FIND THE CENTER. HEAR THE DIFFERENCE.
A different angle
on the same recording.
01 / THE MIX
One file · 80 MB maximum · up to 2 minutes · mono or stereo · formats this browser can decode. Video needs an audio track.
The center track contains sound shared by left and right. The instrumental track reduces that shared sound. Centered drums, bass and instruments are affected too; stereo vocals, reverb and effects may remain. Mono files can be cleaned but cannot be separated this way.
02 / LISTEN AGAIN
Add a stereo recording or try the test mix.
Preview the shared center and reduced-center stereo output before downloading.
Shared sound, with your selected voice cleanup settings.
Download center WAV ↓Stereo output with shared sound reduced by the selected amount.
Download instrumental WAV ↓Outputs are uncompressed 16-bit PCM WAV and can be larger than the original. This process cannot recover original studio stems. Check the previews before using a result.
This local tool uses the relationship between left and right channels. At full reduction, the instrumental output is the difference between them. This can reduce a lead vocal placed in the center of a stereo mix, but it also affects instruments in that position. Results depend on how the original was mixed.
Use the high-pass filter to reduce low-frequency rumble and the noise gate to soften quiet background noise between sounds. Normalization changes the output level; it does not improve separation. These controls are not a neural source-separation model or a guarantee of a clean vocal.
A mono file has no left/right difference. The tool provides a cleaned mono track and skips the instrumental output. A stereo file with identical channels also has no stereo difference, so its full-reduction instrumental result can be silent.
The test is generated on your device: a 440 Hz tone shared by both channels and an 880 Hz tone with opposite left/right phase. With full center reduction, the 440 Hz sound disappears from the instrumental output. This demonstrates the method; it is not a promise about music or speech.
