How to tune vocals (pitch correction basics)

Pitch correction is everywhere in modern music, from invisible touch-ups that just fix a few flat notes to the hard, robotic effect that's a sound in its own right. Both come from the same tools, and the difference is entirely in how you set them. Here's how to tune vocals — the two main approaches, how to keep it natural when you want to, and where tuning fits in the vocal chain.

Key Takeaways

• Two approaches: automatic (fast, real-time) and graphical (manual, note-by-note control).
• Retune speed is the main dial — slow for natural, instant for the hard "autotune" effect.
• Tune early, before tone shaping, so you're processing the corrected performance.
• The best tuning is invisible: fix the pitch, keep the human phrasing.

The two ways to tune

Automatic pitch correction works in real time: you set the key and scale, and the tool snaps notes toward the nearest correct pitch as they play. It's fast, great for quick fixes and for the deliberate hard-tuned effect, and it's what most people mean by "autotune."

Graphical pitch correction shows the vocal as editable notes on a grid, and you move and reshape each one by hand. It's slower but far more musical — you can correct only the notes that need it, preserve natural pitch drift and vibrato, and avoid the artifacts automatic mode can introduce. For a natural result on an exposed lead, graphical wins.

Many mixers use both: automatic for a first pass, graphical to fix what the automatic mode got wrong or made unnatural.

Retune speed: the one dial that matters most

Whatever tool you use, retune speed (how fast a note is pulled to pitch) is the control that defines the sound:

• Slow retune — the pitch is corrected gradually, preserving the natural slides, scoops, and vibrato of the performance. This is the setting for invisible tuning.
• Instant retune — the pitch snaps immediately to the grid, removing the natural transitions and creating the hard, robotic "autotune" effect. This is a creative choice, not a correction.

If you want tuning nobody notices, slow it down. If you want the effect, speed it up. Most "why does my tuning sound robotic when I didn't want it to" problems are just retune speed set too fast.

Keep it natural (when that's the goal)

For transparent tuning: use a slow retune speed, correct only the notes that are actually off (not every note), and preserve vibrato rather than flattening it. The goal is to fix the pitch while keeping the human phrasing — the little imperfections that make a vocal feel real. Tune the performance, don't erase it.

Where tuning fits in the chain

Pitch correction goes early in the vocal chain — typically right after cleanup and before compression, de-essing, and tone shaping. You want to correct the raw performance first, then process the corrected version, so your compressor and EQ are reacting to the final notes rather than the flat ones. Tuning is part of getting a solid, finished vocal before the character stages.

One note: aggressive tuning can slightly shift the character of consonants and esses. It won't usually change your de-essing much, but as always, de-ess the tuned vocal (not the raw one) so you're treating what you'll actually hear. A transparent de-esser like Sibilance handles a tuned vocal cleanly, and it's free to start.

The short version

Tune vocals with automatic correction for speed or graphical for musical, note-by-note control, and let retune speed decide the sound — slow for invisible, instant for the hard effect. Correct only the notes that need it, preserve vibrato, and tune early in the chain so everything downstream processes the corrected performance. The best tuning fixes the pitch and keeps the human.

Sibilance is the single-authority de-esser this comes from — a real free tier, and a 7-day Pro trial with no card.

Related reading