Stereo width and correlation
Correlation measures how alike the left and right channels are. +1 means identical, which is mono. 0 means unrelated, which is very wide. Negative values mean the channels are fighting each other and will cancel when summed to mono. A good mix lives mostly between +0.3 and +0.9.
Reading the meter
| Reading | Meaning | In mono |
|---|---|---|
| +1 | Left and right identical | No change |
| +0.5 to +0.9 | Solid centre with stereo content around it | Sounds much the same, a little narrower |
| 0 to +0.5 | Wide. Lots of difference between channels | Wide elements drop in level |
| 0 | Channels unrelated, like two different recordings | Wide elements drop about 3 dB |
| Below 0 | Opposite polarity dominates | Content cancels. Parts go missing |
| −1 | Left is the inverse of right | Silence |
Width is the other view of the same thing
Width figures compare the energy in the side signal, what differs between left and right, with the energy in the mid signal, what they share. A mono track has zero side energy and zero width. A mix where everything is panned centre, with the faders just pushed up, measures near zero too.
Typical finished mixes carry most of their energy in the mid, because kick, bass, snare and lead vocal are all central. Width comes from what you place around them.
Where healthy width comes from
- Panning different parts apart. Two different guitar takes, left and right. This is the best kind: it's wide, and in mono both takes are still fully there.
- Stereo recordings. Overheads, piano, room mics.
- Stereo reverb and delay. Returns that differ left to right.
- Doubles. The same part performed twice and panned apart.
Where trouble comes from
- Stereo widener plugins that work by delaying or inverting one channel. They raise width by lowering correlation, and what they add disappears in mono.
- A single take copied to two tracks with one delayed. Wide in stereo, comb-filtered in mono.
- One channel with flipped polarity, from a miswired cable or an accidental button press. The meter sits near −1 and the centre of the mix is hollow.
- Chorus and wide patches on bass. See mono bass.
A mono check, properly
- Sum the master to mono.
- Listen for what gets quieter. Lead vocal, kick, snare and bass should be unchanged.
- Guitars, keys and reverb will drop a little. That's expected.
- If something nearly vanishes, find that track and look at how its width was made.
How wide to go
There's no target number. A dense rock or pop mix with doubled guitars is wide. A jazz trio or a rap vocal over a sparse beat is narrow, and that's right for it. The useful test is comparative: a mix that measures far narrower than commercial tracks in the same style probably has everything panned centre.
Quick answers
- What should a correlation meter read?
- Mostly between +0.3 and +0.9 for a full mix. Brief dips toward 0 are normal. Sustained readings below 0 mean the mix will lose content in mono.
- Why does my mix sound different in mono?
- Anything that differs between left and right is reduced or cancelled when the channels are summed. Wide reverbs, stereo wideners and out-of-phase doubles are the usual causes.
Audizer does this when you ask.A Mac DAW with an engineer built in. Say what you want and it makes the edit. Free download.
Get Audizer