extract mid side channel audio
Mid-Side (MS) encoding decomposes a stereo signal into two orthogonal components: Mid (M = (L+R)/2), which contains what is common to both channels, and Side (S = (L-R)/2), which contains only the content unique to each channel. This decomposition is widely used in mastering to control stereo width independently from the center.
If the Side channel is very low in volume, the mix is essentially mono. If very loud, there is aggressive stereo content.
Use the result as technical or educational support, keeping the tool limits explicit in the workflow.
Averaging sums L and R and halves the result, which preserves centred content (vocals, kick) but can cancel anything recorded in opposite phase. The L and R modes simply discard one channel: you lose whatever only existed on the other side, but you gain a guarantee that nothing will cancel. When the correlation strip shows negative stretches, that is exactly the trade-off to weigh.
Yes. Songs with little stereo content (voiceovers, centered studio recordings) will have a very quiet Side channel. Songs with heavy reverb, wide-stereo synthesizers, or panning effects will have a louder Side channel.
No. It helps explain the scenario and use the tool more safely, but real decisions should consider official sources, full context and qualified guidance when needed.
Your audio is decoded and processed in your browser. No file is ever uploaded to the server.