Vocal Remover

Turn a song into a karaoke instrumental by removing the lead vocal, or do the opposite and keep mainly the voice. Drop in a stereo MP3, WAV or M4A, choose the mode and strength, and download the result – the song never leaves your device.

  • Files stay on your device
  • Free, no sign-up
  • Instant
Drop an audio file here MP3, WAV, M4A, OGG, FLAC or WebM

Processed on your device. Your files are not uploaded to any server.

Quick guide · 4 steps

How to use the Vocal Remover

When to use the Vocal Remover

Karaoke at home

Make an instrumental of a favourite song and sing over it at a party or family gathering.

Singing practice

Practise a song with the backing track alone so you can hear your own voice clearly.

Learning an instrument

Remove the vocal to hear the guitar, piano or bass parts more clearly while you learn them.

Remix sketches

Pull out a rough acapella to try new beats under it – for personal experiments and study.

Features and benefits

  • Karaoke (instrumental) and acapella (vocal) modes
  • Frequency-by-frequency separation that keeps bass and drums much better than simple phase cancelling
  • Adjustable strength to balance vocal removal against music quality
  • MP3 or WAV download, processed privately on your device

Example

Input
pop-song.mp3 – 3:28, stereo studio mix
Output
pop-song-instrumental.mp3 – 3:28, 192 kbps, lead vocal strongly reduced

How the Vocal Remover works

The song is cut into overlapping slices about a tenth of a second long. Each slice is converted to its frequencies with a fast Fourier transform, separately for the left and right channels. For every frequency in the voice range, the tool compares the two channels; when they match almost perfectly, that part of the sound is considered centred and therefore likely to be the lead vocal. In karaoke mode those parts are reduced according to the strength setting; in acapella mode everything else is reduced instead. The slices are then converted back to sound and smoothly overlapped.

A real-life example

Fatima was preparing to sing at her cousin’s wedding and had only the original song. She dropped the MP3 into the vocal remover, chose Remove vocals at 90% strength and downloaded the instrumental. A faint echo of the singer remained in the quiet intro, but once she sang over it nobody could hear it. She then used the Key Changer to lower the instrumental by two semitones to suit her voice.

Good to know

In most studio mixes the lead vocal is placed exactly in the centre, so it is identical in the left and right speakers, while guitars, keyboards, backing vocals and reverb are spread across the stereo field. The vocal remover splits the song into thousands of small time-and-frequency pieces and measures how centred each piece is. Pieces that are the same in both channels, within the frequency range of the human voice, are treated as vocal and removed – or kept, in acapella mode.

Older karaoke tricks simply subtracted the right channel from the left. That removes the voice but also removes the bass, kick and snare, which are usually centred too, leaving a thin, hollow sound. Because this tool only works on the vocal frequency range and leaves low frequencies alone, bass and drums survive far better. No method is perfect without the original multitrack recording, so expect some vocal reverb or faint echoes to remain.

Results depend on how the song was mixed. Modern pop, rock and hip-hop with a dry, centred lead vocal usually work very well. Live recordings, mono files, songs with wide stereo vocal effects and some older recordings separate less cleanly. Please use separated tracks for personal practice, karaoke at home and study, and only publish versions of songs you own or have permission to use.

Tips for the best results

  1. Use the highest quality file you have – a 320 kbps MP3 or WAV separates better than a low-quality copy.
  2. Start at 90% strength; go higher for cleaner removal or lower if the music sounds watery.
  3. Songs with dry, centred vocals work best; heavily layered or wide-stereo vocals are harder.
  4. For karaoke, combine with the Key Changer to fit the song to your voice.
  5. Acapella mode gives a rough vocal; it is great for study but not studio-quality.

Common mistakes to avoid

Using a mono recording

The tool needs a stereo file. Look for an official stereo release of the song.

Expecting a perfect studio stem

Without the original multitrack, some reverb and backing vocals remain. Treat the result as a practice track.

Publishing someone else’s instrumental

Removing vocals does not give you rights to the song. Keep it for personal use unless you have permission.

Vocal Remover – frequently asked questions

Why does my result still have some voice in it?

Reverb and echo on the vocal are spread across the stereo field, so they cannot be told apart from the music. Raise the strength, or accept a little ghosting – it is usually hidden when you sing over the track.

Why did it say my file is mono?

A mono file has only one channel, so there is no left–right difference to separate. You need a stereo version of the song.

Can I upload a YouTube link?

No. The tool works on audio or video files on your device, which keeps processing private and fast.

Can I post the karaoke version on YouTube?

Only if you have the rights to the song. Removing vocals does not change who owns the music and backing track.

How long does it take?

A typical 3–4 minute song takes 5 to 20 seconds, depending on your device.

Can it remove the voice from a video?

Drop the video in; the audio is processed and saved as an audio file. Use MP3 to MP4 to add a picture back.