Browser 106

max 1GB

Inspector media/lufs-meter

LUFS Loudness Meter

Overview

A gated loudness meter that follows ITU-R BS.1770-4 exactly: K-weighting, 400 ms blocks at 75% overlap, an absolute gate at −70 LUFS and a relative gate ten units below the mean of what survives it. The file is decoded at 48 kHz specifically so the recommendation's own coefficient table applies verbatim, and the implementation is checked against EBU Tech 3341 cases 1 to 6 to within 0.1 LU. Alongside the integrated figure you get the loudness range, the momentary and short-term traces over time, and a true peak measured with 4× oversampling — which is the number that tells you whether a master will clip the first encoder it meets, and the one a sample-peak reading cannot give you.

Check a master before releaseSee the integrated loudness and true peak against the platform you are delivering to.
Compare a mix to a referenceMeasure both and see the loudness and range side by side.
Deliver to broadcastConfirm a programme meets EBU R128 at −23 LUFS before you send it.

Set

Media Tools 16

Neighbours

01 Audio Converter 02 BPM Detector 03 Compress Video 04 Extract Audio from Video 05 Audio HLS Stream Generator 06 HLS Stream Generator

Specification

Command: media.lufs_meter
Set: Media
Input: MIXED
Output: —
Max file: 1GB
Batch: single file

MPEG WAV FLAC MP4 OGG OPUS AAC WEBM

Runtime

Engine: ffmpeg-wasm 0.12
Threads: — workers
Isolated: —
WebGPU: —
Memory: —

Transport 0 B

The file is opened, worked on and written back inside this tab. No upload step exists in this tool — there is nowhere for the bytes to go.

Keys

Search: ⌘K
Panels: ⌘1 · ⌘2
Full width: F
All keys: ⌘/

How it works

Drop a mix or a master

Decoded on your device at 48 kHz. Nothing is uploaded.

Read the integrated figure

One number for the whole programme, gated so leaders and fades do not drag it down.

Compare against a target

Pick the platform you are delivering to and the target is drawn across the meter.

Frequently asked

How do I know the numbers are right?

The implementation is verified against EBU Tech 3341 cases 1 to 6 — the reference test signals the EBU publishes for exactly this — and agrees to within 0.1 LU. The loudness range is checked against the four EBU Tech 3342 cases. The file is decoded at 48 kHz so the coefficients in BS.1770-4 apply as printed, with nothing re-derived.

What is the difference between true peak and sample peak?

Sample peak is the largest number in the file. True peak is the highest level the reconstructed waveform reaches between those numbers, and it can be meaningfully higher — a master reading −0.1 dBFS at the samples can hit well above 0 dBTP between them and clip a converter or a lossy encoder downstream. This tool reports both, and the gap between them.

Which target should I use?

Most streaming services normalise to around −14 LUFS and turn louder masters down, so mastering far above that buys nothing but lost dynamics. Broadcast in Europe is −23 LUFS under EBU R128. The targets are listed with their values so you can see what you are aiming at.

Is my audio uploaded?

No. The decode and every measurement happen in your browser, in a worker.

Keys

Search: ⌘K
Panels: ⌘1 · ⌘2
Full width: F
All keys: ⌘/

Related · Media Tools