Loudness Meter — Free LUFS Tool

This online loudness meter measures integrated LUFS, short-term and momentary loudness, loudness range and peak level from an audio file using the ITU-R BS.1770 algorithm with dual-stage gating. Drop in an MP3, WAV, FLAC or M4A and the whole analysis runs in your browser; the file is never uploaded.

How to use it

1. Load audio. Drop a track onto the dropzone or click it to browse. The file is decoded locally, so long tracks take a moment before the numbers appear.
2. Pick a preset or keep your file. The four presets are calibrated demo masters: Pop (−14 LUFS), Dynamic (−18 LUFS), Broadcast (−23 LUFS) and Club (−9 LUFS). Each loads a 4-second 1 kHz tone at that level, which is useful for seeing how the readouts behave before you have a file handy.
3. Press Play Preview to hear the current audio, and Stop to end playback. Output volume does not change the measurement.
4. Read the badge first. The large number is integrated LUFS for the whole programme. Short-Term (3s), Momentary (400ms), LRA and Peak Level follow in the readout grid.
5. Check the target table. Each row shows a platform, its target in LUFS, the gain difference for your track in dB, and whether the platform would turn it down.
6. Use the bar scale from −36 LUFS to 0 LUFS, whose markers sit at the Apple −16, Spotify −14, EBU −23 and Club −9 positions, to see where the result lands.

How the measurement works

Loudness is not level. A 60 Hz tone and a 3 kHz tone at the same dBFS do not sound equally loud, because the ear is far more sensitive in the midrange, so the standard filters the signal before it measures energy.

K-weighting

The ITU-R BS.1770 algorithm passes audio through two filter stages. Stage one is a high shelf with about 4 dB of gain centerd near 1.7 kHz, modeling the acoustic effect of the head and ear canal. Stage two is an RLB high-pass that rolls off below roughly 40 Hz, because subsonic energy carries real signal power but contributes little perceived loudness.

400 ms blocks and two gates

Filtered energy is summed in 400 ms blocks with 100 ms hops, so consecutive blocks overlap by 75%. The absolute gate discards every block below −70 LKFS, which keeps silence and noise floors out of the average, and the relative gate then discards blocks more than 10 LU below the mean of what survived. A −0.691 dB offset inside the log conversion anchors the result to the standard’s scale.

Measure	Analysis window	What it tells you
Momentary	400 ms	The loudness now, useful for watching a single hit or a fade
Short-Term	3 s	The level of a phrase or chorus, close to what a listener would ride by hand
Integrated	Whole programme with gating	The single number to compare against a delivery target
Loudness Range (LRA)	10th to 95th percentile of short-term values above a −20 LU gate	How much the loudness moves, in LU
Peak Level	Highest sample magnitude	The digital ceiling of the decoded file in dBFS

EBU Tech 3342 defines the LRA percentile method used here, and EBU R 128 applies it to broadcast programmes.

Delivery targets and what platform gain does

Streaming normalization applies playback gain; it does not restore dynamics. Spotify publishes a target of −14 dB LUFS measured with the ITU 1770 method, and its mastering guidance asks for true peak below −1 dBTP, or below −2 dBTP when a master is louder than −14 LUFS. A track that integrates at −9.5 LUFS therefore sits 4.5 dB above the target, and the meter’s gain column will show roughly −4.5 dB: the platform turns it down, the compressed dynamics remain, and the loudness advantage disappears.

Target	Integrated loudness	Notes
Spotify (published)	−14 LUFS	Spotify’s own guide; keep true peak below −1 dBTP, or below −2 dBTP if the master is louder than −14 LUFS
Apple Music (Sound Check)	−16 LUFS	The meter’s preset, built on the long-standing Sound Check convention
YouTube	−14 LUFS	The meter’s preset and the common delivery convention
Broadcast (EBU R 128)	−23.0 LUFS	EBU R 128 target, with a true-peak ceiling of −1 dBTP, a ±1.0 LU tolerance where the target is not achievable in practice, and ±0.2 LU allowed for QC measurement error
Club (meter preset)	−9 LUFS	A playback-room preset for loud venues, not a standards-body target

Targets and platform behavior change, so treat the table as a starting point and confirm current delivery specs with your distributor.

Reading a result

An integrated figure above the target usually means the track is limited hard enough that the platform will pull it down, while a figure far below the target means the platform may lift it, subject to its own headroom rules. LRA describes the spread: a value near 1 LU means the loudness barely changes, which is typical of a heavily limited pop master, while a wide LRA means the quiet sections really are quiet. Short-term peaks above the integrated value are normal on tracks with a loud chorus and a quiet verse, and the difference between momentary maximum and integrated loudness is a quick guide to how much the loudest moments stick out.

Accuracy, limits and troubleshooting

• Peak Level is sample peak, not true peak. It reports the largest sample magnitude in the decoded file. A band-limited signal can overshoot between samples, so the true peak can be higher; matching a −1 dBTP delivery ceiling needs oversampled true-peak metering.
• Channel weighting. Decoded channels are averaged with equal weight. The surround weighting defined in BS.1770 for multichannel material is not applied, so results are most reliable for mono and stereo files.
• Short files. Anything under 400 ms is measured as a single block with no gating, which makes the result indicative rather than standard.
• Quiet material. Gated blocks below −70 LKFS are discarded, so a near-silent file can show a dash instead of a number.
• Decoding depends on the browser. MP3, FLAC, AAC and WAV support varies between Safari, Chrome and Firefox; a file that fails to decode is reported rather than analysed.
• Lossy sources. A decoded MP3 or AAC file can differ slightly from the original around sharp transients; use a lossless WAV or FLAC when the exact number matters.
• This is a measurement, not a certification. Broadcast compliance uses defined meter scales and test signals (see EBU Tech 3341 for loudness metering), and a browser tab cannot replace a calibrated facility meter.

Related tools

• Dynamic Range Test: compare peak and RMS behavior across a file to see how much limiting it has already taken.
• Decibel Meter: check the acoustic level of your monitoring room while you work.
• Audio Spectrum Analyzer: look at the spectral balance behind a loudness reading.