Normalize your audio
for any platform.

LUFS stands for Loudness Units relative to Full Scale. It's a measure of how loud audio sounds to human ears — not just how loud it peaks.

Think of it like the difference between average brightness and the brightest pixel in a photo. A photo can have one extremely bright pixel (a specular highlight) while the overall image is quite dark. Audio works the same way: a single loud transient (a snare hit, a clap) can peak near 0 dBFS while the song's perceived loudness is much lower.

LUFS measures perceived loudness by applying a K-weighting filter — it de-emphasizes bass frequencies (which sound louder at lower levels) and boosts presence frequencies, matching how human hearing actually works. It then measures the average power of the signal over time, ignoring the very quiet passages between notes.

Every major streaming platform — Spotify, Apple Music, YouTube, Tidal — automatically adjusts the playback volume of every track to a target level. This is called loudness normalization, and it's always on.

This creates a trap for audio creators:

The solution is to deliver audio that's already at the target level. When you hand the platform a track at exactly −14 LUFS (Spotify's target), they have nothing to adjust. Your audio plays back exactly as you intended.

This tool measures your audio's integrated loudness, calculates the exact gain needed to hit the target, and applies it — preserving 100% of your dynamic range in the process. No compression, no limiting. Just a linear volume shift.

Everything runs in your browser using the Web Audio API and WebAssembly (FFmpeg compiled to WASM). Encoding to MP3, AAC, FLAC, or OGG happens client-side — no files are transmitted to any server because there is no server.

The loudness measurement implements the full BS.1770-4 standard: K-weighting IIR filters → 400ms overlapping blocks → absolute gate at −70 LUFS → relative gate at −10 LU below ungated loudness → integrated result.

Digital audio is stored as discrete samples. Between those samples, the actual waveform can reach levels higher than any individual sample — these are called intersample peaks or true peaks.

When a DAC (digital-to-analog converter) reconstructs the continuous waveform from samples, these peaks become real. If they exceed 0 dBFS, the output clips — you hear distortion even though your audio file's sample values looked fine.

This problem is especially acute after lossy encoding (MP3, AAC). The encoding process can create intersample peaks that didn't exist in the original. A WAV file with a sample peak of −1 dBFS can encode to an MP3 with a true peak of +0.5 dBTP.

Your audio files are processed entirely in your browser. The loudness analysis runs in a Web Worker thread. The encoding runs in WebAssembly (FFmpeg compiled to WASM). Neither of these technologies sends data anywhere.

We have no server to receive your files. There is no analytics, no tracking, no accounts. The only external requests this page makes are to load the Google Fonts CSS and the FFmpeg WASM binary from a CDN — neither of which involves your audio.