Free LUFS meter. Measure before changing the level.

A peak reading answers how high a decoded sample reaches. Integrated loudness uses frequency weighting and gated energy across the measured selection. Two files can share a sample peak and still have different integrated loudness. Listen to and download an equal-peak comparison to try that distinction with the meter.

For an original numerical check, our six-second 1 kHz mono tone at sample peak −20.0 dBFS measured −23.00 LUFS. Putting the same tone on both stereo channels measured −19.99 LUFS, while the sample peak stayed −20.0 dBFS. FFmpeg's independent ebur128 filter reported −23.0 and −20.0 LUFS for those files. These are synthesized measurement checks, not music-quality ratings or AI isolation results.

Keeping both channels matters. This meter adds their weighted energies rather than averaging their waveforms into mono. Equal opposite-polarity stereo tones therefore remain measurable; a downmix can cancel them. The channel splitter and its listening examples explain that separate operation.

The implementation uses 48 kHz K-weighting, complete 400 ms blocks spaced 100 ms apart, an absolute −70 LUFS gate, and a relative gate 10 LU below the absolute-gated result. Incomplete trailing blocks are discarded. The EBU Tech 3341 reference describes integrated loudness gating and the wider requirements of a complete EBU-mode meter.

This tool implements the integrated measurement for mono and stereo, plus sample peak. It does not provide every measurement required by that wider specification. There is no true-peak estimate, loudness range, multichannel weighting, microphone calibration, or compliance badge. Decoding or resampling can change the samples compared with a meter operating directly on another source format.

Silence is not a finite LUFS reading. If no complete block passes the gate, the tool shows No reading and retains a separate sample-peak result where possible. Quiet passages excluded by loudness gating are still part of your recording; the tool neither deletes them nor determines whether you should edit them.

Read the actual delivery requirement for your project. This meter does not predict what a streaming service will do or tell you that one universal LUFS value suits every recording. A final delivery may need additional true-peak and other checks that are outside this tool.

If your task is one sample-peak adjustment, the free peak normalizer can apply linked gain and export a WAV. It does not set a LUFS target. To adjust integrated loudness instead, use the free LUFS normalizer and check its measured export; peaks may constrain the requested gain. Use the peak and loudness normalization guide to choose the operation before changing levels.

For an isolation comparison, listen to the original and processed recording at similar playback levels. Also check quiet words, overlap, and tone: a louder or numerically closer candidate can still lose wanted detail. Keep the original and measure the final export with the appropriate tools when a delivery specification matters.