Vocal Remover screenshot

Vocal Remover

Remove centre-panned content from a stereo track with mid/side cancellation, hear what remains, and download it as a mono WAV. No model, no upload.

Privacy First
0 Runs
Updated 2026-11-22
Runs in your tab
No model
Mid/side only
Mono WAV

What this page makes

Computes (left − right) / 2 so anything identical in both channels cancels, including a centred vocal and centred instruments.

How to use it

Choose a stereo file, press Remove centre, then Play result or Download result WAV. The readout reports the mono length and the RMS of the result beside the left channel's RMS; a much lower level is the cancellation working.

Worked example: a stereo mix with the lead vocal centred and a guitar panned left becomes the centred vocal cancels, the panned guitar remains at half amplitude, and the result is mono.

Method

The decoded left and right channels are subtracted and halved: out = (L − R) / 2, written as one mono channel. A signal at the same level and phase in both channels cancels exactly; a signal panned to one side survives at half amplitude. There is no separation model and no vocal detector.

What it refuses

A mono file is refused with the reason that its channels are identical and the result would be silence. The page refuses to call the output an isolated vocal: it removes the centre, and a centred bass, kick or snare goes with it.

Where it runs

The file is decoded and subtracted in this tab: it is never uploaded and no copy is stored. The browser requirement is stated on the page: Decoding needs decodeAudioData and playback needs an AudioContext. The cancellation is one subtraction per sample pair.

The rest of the family

The other audio and recurrence pages are listed under Calculators: tone and binaural generation, the virtual piano, pitch shifting, looping and reversal, the stated noise filter chain, mid/side cancellation, microphone recording and the closed-form recurrence solver.

Computes (left − right) / 2 so anything identical in both channels cancels, including a centred vocal and centred instruments.

The decoded left and right channels are subtracted and halved: out = (L − R) / 2, written as one mono channel. A signal at the same level and phase in both channels cancels exactly; a signal panned to one side survives at half amplitude. There is no separation model and no vocal detector. The file is decoded and subtracted in this tab: it is never uploaded and no copy is stored.

Worked example: a stereo mix with the lead vocal centred and a guitar panned left becomes the centred vocal cancels, the panned guitar remains at half amplitude, and the result is mono.

The readout prints the result's RMS beside the left channel's, so the cancellation is visible as a number.

How to use the vocal remover

Choose a stereo file, press Remove centre, then Play result or Download result WAV. The readout reports the mono length and the RMS of the result beside the left channel's RMS; a much lower level is the cancellation working.

The file is decoded and subtracted in this tab: it is never uploaded and no copy is stored.

  • Domain: One stereo audio file the browser can decode; the output is mono WAV.

What the page proves

Computes (left − right) / 2 so anything identical in both channels cancels, including a centred vocal and centred instruments.

The decoded left and right channels are subtracted and halved: out = (L − R) / 2, written as one mono channel. A signal at the same level and phase in both channels cancels exactly; a signal panned to one side survives at half amplitude. There is no separation model and no vocal detector.

  • Browser requirement: Decoding needs decodeAudioData and playback needs an AudioContext. The cancellation is one subtraction per sample pair.
  • The page states its domain: One stereo audio file the browser can decode; the output is mono WAV.

What it refuses, and what it leaves out

A mono file is refused with the reason that its channels are identical and the result would be silence. The page refuses to call the output an isolated vocal: it removes the centre, and a centred bass, kick or snare goes with it.

  • Not the right page for: isolating a vocal while keeping centred instruments
  • Not the right page for: a vocal that is panned or doubled wide
  • Not the right page for: a mono recording, which has no side signal
  • Reported as a named state: a mono file produces silence and is refused
  • Reported as a named state: a centred instrument disappears with the vocal
  • Reported as a named state: a wide or doubled vocal may survive instead of cancelling

What to Expect

Best for

  • making a rough karaoke backing from a centred vocal
  • hearing what sits in the centre of a mix
  • a quick demonstration of mid/side arithmetic

Not ideal for

  • isolating a vocal while keeping centred instruments
  • a vocal that is panned or doubled wide
  • a mono recording, which has no side signal

What this tool keeps

  • the difference between the channels, and nothing that is identical in both

What may need cleanup

  • expect a quieter, mono result; there is no level restoration
  • a centred bass or kick may need re-adding from another source

Common errors

  • a mono file produces silence and is refused
  • a centred instrument disappears with the vocal
  • a wide or doubled vocal may survive instead of cancelling

Example use cases

  • make a rough backing track
  • check whether a vocal is centred
  • demonstrate what mid/side removes

Sample input

a stereo mix with the lead vocal centred and a guitar panned left

Sample output

the centred vocal cancels, the panned guitar remains at half amplitude, and the result is mono

Who this is for

  • karaoke singers making a rough backing
  • audio learners seeing mid/side cancellation

Frequently asked questions

Is my data sent anywhere?

The file is decoded and subtracted in this tab: it is never uploaded and no copy is stored.

Which browser API does this page need?

Decoding needs decodeAudioData and playback needs an AudioContext. The cancellation is one subtraction per sample pair.

What can I download?

Download result WAV writes the mono cancellation as 16-bit PCM.

Does this extract the vocal?

No. It removes the centre, which is usually where the vocal sits. The output is the rest of the mix, and a centred bass, kick or snare is removed too. Real separation needs a model, which this page does not have.

Why is the result mono?

Because L − R is one signal. Writing it to both channels would be the same audio twice, so the page writes one mono channel and says so.

Why is it so quiet?

Cancellation removes energy. A centred vocal and any centred instruments are subtracted, so the remaining panned content is quieter. The page does not boost it back, because that would change the level you are judging.

Developer program

Build custom tools for your store

Join the SDK program to create useful ecommerce utilities for your store.