Learning how to use autotune comes down to three settings: tell the plugin what key and scale the song is in, pick between transparent correction or the hard “T-Pain” effect with the retune-speed control, and then fix any notes the plugin snaps to the wrong pitch. Get those right and autotune becomes an invisible polish on a good vocal take rather than a robotic gimmick. This guide walks through the whole workflow — from loading the plugin to catching the mistakes that make tuned vocals sound worse than the raw take.
Violet Recording is reader-supported — we may earn a commission from links on this page, at no extra cost to you.
What autotune actually does to a vocal
Autotune is a pitch-correction effect. It listens to the incoming vocal, works out the pitch of each note, and nudges that pitch toward the nearest “correct” note in the scale you’ve chosen. A singer who lands slightly flat on a sustained note gets pushed up to the true pitch; a note that drifts sharp gets pulled down. The listener hears a vocal that sits perfectly in tune with the track.
The term “autotune” is used generically now, but it started as a specific product — Antares Auto-Tune. Today the same job is done by many plugins: Antares Auto-Tune, Celemony Melodyne, Waves Tune, and your DAW’s built-in pitch tools all correct pitch, just with different interfaces and philosophies. Everything below applies whichever one you own, because the core controls are the same across the board.
How to use autotune: the basic workflow
Insert the plugin as the first effect on your vocal track, before any EQ, compression or reverb. Pitch correction wants the cleanest possible signal to analyse — reverb tails and heavy compression confuse the pitch-detection, so correct first and colour afterwards. Then follow these steps in order:
- Set the key and scale. Tell the plugin the song is in, say, A minor. Now it only snaps notes to the pitches that belong in A minor, so it can’t accidentally shove your vocal onto a note that clashes with the chords.
- Choose your reference — chromatic vs. scale. Chromatic mode allows all twelve notes and makes the gentlest moves. Locking to the actual key is more aggressive but keeps stray notes musical. Start on the correct key; drop to chromatic if the correction sounds forced.
- Set the retune speed. This is the single most important control — it decides whether the effect is invisible or obvious (more on it below).
- Play the vocal back against the track and listen for notes that snapped to the wrong pitch, then fix them by hand.
That’s the entire loop. Everything else is refinement.
Retune speed: the difference between natural and robotic
Retune speed controls how quickly the plugin drags a note to the correct pitch. It is the dial that separates a subtle fix from the hard, gliding, “digital” sound people associate with modern pop and rap.
Slow retune speed (higher millisecond values) lets the voice ease into pitch. Natural vibrato, scoops into notes, and expressive slides survive, because the plugin doesn’t react fast enough to flatten them. This is what you want for transparent correction on a pop, rock, folk or acoustic vocal — the goal is a listener who never suspects autotune was used.
Fast retune speed (near zero) snaps every note instantly to the grid. It kills the natural glide between pitches, producing the stepped, quantised sound — the deliberate “T-Pain effect” heard across hip-hop, trap and hyperpop. If you want the effect as a creative choice, set retune speed to its fastest, lock to the song key, and make sure the singer commits fully to each note.
The mistake beginners make is leaving retune speed fast when they actually wanted transparency. If your corrected vocal sounds robotic and you didn’t intend it to, slow the retune speed down until the movement disappears.
Automatic mode vs. graphical (manual) mode
Most serious pitch plugins offer two ways of working. Automatic mode corrects the whole track in real time using your key and retune-speed settings — fast, and perfect for a live vocal, a rough mix, or the hard-tuned effect. Graphical or manual mode (Auto-Tune’s Graph mode, or Melodyne’s whole approach) shows every note as a blob on a piano-roll grid, and you move, straighten and re-time each one by hand.
Graphical mode is more work but far more musical. You can correct a single flat word while leaving the rest of a beautiful take untouched, preserve intentional vibrato, and fix pitch without the collateral damage automatic mode sometimes causes. For a lead vocal you care about, comp your best take first, then tune it note by note in graphical mode. For backing vocals and quick jobs, automatic mode is usually fine.
Common mistakes that make autotune sound worse
Over-correcting is the biggest one. If the raw take is already close, heavy tuning removes the tiny pitch variations that make a voice sound human, leaving it sterile. Correct only what needs correcting.
Getting the key wrong is the second. Set the plugin to the wrong scale and it will confidently tune notes to pitches that fight the chords — the vocal sounds “in tune” with itself but wrong against the track. Always confirm the song’s key first.
Feeding the plugin a bad signal is the third. Autotune fixes pitch, not tuning drift caused by sibilance, breaths, or two singers bleeding onto one track. Comp a clean take, and consider putting a de-esser or gentle gate after the correction, not before it.
Finally, remember that autotune can’t fix timing, tone or a poor performance. A confident, in-the-ballpark take tuned lightly beats a shaky take rescued by aggressive correction every time. Track well first — our guide to mixing vocals covers the EQ and compression that come after tuning.
Which plugin should you use, and where to get one
Antares Auto-Tune remains the industry standard, especially for the classic hard-tuned effect, and its real-time mode is built for live use. Celemony Melodyne is the choice when you want surgical, natural correction and the ability to edit pitch, timing and even individual notes inside a chord — it’s a different way of working, closer to editing than to a live effect. Waves Tune and your DAW’s stock pitch tool (Logic’s Flex Pitch, Cubase’s VariAudio, Studio One’s Melodyne integration) cover most jobs without extra spend.
If you’re buying, you can compare the main paid options and watch for sales at a store like Plugin Boutique, which carries Antares, Celemony and Waves alongside frequent bundle discounts. For a deeper breakdown of the paid field, see our roundup of the best autotune plugins.
The best free autotune plugins worth trying
You don’t have to spend anything to start. A few genuinely useful free pitch-correction plugins:
- GSnap (GVST) — the long-standing free classic on Windows. It does clean automatic correction and, pushed hard, a convincing hard-tuned effect. The go-to for producers wanting the “T-Pain” sound for free.
- Graillon 2 Free (Auburn Sounds) — a free tier of Graillon with pitch correction plus a pitch-shift/octave “bitcrusher” character. Great for both correction and creative vocal effects on any platform.
- MAutoPitch (MeldaProduction) — a free, tidy automatic pitch-corrector with a formant-shift control and a width knob, good for both subtle fixes and thickening.
- Voloco — better known as a mobile app, but a fast route to hard-tuned vocals if you’re working on a phone or tablet.
Any of these is enough to learn the workflow. Once you understand key, retune speed and manual fixes on a free tool, the paid plugins simply give you cleaner detection and more control.
Fitting autotune into the rest of your mix
Pitch correction is one link in the vocal chain, not the whole thing. Correct pitch first, then shape the sound with EQ and compression, then add reverb and delay for space. If you’re new to that order, our primers on EQ and compression fundamentals and using reverb and delay show how the pieces fit together. For the wider picture of getting a mix over the line, start with the beginner’s guide to mixing, and browse the rest of our mixing and mastering guides.
Frequently asked questions
Does using autotune mean you can’t sing?
No. Autotune is used on the vast majority of commercial vocals, including by excellent singers, as a final polish. It corrects small, unavoidable pitch drift the same way a spellchecker tidies good writing. It only sounds obvious when you deliberately push the retune speed to maximum for effect. A strong performance still matters — correction refines a take, it doesn’t replace singing.
How do I get the hard “T-Pain” autotune effect?
Set retune speed to its fastest (near zero), lock the plugin to the exact key and scale of the song rather than chromatic, and have the singer commit fully to each note with minimal sliding. The instant snapping between fixed pitches, combined with a confident performance, produces the stepped robotic glide. GSnap and Graillon do this well for free.
Should I autotune before or after compression?
Before. Place pitch correction first in the chain so the plugin analyses a clean, uncompressed signal — compression and reverb smear the pitch information the detector relies on. Correct the pitch, then add EQ, compression and effects afterwards. The one exception is a light de-esser, which can sit after correction to tame any sibilance the tuning exaggerates.
