Mid/side EQ basics: what mid and side actually contain

Worked example: take a sample value of 1 in the left channel and 0 in the right. Its mid and side values are both 0.5, so a hard-left source contributes to both. With identical left and right values, side is zero. This simple calculation explains why a side cut can alter a panned instrument while leaving an identical-channel signal alone.

Use documented component audition controls at a sensible monitor level. Identify the content and return to stereo before EQ. A centered vocal may dominate mid while stereo ambience contributes to side; microphones, panning, and effects determine the actual content.

On a drum bus, listen for which cymbal wash, toms, and room details appear in each component. Do not assume side contains only the overheads or mid only kick and snare. Several unrelated sources can share the component you are about to change.

Suppose the drum image has a bright outer wash while the centered snare works. Audition a restrained side shelf and compare in normal stereo at similar loudness. Check crash placement, snare room, and panned toms. If the shelf also removes their useful character, reduce it or treat the contributing source instead.

If both components need the same broad correction, use ordinary stereo EQ. Mid/side earns its place when different component treatment solves an audible problem, not because it exposes more controls.

A side low cut reduces low-frequency channel differences. Choose cutoff and slope from bass notes, low toms, and room tone rather than a fixed bass-mono rule. Check whether the wanted body or spatial cue changes before keeping it.

Different filtering of mid and side can change their reconstructed channel relationships. Compare gentler settings or another phase mode if the image shifts unexpectedly. Linear-phase processing has latency and possible pre-ringing tradeoffs; it is not an automatic requirement.

In the basic reconstruction, summing L and R gives 2M: side cancels. A side-only boost can therefore make stereo more exciting while leaving the basic mono sum unchanged. Check whether the hook and balance still work without that added side emphasis. A later processor that changes or mixes components can alter this simple relationship, so test the finished chain.

Check a sparse section after the chorus. If a chorus correction thins the intro, automate it or move it closer to the source. Keep the bypass version available so you can verify the correction in normal stereo rather than judging a pleasant-sounding solo component.