Mid/side processing is a stereo technique that separates a stereo signal into two components: the mid (the center information that is the same in both channels) and the side (the difference between left and right, which is the stereo information). Once separated, you can EQ, compress, or otherwise process the mid and the side independently, then sum them back to stereo. The result is a level of stereo control that is impossible with traditional left-right EQ.
How mid/side works mathematically. The mid signal is L+R (left summed with right, then halved). The side signal is L-R (left with right inverted, then halved). To convert back to stereo: left becomes mid+side, right becomes mid-side. This sum-difference encoding has been used in FM radio broadcasting since the 1960s and in stereo recording since the M/S microphone technique was developed.
The most common mid/side EQ move: cut low frequencies on the side channel. Low-end content (below 200 Hz) in the side channel causes mono compatibility problems and phase issues. By high-pass-filtering only the side channel at 100 to 200 Hz, you keep the low end mono (where it belongs) without affecting the wider stereo image of the mids and highs. This is essentially a more elegant version of "mono-summing the bass" that you can apply to the entire master bus.
The second-most-common move: boost high frequencies on the side channel. A 1 to 3 dB boost at 8 to 15 kHz on only the side channel widens the perceived stereo image of the mix. The center information (vocals, kick, bass, snare) stays where it was; the side information (overheads, reverb tails, wide synths, doubled guitars) gets brighter and more spacious. The result feels wider without actually moving anything off-center.
Mid/side processing on the master bus. The standard mastering chain often includes a mid/side EQ at the end. High-pass the side channel at 120 Hz. High-shelf boost the side channel at 12 kHz. This combination creates a polished, wide, mono-compatible master with minimal effort.
Mid/side EQ on individual tracks. Less common but useful for stereo recordings. On a stereo overhead pair, you can boost the side channel's 10 kHz to make the cymbals wider without affecting the mid (where snare and kick bleed live). On a stereo synth pad, you can high-pass the side channel at 300 Hz to keep the pad from competing with the bass while letting the upper-mid and high content stay wide.
Mid/side compression. Compress the mid channel separately from the side. The mid channel compression makes the lead vocal and kick punch through; the side channel compression preserves the dynamics of the stereo information. This is mostly a mastering technique — most mixing situations don't need this level of separation — but it is a powerful tool when you do.
The phase trap. Mid/side processing only works if your DAW handles the encoding and decoding precisely. If you have a phase issue between left and right (which is more common than people think — out-of-phase backing vocals, badly recorded stereo drums, mis-aligned multi-mic recordings), mid/side processing can amplify the phase issue dramatically. Always check your stereo image with a goniometer or phase meter before mid/side processing.
Mono compatibility check. After applying mid/side EQ, sum your mix to mono and listen. The mid channel content should still be there. If parts of the mix disappear in mono, you have a phase issue — likely caused by side-channel content that is mostly out-of-phase information. The fix is to reduce the side-channel level or to identify and fix the source recording.
Common mistakes: applying mid/side EQ without understanding the source's stereo content (you may be processing nothing because there is no real stereo information), boosting side channel low end (creates mono compatibility problems), excessive side-channel boosts (creates a hollow center where the vocal should be), forgetting to check mono compatibility (the mix may sound great in stereo but disappear on a phone speaker).