The LUFS meter measures the loudness of a supported WAV recording so you can compare a short export with another version before delivery. It reports gated, K-weighted integrated loudness, maximum momentary loudness, sample peak, duration and channel count. The file stays in your browser, and a JSON report records the measurements. This LUFS meter accepts a specific 48 kHz mono or stereo WAV profile; it does not measure true peak, normalize the audio or certify that a streaming or broadcast delivery requirement has been met.
How to use the LUFS meter
Choose a supported .wav file or drag it into the file area. Selection starts local analysis. Try an example loads the bundled demonstration tone so you can see the measurement labels before checking a real recording. No target-loudness selector or gain control is offered: the interface analyzes the bytes rather than changing them.
Read Integrated loudness (LUFS) for the gated whole-file measurement, then compare the maximum momentary value and sample peak. Duration and channels help confirm that you selected the intended export. Download loudness.json to retain the measured values and scope note. A short excerpt and a complete mix may yield different integrated readings because they contain different material.
Supported WAV inputs and duration limits
The accepted profile is 48,000 Hz mono or stereo WAV with PCM 16-bit, 24-bit or 32-bit samples, or 32-bit floating-point samples. Duration must be at least 0.4 seconds and at most five minutes. The file-area size limit also applies, so an otherwise supported uncompressed export may need to be shorter or smaller to fit the configured limit.
MP3, FLAC, AAC, other sample rates, surround audio and unsupported WAV encodings are not decoded by this implementation. A .wav filename alone does not establish the internal format. Export an appropriate analysis copy from your audio editor when necessary, while retaining the original master. Changing an extension does not change the sample format or rate.
What integrated and momentary loudness tell you
This implementation applies K-weighting and uses 400 ms analysis blocks stepped by 100 ms. Absolute and relative gates exclude blocks below the supported loudness thresholds from the integrated calculation. The integrated result therefore describes the included programme material, not a simple arithmetic average of every sample’s amplitude.
The maximum momentary value is the largest supported 400 ms block reading. It can reveal a loud passage that a whole-file number hides. Two tracks with similar sample peaks can have different integrated loudness because sustained energy and programme structure differ. These measurements are useful diagnostic context, but they do not judge artistic quality or replace listening to the actual export.
Worked example: comparing two gain versions
Imagine a supported WAV whose integrated reading is −18 LUFS. If you raise the entire source by 3 dB in an editor without clipping and its gated block membership remains unchanged, the expected integrated reading is approximately −15 LUFS. Its sample peak rises by the same 3 dB. Reanalyze the exported result rather than assuming the editor’s gain value proves the final file has that reading.
This is an illustrative gain relationship, not a promise about the bundled sample or a recommendation to target −15 LUFS. Compression, limiting, changed silence and different edits can change the programme measurement in other ways. The correct destination target must come from the receiving service or specification for your particular job.

Sample peak is not true peak
Sample peak is the highest stored sample amplitude expressed in dBFS. True peak estimates the reconstructed signal between samples and requires a different measurement process. This LUFS meter reports sample peak only. Do not interpret a sample-peak result below zero as proof that intersample peaks or later lossy-encoding peaks stay below a delivery ceiling.
The LUFS meter does not calculate loudness range, perform a true-peak oversampling check or provide certified EBU or broadcaster compliance. For final delivery, use an appropriate production meter and the current receiving specification. A number displayed to two decimal places describes output formatting; it is not a claim of a certified accuracy tolerance.
Silence, rejected files and a practical media workflow
Silence or material below the gate can produce a Below gate / silence integrated result. That is more meaningful than inventing a finite loudness for a file with no included blocks. Truncated RIFF data, inconsistent WAV headers or invalid samples are input errors. Re-export from the original session instead of removing an error check.
This audio analysis does not generate video captions. In a captioned-media workflow, use SRT to VTT for a separate web-caption file or subtitle repair for supported subtitle structure fixes. Those steps do not affect the WAV loudness. Keep measurements tied to the exact audio version analyzed, and clear local results when a confidential recording is no longer needed in the tab.
Common LUFS meter errors and how to fix them
When the LUFS meter cannot use a file, it stops with one clear message instead of producing a half-finished download. These are the problems people run into most often, with the fix for each:
- Use 48 kHz mono/stereo PCM 16/24/32-bit or float32 WAV. The WAV uses another sample rate (such as 44.1 kHz), more than two channels, or an unsupported encoding. Export a 48 kHz copy from your editor.
- WAV duration must be between 0.4 seconds and five minutes. Integrated loudness needs at least one 400 ms block; longer programmes need to be measured in sections.
- The file ends unexpectedly. It may be truncated or corrupt. A chunk size points beyond the end of the file, typically from an interrupted export or copy.
- The WAV file is malformed or contains invalid samples. The RIFF header size is wrong, chunks are duplicated or float samples are not finite.
Privacy: the LUFS meter runs in your browser
LUFS meter processing happens with JavaScript in a background worker on this page. Your file is not uploaded to a FileFix server, and the results you download are created in your browser’s memory. Unreleased video scripts and captions stay on your computer. When you finish on a shared computer, clear the selection or close the tab.
For the underlying format rules, see the ITU-R BS.1770 loudness algorithm. The LUFS meter follows the subset described on this page rather than every optional feature in those documents.
Sources
The LUFS meter follows these specifications and format guides:
Questions about this tool
Can I measure an MP3 or a 44.1 kHz WAV?
Not in this implementation. It accepts the stated 48 kHz mono/stereo PCM or float32 WAV profile. Export a supported analysis copy in an audio editor.
Is sample peak the same as true peak?
No. Sample peak reads stored samples. This meter does not estimate between-sample true peak or certify a true-peak delivery ceiling.
Does the meter normalize my recording?
No. It analyzes the file and offers a JSON report. It does not change gain or download a processed audio file.
Why does a silent file have no finite integrated LUFS value?
Silence or material below the gate can leave no included analysis blocks. The result reports that condition instead of manufacturing a finite loudness.
Can I certify broadcast compliance from this report?
No. The tool has a bounded diagnostic scope, without true peak, loudness range or certified delivery-compliance verification.
Review the supported formats and limitations before using an output file. File processing happens in your browser.