Check a track against -14 LUFS
Drop an audio file and read its integrated loudness against the -14 LUFS figure Spotify, YouTube, and Amazon normalise towards. The number beside the target is the gain change that would land the file exactly on it.
What normalisation actually does
Streaming platforms measure each track and apply a single gain so everything plays at a similar loudness. A master at -8 LUFS is turned down 6 dB to reach -14. A master already at -14 is left alone.
The consequence is that loudness competition stops paying. Pushing a limiter harder raises the measured figure, the platform removes exactly that amount, and what remains is a track with less dynamic range playing at the same volume as everyone else.
True peak is the real risk
The meter oversamples 4x to find the true peak, because a converter reconstructs a continuous curve between samples and that curve can rise above the highest stored sample. A file reading -0.1 dBFS on samples can be well over 0 dBTP.
That matters after normalisation. If the true peak would cross -1 dBTP once the gain change is applied, the reading turns a warning colour. Lossy encoding decodes to slightly different values than the original samples, and a file sitting on the ceiling clips when it does.
Reading the other figures
- Integrated is the whole-file loudness, gated so silence between phrases does not drag it down. This is what the platform measures.
- Short term max is the loudest 3 second window, and momentary max the loudest 400 ms.
- Loudness range in LU describes how much the loudness varies across the track. A very low figure on a full song usually means heavy limiting.
Other targets
Apple Music aims at -16 LUFS, podcasts at -16, and European broadcast at -23. To see all four at once, use the full Loudness Meter.