Skip to content

Stereo correlation: what the reading tells you

Send feedback

Last updated 5 min readTags:MixingStereoMastering

Learn what stereo correlation readings mean, how to check your mix in mono, and what to change when vocals, bass or instruments lose definition.

Stereo correlation measures how closely the left and right channels of an audio signal vary together. The scale runs from −1 to +1.

The reading can flag a risk of cancellation when the channels are combined for mono playback. It cannot tell you which instrument is affected or whether the change is audible. Use it alongside a mono listening check.

How to read stereo correlation

Reading What it means What to check
+1 The channels have the same waveform shape and polarity. Their levels may differ. The signal has no independent left/right content. Decide whether that suits the recording.
Between 0 and +1 There is some positive relationship between the channels. Listen for changes in individual instruments when you switch to mono.
0 There is no overall linear correlation between the channels over the measured period. Check the mono sound. Zero alone does not establish a fault.
Between −1 and 0 There is an overall negative relationship between the channels. Check for cancellation, particularly if the reading stays negative.
−1 The channels have matching waveform shapes with opposite polarity. Equal-level versions cancel completely when summed to mono.

These values describe the relationship between the channels. They are not mix-quality grades.[1]

Is there an ideal value?

There is no single value that every mix should reach. A sparse recording may sit close to +1. A mix with prominent stereo ambience may read lower.

A reading of 0.65 does not prove that every instrument will survive mono playback. Nor does 0.15 prove that something is wrong. The arrangement, processing and affected sounds all matter.

Brief negative readings deserve a listen, but they do not automatically require a correction. A sustained negative reading is more concerning, especially if a lead instrument or bass part becomes noticeably weaker in mono.[2]

Why mono playback changes the sound

Mono summing combines the left and right channels. Shared content remains, while content with opposing polarity can cancel.

For example, a widened synth may sound prominent in stereo but become thin in mono. The stereo effect may introduce differences that cancel when the channels are combined.

Not every change is a phase problem. Mono removes left/right separation, so instruments may mask one another more. A narrower sound is expected; a missing vocal, hollow bass line or substantial change in tone needs investigation.

What a whole-track reading can miss

One value for an entire track cannot show when a problem occurs. A wide chorus or short effect may behave quite differently from the rest of the song.

A strong centre signal can also dominate the result. The vocal, kick and bass may keep the overall reading positive while a quieter stereo instrument loses definition in mono.

Full-band correlation does not isolate bass behaviour either. Use the sub stereo width result alongside it when checking low-frequency imaging.

For a particular passage, use a correlation meter in your DAW and listen to that passage in mono. A meter that updates during playback provides information that a single whole-track result cannot.[2]

How to check and correct a problem

1. Listen to the mono sum

Use your DAW or monitor controller to sum left and right to mono. Listening to only the left or right channel is a different check.

Keep the listening level reasonably consistent. Focus on the vocal, kick, bass and main musical parts. Note exactly what becomes weaker or changes tone.

2. Find the source

Loop the affected passage. Bypass stereo wideners, short delays, chorus effects and other spatial processing one at a time.

If several microphones capture the same source, check their timing and polarity relationships. Test the suspected source on its own, then return it to the mix.

3. Adjust the cause

If a widening effect causes the loss, reduce its amount or change its settings. Check delay times and the balance between dry and processed sound.

A polarity reversal may solve a specific microphone or routing error. It is not a general remedy for a low correlation reading.

A master-bus width control may narrow the stereo image without restoring content that cancels in mono. Make the correction on the source or effect responsible.

4. Compare both versions

Listen again in stereo and mono. Keep the change if it preserves the important parts without spoiling the stereo mix.

The aim is for the arrangement to remain clear in both versions. The number alone cannot establish that.

What if the reading is close to +1?

A high reading is not inherently a fault. Mono recordings and deliberately narrow arrangements can be entirely appropriate.

If the mix sounds narrower than you intended, review the panning and stereo effects on supporting parts. Compare any changes in mono before keeping them.

How is stereo correlation calculated?

One method is the Pearson correlation coefficient, calculated from corresponding samples in the left and right channels. It measures how the signals vary together, normalised by the variation in each channel.[3]

The measured period matters. A calculation across the whole file gives one overall result; a meter that uses short windows shows changes during playback. These readings need not match.

If either channel has no variation, such as a completely silent channel, the Pearson calculation is undefined. That is a measurement limitation, rather than evidence of a stereo fault.

Sources

  1. Apple Logic Pro: Correlation Meter, for the correlation scale and mono-cancellation behaviour.
  2. iZotope Ozone 11: Imager, for correlation monitoring, widening and the interpretation of brief low readings.
  3. NumPy: corrcoef, for the Pearson calculation.

Was this page helpful?

    Type to search titles, headings and summaries.