Free audio tool

Loudness meter.

Check a mix or master against broadcast and streaming loudness specs. Integrated LUFS, true peak and loudness range, measured to ITU-R BS.1770 and checked against the EBU test set. Your file never leaves your device.

WAV, MP3, AAC, FLAC. WAV is read sample-exact at its own rate.
Nothing is uploaded. Measured in your browser.
Published accuracy. Every synthetic EBU test case, results below.

Drop an audio file hereor choose a file · WAV, MP3, AAC, FLAC · processed on your device

From the studio

Didn’t pass, or not sure what to change?

We mix and master to delivery specs for broadcast, streaming and cinema. Send the file and the spec sheet, and you get it back measured and compliant. Delivering stems or a reference to us? Measure them here first so we start from the same numbers.

Start a project Mixing and mastering WhatsApp the team

Reading the result

What LUFS, true peak and LRA mean.

LUFS (and LKFS)

Loudness Units relative to Full Scale, defined by ITU-R BS.1770. The signal is K-weighted to approximate hearing, and quiet passages are gated out, so the figure describes foreground loudness. LKFS is the same unit under the North American name; one LU equals one dB.

True peak (dBTP)

The waveform's highest point between samples, found by oversampling. It can sit above the highest sample value, and lossy encoding can push it higher still, which is why specs cap it at −1 dBTP rather than 0.

Short-term and momentary

Loudness over the last 3 seconds and the last 400 milliseconds. EBU R 128 s1 limits short-term loudness on adverts to −18.0 LUFS, and French law limits it to −20 LUFS on commercials.

Loudness range (LRA)

How much the short-term loudness moves across the programme, in LU (EBU Tech 3342). UK DPP asks programmes to aim for no more than 18 LU, and France caps programmes over two minutes at 20 LU.

Broadcast: a requirement.

A broadcaster measures the delivered file and rejects it if it misses the spec. EBU R 128 sets −23.0 LUFS with true peak no higher than −1 dBTP; UK DPP narrows the tolerance to ±0.5 LU for non-live programmes; adverts add a short-term ceiling (−18.0 LUFS under EBU R 128 s1, −20 LUFS under French law). In the US, ATSC A/85 sets −24 LKFS with true peak at or below −2 dBTP, measured on dialogue for programmes and on the full mix for adverts. An advert can hit −23 on average and still fail on one loud passage.

Streaming: a playback adjustment.

Spotify never rejects a master for its loudness. It normalises playback to −14 LUFS, turning louder tracks down and lifting quieter ones only as far as their true peak allows. So a very loud master gains nothing on Spotify, and a quiet one with peaks near full scale can end up playing quieter than the tracks around it.

Step by step

How to check a file against a loudness spec.

01
Export the file you will actually deliver
Full length, at the sample rate and format going to the broadcaster or platform. Loudness is measured over the whole programme, so a cut-down or a clip measures differently from the full master.
02
Drop it into the meter
WAV, MP3, AAC (M4A) or FLAC. The file is read and measured in your browser; nothing is uploaded.
03
Read integrated loudness and true peak
Integrated loudness (LUFS) is the programme's overall level after gating. True peak (dBTP) is the highest level of the waveform between samples, which is what trips a broadcaster's peak limit.
04
Check the row for your destination
Each delivery check applies its spec as written, with the tolerance and a link to the source. Streaming rows tell you how the platform will change the playback level instead of pass or fail.
05
Fix the right thing
Too loud or too quiet overall is a level change. A true-peak failure is a limiter ceiling problem. A short-term failure on an advert means one passage is too loud even if the average passes, which usually needs the mix itself adjusted.

Accuracy

Tested against the EBU test set.

EBU Tech 3341 and Tech 3342 define the signals a loudness meter must read correctly, and the tolerance it is allowed. These are this meter’s readings on every synthetic case, at 48 kHz and 44.1 kHz. Two cases in each document use real programme recordings from the EBU’s own download, which are not part of this table.

Loudness meter readings on the EBU Tech 3341 and 3342 test cases
Case	Signal	Measure	Expected	48 kHz	44.1 kHz
3341 #1	Stereo 1 kHz sine, −23 dBFS, 20 s	Integrated	−23 LUFS ±0.1	−22.99	−22.99
3341 #2	Stereo 1 kHz sine, −33 dBFS, 20 s	Integrated	−33 LUFS ±0.1	−32.99	−32.99
3341 #3	10 s −36 / 60 s −23 / 10 s −36 dBFS	Integrated	−23 LUFS ±0.1	−23.01	−23.01
3341 #4	−72 / −36 / −23 (60 s) / −36 / −72 dBFS	Integrated	−23 LUFS ±0.1	−23.01	−23.01
3341 #5	20 s −26 / 20.1 s −20 / 20 s −26 dBFS	Integrated	−23 LUFS ±0.1	−22.98	−22.98
3341 #6	5.0: L/R −28, C −24, Ls/Rs −30 dBFS	Integrated	−23 LUFS ±0.1	−23.02	−23.01
3341 #9	(1.34 s −20 / 1.66 s −30) ×5	Short-term (steady)	−23 LUFS ±0.1	−22.99	−22.98
3341 #10	20 × (i·0.15 s silence, 3 s −23, 1 s silence)	Max short-term, 20 segments	−23 LUFS ±0.1	−22.99	−22.99
3341 #12	(0.18 s −20 / 0.22 s −30) ×25	Momentary (steady)	−23 LUFS ±0.1	−22.96	−22.96
3341 #13	20 × (i·20 ms silence, 400 ms −23, 1 s silence)	Max momentary, 20 segments	−23 LUFS ±0.1	−22.99	−22.99
3341 #15	Sine fs/4, 0.5 FFS, phase 0.0°	True peak	−6 dBTP +0.2 / −0.4	−6.02	−6.02
3341 #16	Sine fs/4, 0.5 FFS, phase 45.0°	True peak	−6 dBTP +0.2 / −0.4	−6.06	−6.06
3341 #17	Sine fs/6, 0.5 FFS, phase 60.0°	True peak	−6 dBTP +0.2 / −0.4	−6.04	−6.04
3341 #18	Sine fs/8, 0.5 FFS, phase 67.5°	True peak	−6 dBTP +0.2 / −0.4	−6.03	−6.03
3341 #19	Sine fs/4, 1.41 FFS, phase 45.0°	True peak	3 dBTP +0.2 / −0.4	2.94	2.94
3341 #20	fs/6 carrier + one fs/4 period at 1.0, offset 0	True peak	0 dBTP +0.2 / −0.4	0.03	0.03
3341 #21	fs/6 carrier + one fs/4 period at 1.0, offset 1	True peak	0 dBTP +0.2 / −0.4	0.01	0.01
3341 #22	fs/6 carrier + one fs/4 period at 1.0, offset 2	True peak	0 dBTP +0.2 / −0.4	0.05	0.05
3341 #23	fs/6 carrier + one fs/4 period at 1.0, offset 3	True peak	0 dBTP +0.2 / −0.4	0.07	0.07
3342 #1	LRA: -20 dBFS / -30 dBFS (20 s each)	Loudness range	10 LU ±1	10	10
3342 #2	LRA: -20 dBFS / -15 dBFS (20 s each)	Loudness range	5 LU ±1	5	5
3342 #3	LRA: -40 dBFS / -20 dBFS (20 s each)	Loudness range	20 LU ±1	20	20
3342 #4	LRA: -50 dBFS / -35 dBFS / -20 dBFS / -35 dBFS / -50 dBFS (20 s each)	Loudness range	15 LU ±1	15	15

How this meter works.

Measurement follows ITU-R BS.1770-5: K-weighting redesigned exactly for the file's own sample rate, 400 ms gating blocks with 75% overlap, the −70 LUFS absolute gate and the −10 LU relative gate. Loudness range follows EBU Tech 3342. True peak is measured by oversampling (8× below 88.2 kHz) with a windowed-sinc interpolator.

It is checked against every synthetic test case in EBU Tech 3341 and Tech 3342 at both 44.1 and 48 kHz, and against ffmpeg's EBU R 128 meter on every test file: 204 of 204 checks pass. True peak agrees with 32× brute-force oversampling to within 0.03 dB.

In the browser it is tested with real file uploads in Chrome, Firefox and Safari on desktop and phone, where every browser returns identical readings to a hundredth of a decibel — including a check that no audio, file name or result leaves the page.

WAV files are read sample-exact at their own rate. Compressed formats are decoded by your browser at the rate read from the file's header, so nothing is quietly resampled.

Built and tested by Abu Bakar Zahid, Founder, Managing Director & Music Producer at Elephant Audio. Last updated 24 September 2026.

Common questions

Frequently asked questions.

What LUFS should my song be for Spotify?

There is no required level. Spotify normalises playback to −14 LUFS: louder tracks are turned down, and quieter ones are lifted only as far as their true peak allows. Spotify advises keeping true peak below −1 dBTP, or below −2 dBTP for masters louder than −14 LUFS.

What loudness do TV broadcasters require?

In Europe, EBU R 128 sets −23.0 LUFS with true peak no higher than −1 dBTP. UK DPP requires −23.0 LUFS ±0.5 LU for non-live programmes, and adverts under EBU R 128 s1 also have a −18.0 LUFS short-term ceiling. In the US, ATSC A/85 recommends −24 LKFS with true peak at or below −2 dBTP; the meter checks US adverts and promos, while US programmes are measured on dialogue.

My advert measures −23 LUFS. Why was it rejected?

The integrated figure is only one of the checks. Adverts also face a short-term ceiling (−18.0 LUFS under EBU R 128 s1, −20 LUFS under French law), a true-peak limit, and in the UK a Clearcast phase check. A spot can hit −23 LUFS on average and still have one passage that is too loud.

Is LUFS the same as LKFS?

Yes. Both are the unit defined by ITU-R BS.1770 and measured by the same algorithm; the EBU calls it LUFS and ATSC calls it LKFS. A file measuring −23 LUFS measures −23 LKFS.

Is my audio uploaded anywhere?

No. The file is decoded and measured inside your browser, and nothing about it — not the audio, the file name or the results — is sent to a server.

How accurate is this meter?

It passes every synthetic test case in EBU Tech 3341 and 3342 at 44.1 and 48 kHz, agrees with ffmpeg's EBU R 128 meter on every test file, and gives identical readings in Chrome, Firefox and Safari. The method section above lists what was tested.

Why does my MP3 read slightly differently from the WAV?

Lossy encoding changes the waveform a little, usually by around a tenth of a loudness unit, and can raise true peak. Always measure the exact file you are delivering.

Which file formats can I check?

WAV (including 24-bit, 32-bit float and 5.1), MP3, AAC or M4A, and FLAC work in every modern browser. Ogg and Opus depend on the browser. Surround files are measured with the LFE channel excluded and surrounds weighted, as BS.1770 specifies.

Delivery specs, sourced

The specs behind the checks.

Every rule the meter applies comes from one of these reference pages, each written from its primary document.

Studio credit

Elephant Audio

Audio production · Sound design · Music production

LahorePKEngineering sound since 2018

Last updated: 24 September 2026

Message us on WhatsApp