never pulled to a pitch that does not belong. | Match it to the song's key |
| Retune speed | How fast a note moves to its target pitch once it is detected. Slow lets it glide in naturally; fast makes it snap instantly, which is the robotic sound. | Slow for natural, fast for the effect |
| Amount or strength | How far toward the target the note is moved. Full pulls it all the way; less than full keeps some of the original tuning and character. | Less than full for a natural result |
Setting the key and scale matters more than anything else here. If you tell the tool the wrong key it will happily pull good notes to the wrong pitch, so it is worth knowing the key of your song before you start. If keys and scales are new to you, our guide on what scales and keys are explains them in plain terms.
There are also two ways of driving it. Automatic mode corrects in real time as the track plays, snapping everything to the scale you set, which is quick and hands-off. Graphical or manual mode shows the notes on a grid and lets you move or redraw each one by hand, which is slower but gives you precise, note-by-note control, so you can fix only the notes that are actually off and leave the rest alone.
Set the retune speed to its fastest. At the fastest setting the pitch jumps straight to each target note with no glide between them, and that instant snapping from note to note is the hard, robotic sound. It is sometimes called the T-Pain effect, after the way it was used on a run of pop and hip-hop records.
A few things make it stronger. Set the strength to full so every note is pulled all the way to pitch, and set the key and scale to match the song so it snaps to musical notes rather than random ones. The effect shows up most on held notes and on slides between notes, where you can clearly hear the pitch step instead of slide. Sung phrases that jump across big intervals tend to show it off best.
This is a creative choice, not a repair. The same tool that hides its work on a slow setting becomes an obvious, up-front effect on a fast one, so which one you want depends entirely on the sound you are after.
The line between invisible correction and the robotic effect is mostly just the retune speed. Slow and it glides in unnoticed; fast and it snaps hard enough to hear. If you are chasing the effect, that is the first control to reach for, with the strength turned up behind it.
Use a slower retune speed, correct only the notes that actually drift, and start from a good take. Transparent tuning is meant to fix small errors while leaving the natural pitch movement of the voice intact, so the lighter your touch, the more it sounds like a person and not a machine.
Slow the retune speed. A fast speed snaps every note into place and flattens the small, natural slides and the vibrato that make a voice sound human. A slower speed corrects the pitch while letting that movement through, which is usually what you want for a natural result.
Fix the notes that are off, not all of them. A whole take run through automatic correction at full strength often sounds lifeless, because it clamps the good notes along with the bad. Graphical or manual mode lets you move only the few notes that are actually out and leave the rest alone, which almost always sounds more natural.
Start from a good performance. Pitch correction moves tuning; it does nothing for timing, tone or feel. It cannot turn a poor take into a good one, and pushing it hard to rescue one usually just adds an artificial, warbly quality on top of the original problems. A confident take that is only slightly out is the easiest thing to make sound right. Tuning is really just one early step in the larger job of mixing a vocal.
Most natural-sounding tuning uses a gentle speed, less than full strength and correction only where it is needed. If a tuned vocal sounds artificial and you did not want it to, the fix is almost always to use less: slower, weaker, and on fewer notes.
Pitch correction is for the voice on top. The instruments underneath do not drift the way a singer can, and a part that is written and played in the right key is in tune to begin with. That is the half of the record we make.
We do not make a vocal tuner or a pitch-correction plugin, so this is not that. What we make is instruments. The Collection is our three of them together, ARGISH, SILT and REHEAT, with three separate licence keys. They play in whatever key you set, so they are in tune by nature and the backing under your vocal never needs correcting. ARGISH is a self-playing drone synth, SILT is a tape-loop instrument, and REHEAT writes acid lines. Every one ships as AU, VST3, AAX and a standalone app on macOS, signed and notarized, and a VST3 on Windows.
You can hear one free in your browser first, no install and no account, then decide.
We write these when there is something worth writing down. One email when a new one lands or a new Tunary instrument ships. No newsletter, no schedule.