Loudness Normalization Tool

Integrated loudness, measured in LUFS (Loudness Units relative to Full Scale, per ITU-R BS.1770 / EBU R128), estimates how loud a whole track sounds to humans using a K-weighting filter and gating. The K-weighting curve emphasises the 1–4 kHz midrange (where hearing is most sensitive) and attenuates very low frequencies. Streaming services normalize everyone’s uploads to a common target so listeners don’t have to ride the volume between songs — mastering far louder than the target just gets turned down (and can sound worse, with squashed dynamics). To measure your own track’s integrated LUFS before entering it here, use the online LUFS meter.

To normalize, you apply a single gain equal to target − current in dB. If your track is −9 LUFS and the target is −14 LUFS, you apply −5 dB. Because it’s one level change, the program loudness shifts by that same amount — unless raising the level would push peaks past your ceiling, in which case a limiter caps the peaks and the achieved loudness gain is reduced. The calculation is straightforward, but understanding the ceiling interaction is where many producers get caught.

Why the peak ceiling matters

Lossy codecs (AAC, MP3, Opus) can produce inter-sample peaks slightly above your sample peaks through the reconstruction filter, so masters are usually limited to a true-peak ceiling around −1 dBTP to avoid clipping after encoding. The dBTP (decibels relative to True Peak) scale accounts for these between-sample excursions that a conventional sample-peak meter would miss. If normalizing up would breach the ceiling, you either accept the lower loudness, apply limiting, or master to the platform target in the first place. Use the peak level analyzer to observe your track’s live crest factor and headroom before mastering.

For podcast normalization, platforms like Spotify for Podcasters and Apple Podcasts target around −16 LUFS, which is louder than the broadcast −23 LUFS standard but quieter than most music streams. Dialogue content at −16 LUFS with a −1 dBTP ceiling is the widely accepted de facto podcast standard.

What is the difference between LUFS, dBFS, and true peak?

dBFS (decibels relative to full scale) measures the amplitude of individual digital samples — 0 dBFS is the highest value the format can store, and all audio sits below it. It tells you how high the signal's samples reach, but says nothing about how loud the track sounds to a listener. LUFS (Loudness Units Full Scale) is a different beast: it applies a K-weighting filter that emphasises the midrange frequencies where hearing is most sensitive, then averages over the whole program with a gating step that ignores near-silence. The result tracks perceived loudness rather than peak level, which is why a heavily compressed pop track and a quiet classical piece can share the same peak dBFS yet differ by 10 LUFS or more. Use the online LUFS meter to measure integrated loudness, or the RMS calculator to compare peak and average levels.

True peak (dBTP) is a third, distinct measurement. When a digital signal is converted to analogue or re-encoded as AAC or MP3, the reconstruction filter can create inter-sample peaks that exceed the highest sample value. A sample-peak meter reading −1 dBFS may produce an analogue waveform that briefly exceeds 0 dBFS. True-peak metering uses oversampling (typically 4x or more, per ITU-R BS.1770) to catch those between-sample excursions before they reach the encoder. Practical targets that account for all three: around −14 LUFS integrated for Spotify, YouTube, and Tidal; −16 LUFS for Apple Podcasts; and a true-peak ceiling of −1 dBTP in both cases to leave headroom for lossy encoding without audible clipping.