// effects · tool 14
Vocal Remover
Strip center-panned vocals from any stereo track with phase cancellation. Adjustable strength, bass protection, instant preview — and your song never leaves your device.
Drop a stereo song here
MP3, WAV, M4A, OGG, FLAC and more — or click to browse
[ drop works anywhere on this page ]
Remove vocals from a song — free, instant, and private
This vocal remover runs entirely in your browser using the Web Audio API. Drop in a song, drag the Strength knob, and hear the lead vocal drop out of the mix in real time. When it sounds right, export an instrumental as lossless WAV or compact MP3. Nothing is uploaded: the processing happens on your own device, which is why there's no queue, no file size cap, and no privacy question mark.
How it actually works (honest version)
This tool works by phase-cancelling center-panned vocals — the classic "out of phase stereo" (OOPS) technique that engineers have used since the vinyl era. On the vast majority of commercial stereo mixes, the lead vocal is panned dead center: it is recorded identically into the left and right channels. If you subtract the right channel from the left, everything that's identical in both — the vocal — cancels to silence, while instruments panned off-center survive.
There's a catch: bass guitar and kick drum are also mixed in the center, and naive subtraction wipes them out too, leaving a thin, hollow karaoke track. That's why this tool has a Keep bass below control: frequencies under the cutoff (120 Hz by default) are protected from cancellation, so the low end keeps its weight while the vocal range above it is removed.
Results vary by song — genuinely. It works best on standard stereo mixes with a single, center-panned lead vocal and works poorly when: the vocal is double-tracked and panned wide, heavy stereo reverb carries a "ghost" of the voice, the song is a mono recording, or the file is heavily compressed (lossy encoding subtly decorrelates the channels, weakening cancellation). On a good candidate you'll get a usable karaoke or practice track in seconds; on a bad one you'll hear why AI separation services charge money and need your file on their servers. We'd rather be honest about the physics than pretend.
When to use this instead of an AI vocal splitter
- Privacy matters — demos, unreleased mixes, voice memos. This page can't leak what it never receives.
- Speed matters — cancellation is instant; AI separation takes minutes of upload and queue time.
- Karaoke and practice tracks — a slightly imperfect instrumental is fine for singing or playing along.
- Checking a mix — engineers use mid/side cancellation to audition exactly what sits in the center of a stereo image.
Tips for better results
- Start at 100% strength, then back off until artifacts (swirly, hollow sound) are least noticeable.
- If the kick or bass sounds weak, raise Keep bass below toward 180–250 Hz.
- Use the highest-quality source you can find — FLAC or 320 kbps MP3 cancels noticeably better than 128 kbps.
- For removing everything but the vocals, try our Equalizer to carve the 300 Hz–3 kHz presence band instead.
FAQ
How does the vocal remover work?
It uses center-channel cancellation: on most stereo mixes the lead vocal is identical in the left and right channels. Subtracting one channel from the other cancels everything in the center — including the vocal — while keeping instruments panned to the sides.
Why can I still hear some vocals?
Reverb and echo on the vocal are usually mixed in stereo, so a "ghost" often remains. Double-tracked or wide-panned vocals and mono recordings cancel poorly. Full isolation needs AI source separation, which requires uploading your file to someone's server — this tool trades some quality for total privacy and instant speed.
Why does the bass survive when vocals are removed?
Bass and kick are also mixed center, so naive cancellation destroys them. Everything below the adjustable cutoff (120 Hz by default) is protected, keeping the low end intact while the vocal range is cancelled.
Does it work with mono files?
No — cancellation needs two different channels to subtract. A mono file has only one channel, so there's nothing to cancel against.