Mid-side processing splits a stereo signal into two components. Mid is the sum of the left and right channels, and it contains everything that is common to both. Side is the difference, and it contains everything that is not. The two are mathematically reversible, and most mixing plugins offer them as a display and control format rather than as an effect.
What each component contains
The mid signal is what you hear when the track is played in mono. It carries the bass, the kick, the lead, the vocal, the snare body, and the fundamental of everything significant. It is usually the majority of the energy in a mix, and it is where the mix actually lives.
The side signal carries reverb tails, room information, double-tracked width, and the parts of instruments that exist only because they are spread across the stereo field. It is a smaller portion of the energy, and it is responsible for almost all of the sensation of width.
This is why widening techniques feel cheap when overused. Increasing side energy relative to mid energy makes the image wider, but the side signal is also the part that disappears in mono, so the effect is producing a perception of width in exchange for potentially losing content.
The centre is the priority
A common mistake is treating mid and side as two equal halves. They are not. The mid carries your mix and the side decorates it. A mix that sounds thin and unclear has usually lost mid content rather than side content, and widening it will make the problem worse while the cause remains.
The practical priority is therefore to establish the centre first. Levels, low end, the vocal, the lead, the relationship between bass and kick. Once that is solid, the side signal becomes a genuine enhancement rather than a distraction from a broken centre.
Methods for adding width
Doubling a part and panning the copies is the cleanest approach, because everything remains in the mid when summed and you control the balance directly. Reverb and delay on sends produce width naturally and are usually preferable, since the width comes from a real acoustic event rather than from processing.
Haas processing delays one channel by a small amount, which the ear interprets as width. It works well on some material and badly on others, and its failure mode is comb filtering on anything with significant low-frequency content, which is why it is risky on full mixes. M/S high-passing the side signal is the more useful version, since it widens the top of the spectrum while leaving the low end intact and stable.
Working with the mid and side controls
Solo the mid and judge your mix on it. If the mix does not work in mid alone, width will not save it. This check takes thirty seconds and catches more problems than any other single test.
Then solo the side and listen for anything that should not be there. Side-channel instruments, out-of-phase signals and incorrectly panned low content all show up here and nowhere else.
High-pass the side signal before widening it. Anything below roughly 300Hz in the side signal is phase trouble waiting to happen, and removing it costs nothing perceptually.
Check correlation and mono after every width move. A correlation value below minus 5 on the full mix means content is being lost in translation.
Mid and side content by source
| Source | Primarily | Notes |
|---|---|---|
| Kick | Mid | Should never be in the side signal |
| Bass | Mid | Side content here causes mono loss |
| Lead vocal | Mid | Reverb sends supply the width |
| Snare top | Mid | Room mic supplies the side |
| Overheads | Side | Legitimate and safe above 300Hz |
| Reverb returns | Side | Where most width should come from |
| Delay throws | Side | Natural width, easy to control |
Retuning and the stereo field
A retune is a mono process, so it does not by itself change your stereo image. What it changes is the frequency content that sits in the mid, and since a denser low end is more prone to side-channel problems, the low end is where to look after retuning with 432Hz MASTER. Re-soloing the side signal and checking the region below 200Hz is a quick and worthwhile check.
With RASTA 528Hz, where intervals change rather than slide, the harmonic content of individual instruments changes, which can shift how much of each source ends up in the side signal. If the mix already relied on side-heavy processing, it is worth remeasuring rather than assuming the balance held.
The bottom line
Build the centre and then decide how much space to put around it. Most mixes that sound thin are thin in the mid, and no amount of side energy will change that.
Frequently asked questions
What does mid-side processing do?
It splits a stereo signal into mid, which is the sum of both channels and contains content common to them, and side, which is the difference and contains content unique to each channel. Each can be processed independently, and the two recombine into the original stereo image.
Why does my mix sound thin when I widen it?
Usually because the mid was already lacking. Widening increases side energy relative to mid, so if the centre is weak, widening makes the mix feel more spacious without making it more solid. Build the centre first, then widen.
Is it safe to have instruments in the side signal?
Above roughly 300Hz, generally yes. Below that, side content tends to cancel in mono and can cause instability. Side-channel instruments should not carry significant low-frequency content, and a high-pass on the side signal is the standard safeguard.
Should I widen after retuning my mix?
Re-check rather than assume. A retune is a mono process and does not change the stereo image directly, but denser low end after retuning makes side-channel problems more likely. Soloing the side signal and inspecting the region below 200Hz is a quick check.