MasteringLoudness

What -14 LUFS Actually Means

Streaming platforms turn loud masters down. They do not turn quiet ones up. Here is what the number changes.

All articles

It Is A Playback Setting, Not A Target You Hit

The number gets passed around as a mastering goal, which is backwards. Loudness normalisation is something the platform does to your track on playback. Understanding what it does — and what it does not do — is worth more than any specific figure, because the figures change and the mechanism does not.

Three Numbers, Three Jobs

Integrated LUFS is the average perceived loudness across the whole track. This is the one normalisation acts on, and the only one worth checking at the end rather than during.

Short-term LUFS is a rolling window, useful for seeing how much your chorus jumps above your verse. True peak is measured between samples rather than at them, which matters because digital-to-analogue conversion and lossy encoding both reconstruct a waveform that can overshoot the sample values you can see.

What Normalisation Actually Does

If your track measures louder than the platform's target, playback gain is reduced until it matches. That is the whole operation. Every decibel of dynamic range you crushed to get louder is turned back down at the other end, and all you kept was the damage.

Exact targets vary between services and have changed over the years, and some offer settings that change them again. Treat any specific number you read — including this article's title — as something to verify against the platform's own current documentation rather than as a constant.

Where Loudness Still Wins

  • Club and DJ playback, where files are played directly with no normalisation in the path.
  • Local files, downloads and anything played outside a service that normalises.
  • Unmatched A/B comparisons — in a playlist next to a louder reference, louder still wins for the first few seconds.
  • Platforms and contexts where normalisation is off by default or user-configurable.

A Workflow That Survives Both Worlds

Mix For Dynamics, Not For The Meter

Leave the loudness meter off while you are making balance decisions. A meter running in your peripheral vision changes the choices you make and none of those changes improve the mix.

Set A True Peak Ceiling With Headroom

Around -1 dBTP is a common working ceiling. It leaves room for intersample peaks and for the overshoot introduced when a lossy encoder reconstructs the waveform, which is where distortion that never appeared in your bounce comes from.

Measure Integrated At The End

Check the number once the master is done, to know where you landed. If it is far from where you expected, that is information about your mix, not an instruction to push harder into the limiter.

Always Compare At Matched Loudness

Turn your track and your reference to the same measured level before deciding which one sounds better. Every comparison made at unmatched loudness is a comparison of levels, not of mixes.

Same LUFS, Still Quieter Than The Reference

Loudness measurement models average perceived level. It does not model density, spectral balance, or how much of your energy sits in the range the ear is most sensitive to. Two tracks can measure identically and sound a long way apart.

When your master measures the same as a reference but feels smaller, the answer is almost never more limiting. It is usually mid-range presence, arrangement density, or low end that is taking up headroom without being audible on the systems you are being compared on.

Press & Media

You may also like

Our latest news

More press & Media