AUD · Audio tools

Audio Loudness Meter

Loudness is not level. A dense mix and a sparse one can share a peak of -0.1 dBFS and sound minutes apart in volume, which is why every platform normalises on a loudness measurement instead.

Integrated LUFS

The integrated reading is the loudness of the whole file. The signal is K-weighted, measured in overlapping 400 ms blocks, and gated twice: an absolute gate at -70 LUFS drops digital silence, and a relative gate 10 LU below the running mean drops the quiet passages that would otherwise pull a dynamic master down.

The filters are designed at the file’s own sample rate, so a 44.1 kHz source is measured at 44.1 kHz rather than resampled first.

Short term, momentary, and range

Short term is the loudest 3 second window and momentary the loudest 400 ms. Both are peaks of the loudness curve, not of the waveform.

Loudness range is the spread between the 10th and 95th percentile of the short-term readings, gated per EBU Tech 3342. A pop master usually lands between 3 and 6 LU. A film score or a classical recording can run past 15 LU, and that number is the reason it will feel quiet next to a pop track at the same integrated loudness.

True peak

The true peak reading oversamples 4x before measuring, which is what BS.1770-4 requires and what a sample peak meter cannot see.

Keeping true peak at or below -1 dBTP is the usual delivery rule. The headroom is not superstition: lossy encoders shift sample values slightly, and a file that sat at 0.0 dBTP before encoding can clip on playback afterwards.

Reading the target table

Each row shows the gain change that would land the file on that target. A reading of +3.0 dB at -14 LUFS means the file is currently at -17 LUFS and would need 3 dB more.

The number turns to a warning colour when applying that gain would push the true peak past -1 dBTP. That is the case where the file cannot simply be turned up, and needs a limiter or a quieter mix instead.

What a meter cannot tell you

It measures the file, not the mix. Two masters at -14 LUFS with the same range can still differ in balance, distortion, and how they translate on a phone speaker.

Use it to confirm a delivery spec and to compare against a reference, not to decide whether the mix is finished.

Frequently Asked Questions

Integrated loudness in LUFS, the loudest 3 second and 400 ms windows, loudness range in LU, and both sample peak and true peak. It reads WAV, MP3, M4A, AAC, Ogg, Opus, FLAC, AIFF, and WebM audio.

dBFS measures sample values against digital full scale. LUFS measures perceived loudness: the signal is filtered to approximate how the ear weights frequency, averaged over time, and gated so silence between phrases does not drag the number down. Two files can peak at the same dBFS and differ by 10 LUFS.

It should not be by more than a tenth of a decibel if both follow BS.1770-4. The usual causes of a real difference are measuring a selection rather than the whole file, or a meter that reports ungated loudness.

A converter reconstructs a continuous waveform between the samples, and that curve can rise above the highest sample. This meter oversamples 4x to find it. A file that reads -0.1 dBFS on samples can be well over 0 dBTP, which clips after a lossy encoder decodes it.

Streaming platforms normalise towards -14 LUFS, Apple Music towards -16, podcasts to -16, and European broadcast to -23. A master louder than the target is turned down on playback, so pushing past it costs dynamics and gains no volume.

Explore Our Tools

Browse all tools