Free tool · Video
Loudness Checker: Measure LUFS Online, No Upload
Open an audio or video file and this page reads its integrated loudness in LUFS, following ITU-R BS.1770, plus its sample peak. Spotify says it adjusts tracks to -14 dB LUFS and Apple recommends around -16 dB LKFS for podcasts, so you can see the gain each target needs. The file is decoded in your browser and is never uploaded.
Free, no account required.
Open an audio or video file
Your file never leaves your device. It is decoded and measured in this browser tab.
What LUFS measures
LUFS stands for Loudness Units relative to Full Scale, and one unit is one decibel. Unlike a peak meter, which shows the single highest sample, a loudness meter tries to say how loud something sounds. The method is defined in ITU-R BS.1770, and EBU R 128 builds the European broadcast rules on it. ITU writes the unit LKFS, EBU writes LUFS, and the numbers are the same.
The measurement has four steps. First the audio is filtered with a K-weighting curve, which boosts the highs a little and cuts the lows, to reflect how the ear responds. Second, the filtered signal is cut into blocks of 400 milliseconds that overlap by 75 percent. Third, each block gets a loudness from the mean square of its samples. Fourth, the blocks go through two gates: anything quieter than -70 LUFS is dropped, then anything more than 10 loudness units below the average of what remains is dropped as well. The loudness of the surviving blocks, averaged in power, is the integrated loudness.
The gates are why a podcast with long pauses is not measured as quiet. Silence and room tone between sentences are removed, so the number describes the speech. The same gating means two files with different amounts of pause can share a LUFS value while one would sound sparse, so loudness is a level statement, not a quality one.
As a check on the scale, a full scale 997 Hz sine wave on one channel reads -3.01 LUFS, and each 20 dB you lower it lowers the reading by 20. This page applies exactly that calibration. Channels beyond the first two are counted with a weight of 1.0 rather than the surround weights from the standard, so multichannel results are an approximation and stereo results are standard.
Reading the result is simple. Compare the integrated loudness with the target you care about. A file at -20 LUFS is 6 dB under a -14 LUFS target, so it needs +6 dB, and a file at -9 LUFS needs -5 dB. Because the gain is applied to the whole file, it moves the sample peak by the same amount, which is why the peak reading sits next to the loudness reading.
Platform loudness targets and where they come from
A target is the loudness a platform plays your audio at after normalization. Louder files are turned down, and on some platforms quieter files are turned up. Only two of the five platforms below state a number in their own documentation, and both pages were checked October 2026. For the other three the figure is a convention, repeated by mastering guides, and the table says so rather than presenting it as a platform rule.
| Platform | Target | Peak guidance | Basis |
|---|---|---|---|
| Spotify | -14 LUFS | True peak at most -1 dB | Platform documentation, checked October 2026 |
| Apple Podcasts | -16 LUFS | True peak at most -1 dB | Platform documentation, checked October 2026 |
| YouTube | -14 LUFS | None published | Convention, not stated by the platform |
| Amazon Music | -14 LUFS | None published | Convention, not stated by the platform |
| TikTok | -14 LUFS | None published | Convention, not stated by the platform |
Spotify's artist page says: "We adjust tracks to -14 dB LUFS, according to the ITU 1770 standard." Apple's audio requirements say to keep the true peak at -1 dB FS or lower and, for loudness, that "the overall loudness remains around -16 dB LKFS, with a +/- 1 dB tolerance". The Apple figure is a recommendation for Apple Podcasts.
For the three conventions: YouTube's help page on video volume describes a Stable Volume setting and gives no LUFS number. Amazon's help page for Amazon Music loudness normalization explains how to switch the feature on and gives no number either. We found no TikTok document that states a loudness target. Mastering at around -14 LUFS for those three is common practice, and it is also Spotify's published figure, so it is a sensible single target, but nobody here has promised it.
Two cautions when using the table. A platform that only turns loud audio down will not lift a quiet file, so mastering quieter than the target only costs you level against louder neighbours. And a target is not a quality bar: a dynamic recording at -18 LUFS is not worse than a crushed one at -9 LUFS, it is just played at a different gain.
Sample peak vs true peak
This page reports sample peak: the largest absolute sample value in the decoded audio, shown in dBFS. It does not report true peak. A digital file only stores values at fixed instants, but when it is turned back into a continuous wave, the wave can swing higher between two stored samples. A true peak meter estimates that by oversampling, usually four times, and reports in dBTP.
The practical effect is that true peak is never lower than sample peak, and can be a decibel or two higher on bright, heavily limited material. If you are aiming at a ceiling such as the -1 dB true peak that Spotify and Apple both state, leave margin: a sample peak of -1.0 dBFS does not prove the true peak is below -1 dBTP. Lossy encoding to AAC or Ogg Vorbis can add a little more.
We chose to show sample peak and say so rather than approximate the oversampling and risk giving a number that looks authoritative. Sample peak is still useful. It tells you how close the file is to digital full scale, and it tells you whether the gain a target needs will clip the file.
How to fix loudness that is off target
If the table says you need a negative gain, the file is louder than the target and the platform will turn it down. Lowering it yourself gives the same result with no processing by the platform. If it says you need a positive gain, check the sample peak first. Adding 4 dB to a file peaking at -2 dBFS pushes it past full scale, so you need a limiter or a dynamics change, not just a volume knob.
The quickest repeatable fix is ffmpeg's loudnorm filter, which normalizes to an EBU R 128 style target. The FFmpeg Command Generator has a Normalize loudness recipe that builds the command for any target you choose, copies the video untouched and caps true peak at -1.5 dBTP. It is a single pass, so the result lands close to the target rather than exactly on it.
After any change, drop the new file back into the checker. Measuring again is the only way to know the result, because limiters, encoders and normalizers each move loudness in ways the settings do not fully predict.
How to check the loudness of an audio file
Step 1
Choose an audio or video file
Press Choose file and pick an MP3, WAV, M4A, AAC or a video such as MP4. The browser decodes it on your device. If it cannot decode the codec, the page says so.
Step 2
Read the integrated loudness
The large number is the loudness of the whole file in LUFS after gating out silence. A file shorter than 400 ms or fully silent shows no measurable level.
Step 3
Check the sample peak
Sample peak is the highest sample value in dBFS. If it is near 0 dBFS, raising the level will clip unless you use a limiter.
Step 4
Compare against the targets
The table gives the gain in dB that moves your file to each platform's level. Apply that gain in your editor or with the ffmpeg loudnorm recipe, then measure again.
When you outgrow this tool
Linkeddit Studio normalizes audio to a LUFS target, and its agent can measure loudness, peak and speech regions before and after an edit.
This checker is free and complete on its own, with no account. Linkeddit Studio is the desktop editor for the audio itself. It normalizes audio to a LUFS target, and its agent can measure loudness, peak and speech regions before and after an edit, so it is not editing blind. Studio is a one-time purchase and your media stays on your disk.
FAQ
Loudness Checker (LUFS Meter) questions
Direct answers on LUFS, platform targets and peak measurement.
What is LUFS?
LUFS means Loudness Units relative to Full Scale. It is a measure of perceived loudness defined in ITU-R BS.1770, which weights the signal to match how the ear responds, then averages its power. One loudness unit equals one decibel. EBU R 128 uses the same measure and calls it LUFS, while ITU and some broadcasters write LKFS for the same quantity.
What is the LUFS target for Spotify?
Spotify's documentation for artists says it adjusts tracks to -14 dB LUFS according to the ITU 1770 standard. It also advises keeping a master below -1 dB true peak. Normalization happens on playback, so Spotify does not change your file; it turns loud tracks down and, within headroom limits, soft ones up.
What loudness should a podcast be?
Apple Podcasts recommends around -16 dB LKFS with a tolerance of plus or minus 1 dB and a true peak no higher than -1 dB FS, measured to ITU-R BS.1770. That is Apple's recommendation for Apple Podcasts. Other podcast apps may use different targets, so check the one your audience uses.
What LUFS does YouTube use?
YouTube does not publish a LUFS number on the help page that describes its volume features. That page describes a Stable Volume setting that is on by default, and nothing more specific. The widely repeated figure of -14 LUFS is a convention from mastering guides, so this page labels it that way. If you need a precise number, treat it as a good approximation, not a platform statement.
Is this a true peak meter?
No. It reports sample peak, the largest sample value in the decoded audio. A true peak meter oversamples the signal, usually by four times, to catch peaks that fall between samples and appear after conversion or encoding. True peak is never lower than sample peak, so a file that reads -1.0 dBFS here could read higher on a true peak meter.
Why does my loudness read differently in another meter?
Small differences of a few tenths of a loudness unit come from file decoding, resampling and how the gate is applied. This page decodes to 48 kHz in the browser, so a file recorded at another rate is resampled first. The implementation is calibrated to the BS.1770 reference, where a 0 dBFS 997 Hz sine reads -3.01 LUFS, and that value is pinned by a test. In our own checks on 48 kHz test signals it matched ffmpeg's ebur128 filter to within 0.1 LU. That was our test, not a certification. A compressed file can also decode slightly differently in different players.
Does a surround file measure correctly?
Not to the full standard. BS.1770 gives the surround channels extra weight, and this tool does not apply those weights. It treats every channel with a weight of 1.0 and adds their power. Stereo and mono files, which are what most video and podcast files are, are measured as the standard specifies.
Is my file uploaded?
No. The browser reads the file from your disk, decodes it with the Web Audio API on your device, and plain JavaScript measures it. There is no account and no network request for the file, so unreleased audio stays private.
More free tools
Every tool on the shelf is free, with no card and no trial. Most need no account at all. See all free tools.
Work out file size from bitrate, the bitrate that fits a size limit, or how long a video can run, with audio included.
Build a correct ffmpeg command for trimming, compressing, converting, cropping and more, with every flag explained and linked to the ffmpeg docs.
Pick a resolution, frame rate and SDR or HDR and get YouTube's recommended container, codec, bitrate and audio settings, with the source linked.
Reduce a resolution to its ratio, scale to a width or height with even dimensions, and see the crop or padding needed to reframe, such as 16:9 to 9:16.