AI Vocal Tuning
Vocal tuning that keeps the performance
Put every note where the singer meant it, as gently or as hard as you like. You set how far and how fast notes move, the singer's tone stays theirs, and every slide and run keeps its shape.
Uses one minute of your allowance per minute of audio: a four-minute track costs four minutes.
How it works
What it measures, what it changes
And what stays yours to decide.
- 01
Every note is heard on its own
The vocal's pitch is tracked continuously and split into notes, in the key and scale you name, or to the nearest semitone if you leave it open. Each note is corrected from where the singer actually landed, instead of the whole take being dragged onto a grid.
- 02
You decide how much, and how fast
Strength sets how far a note moves towards the target; retune speed sets how quickly it gets there. Slow retune is transparent correction nobody notices. Fast retune is the famous hard-tuned sound. Same tool, your call.
- 03
You hear it level-matched, then decide
The tuned take plays beside the original at matched loudness, so all you hear is the correction. If it is not better, keep the original: it is never touched.
The specifics
What you would ask before trusting it with your audio
The formats, standards and numbers behind every result, each one measured.
- Pitch tracking
- A learned pitch tracker, backed by a classical estimator wherever the model is unsure, with each correction restarting at the note's onset. A correction begins where the note begins, and a deliberate bend is corrected from where it lands instead of being fought along its whole length.
- Resynthesis
- Formant-preserving resynthesis, so moving a note does not change the apparent size of the singer's voice. Preservation is a switch, for the rare take where the moved formant is the effect you want.
- Scale and key
- Correction aims at the key and scale you set, major or minor, or, left chromatic, the nearest semitone. Chromatic is the safe choice when you are unsure of the key, and on real singers it finds the note they meant nine times in ten. Name the key when you know it.
- Natural at every setting
- Even the strongest setting stops where the voice still sounds like the singer, so you can push correction hard and still keep the person in the performance.
- Naming
- This is pitch correction and vocal tuning, an AI vocal tuner if that is what you searched for. It is not Auto-Tune, which is Antares' product and trademark. The hard-tuned sound is one setting here: a fast retune speed.
Questions
What people actually ask
- Will it sound obviously tuned?
- Only if you want it to. Retune speed decides: fast is the recognizable effect, slow is invisible correction. The default is gentle, because that is the version most people want.
- Can it tune a vocal that is still in the mix?
- Yes. Switch on whole-song mode and Euphona takes the voice out of the music, tunes it and puts it back, all in one job.
- What about harmonies and doubles?
- Tune each take as its own job against the same key. Give a harmony or double the lead as its reference and its consonants and note starts line up with the lead's, never moved by more than a quarter of a second, while the small differences that make a double sound like two people stay.
- Does it fix a bad take?
- It fixes pitch, and with a lead take as its reference it lines a double up with the lead. The phrasing and the feeling are the singer's own, and that is exactly what it protects.
Try it
Put your own track through it
The free plan runs the whole engine on your whole recording and plays you the result, matched for loudness against the original. No card. You pay when you want the full-quality file, one track at a time, or with a plan if you make a lot of them.
Who it’s for