— Free tool
LUFS Checker
Integrated loudness, true peak and what every platform will do to your master, measured to ITU-R BS.1770-4. Everything runs in your browser — your audio is never uploaded. Drop a whole EP at once if you like.
Drop a track here
analysed in your browser, never uploaded
or click to browse — WAV, MP3, M4A, FLAC, OGG · multiple files welcome
— How to read it
Three loudness numbers, three timescales
LUFS measures perceived loudness — the audio is filtered to match how ears weight frequencies, then averaged over time. Integrated loudness is the whole track boiled down to one gated number, and it is what streaming platforms read. Short-term loudness averages a 3-second window, which is why the chart above shows your drops and breakdowns so clearly. Momentary loudness watches 400 milliseconds and catches the most intense instant. A club master might read −7 integrated with momentary peaks at −5; an ambient piece might sit at −18 with huge swings between the two.
Why true peak matters more than sample peak
Your DAW’s meter reads sample values; your listener hears a reconstructed waveform that swings between them. Push the ceiling to −0.1 dBFS and the reconstruction — and worse, the AAC or Ogg file your distributor actually streams — can pass 0 and clip. Measuring true peak with 4x oversampling, the way this tool does, is how you see that before an encoder finds it for you. Around −1 dBTP is the safe ceiling for streaming; club-only material has more licence.
The −14 LUFS myth
You will read that you must master to −14 LUFS because that is Spotify’s level. You don’t. Normalisation is a volume knob, not a mastering engineer: louder masters are simply turned down, and nothing is lost except the loudness war. Mastering hotter than −14 is fine and completely normal for club genres — what matters is whether the track still hits when it plays at the same loudness as everything else. Master for the sound, keep the true peak sensible, and let the platforms do their arithmetic.
— FAQ
How accurate is the measurement?
The engine implements ITU-R BS.1770-4 and EBU R128 directly and is unit-tested against the standards’ own reference signals — integrated loudness lands within 0.01 dB of the published anchors, loudness range matches the EBU test cases exactly, and true peak reads within a tenth of a dB on deliberate inter-sample-peak fixtures. Two small caveats come from the browser rather than the maths: browsers decode audio at the device’s playback rate, so a 96kHz master is measured at 44.1 or 48kHz (the effect on loudness and true peak is well under 0.1 dB, and the result card shows the rate actually measured), and lossy formats like MP3 decode fractionally differently between browsers, which can move the second decimal place. Files with more than two channels are measured on the first two, and the card says so when that happens.
Should I master to −14 LUFS?
No — master to what the music needs. −14 LUFS is a playback level Spotify and YouTube normalise towards, not a mastering target, and most club music sits far louder: techno, house and drum & bass masters commonly land between −9 and −6 LUFS integrated. Normalisation simply turns the playback gain down; it does not re-limit or repress your track. What actually matters is how your master sounds when it plays at the same loudness as everyone else’s — if it was only exciting because it was louder, normalisation removes that advantage, and if it is dense and clipped it will sound smaller, not bigger.
What’s the difference between LUFS and dB?
A plain dB figure is a ratio — it tells you how far a level sits from a reference, like dBFS measuring distance below digital full scale. LUFS is an absolute loudness scale built on that: the measurement filters the audio to match how human hearing weights different frequencies (K-weighting), averages energy over time, and gates out silence, so one number tracks perceived loudness rather than electrical level. Two tracks at identical peak dBFS can easily sit 6 LUFS apart. The units are the same size — 1 LU = 1 dB of gain — which is why platform deltas are expressed in dB.
What is true peak?
The loudest point of the reconstructed analogue waveform, which can sit between your digital samples. A master whose samples never exceed −0.1 dBFS can still reconstruct to over 0 dBTP, and lossy codecs (the AAC and Ogg your distributor transcodes to) can overshoot further and clip. True peak is measured by oversampling the signal 4x per ITU-R BS.1770-4 and reading the peak of the interpolated waveform — that is why it is quoted in dBTP and why a ceiling around −1 dBTP is the common recommendation for anything headed to streaming.
Is my audio uploaded?
No. The measurement engine is TypeScript running in a Web Worker inside your browser — there is no upload, no server-side processing, and nothing for anyone to store. Unreleased masters never leave your machine. It is also why results come back in seconds: a six-minute WAV measures in a few seconds because there is no transfer time at all.