Stop guessing whether your track is loud enough for streaming. Drop an MP3, WAV or FLAC file onto the tool above and get two numbers that matter: integrated LUFS (perceptual loudness, measured to EBU R 128 / ITU-R BS.1770) and true peak in dBTP. Then pick a target — −14 LUFS for streaming platforms, −9 LUFS for club/EDM, or −16 LUFS for podcast & radio — click it, and download a normalized 16-bit WAV with the exact gain applied. If you are not sure what ELSE is wrong with a file before committing to a full master — tempo drift, an odd key, phase issues in a stem stack — run the same file through our free BPM & key analyzer: it reads BPM, key and Camelot code locally in your browser (nothing uploaded) so you know exactly which side of the problem loudness is not.
To be clear about what "basic mastering" covers here: this is the loudness half of the job — measuring and normalizing to your target level with a true-peak ceiling enforced. It does not run an EQ/compressor/mastering chain; if you want full tonal shaping, that is exactly what the AI mastering service below the tool is for (3 free processes every day).
Everything happens inside your browser tab: decoding, loudness metering and WAV encoding run locally on your device. The audio file is never uploaded anywhere, there is no account and no waiting in a queue.
What the tool shows you
| Metric | Meaning | Why it matters |
|---|---|---|
| Integrated LUFS | Perceptual loudness averaged over the whole track (K-weighted, gated) | This is the number streaming platforms normalize against — Spotify and YouTube target −14 LUFS, Apple Music plays back around −16 |
| True peak (dBTP) | The highest possible instantaneous value between sample points, via 4× oversampled interpolation | Inter-sample peaks can exceed 0 dBFS even when every sample reads "fine" — a common cause of clipping after AAC/MP3 re-encoding on the platform side |
| Sample peak (dBFS) | The highest raw sample value | A quick sanity check; always ≤ true peak by definition, and often much lower |
A track measured above roughly −9 LUFS integrated will usually be turned down by streaming normalization — you lose dynamic range for no loudness gain. A track far below target (for example a quiet demo at −25 LUFS) plays quieter than the rest of the catalog until the platform boosts it, which also pumps up any noise floor in the recording.
How normalization works here
The tool computes the loudness difference between your file and the chosen target, applies that gain, then enforces a −1 dBTP ceiling: if the corrected signal would exceed it, an additional peak-limiting gain keeps everything under the limit so you keep maximum headroom for lossy re-encoding. The result is exported as 16-bit PCM WAV at the file's original sample rate and channel count — the same two-step approach (loudness gain + true-peak ceiling) used in professional EBU R 128 workflows.
For a release, measure → normalize → verify: re-measure the normalized file and confirm it lands within ~0.5 LUFS of target with true peak at or below −1 dBTP before uploading to your distributor. For deeper background on how loudness normalization actually works across platforms — and why true peak headroom is not optional — see our guide LUFS vs True Peak: Loudness Normalization Explained.
Frequently Asked Questions
Is my audio file uploaded to a server?
No. The tool runs entirely in your browser using the Web Audio API. Decoding, loudness measurement and WAV encoding all happen on your device; nothing is transmitted over the network, no account is needed and there is no storage involved at all.
Which audio formats can I use?
MP3, WAV and FLAC are the primary targets, and most browsers will also decode OGG, Opus and M4A/AAC files through the built-in codec stack. Any length works — measurement is done over the full file in real time on a modern machine, so even multi-minute tracks finish in seconds.
What exactly does "−14 LUFS" mean for Spotify or Apple Music?
Streaming services normalize playback to a target loudness (around −14 LUFS for Spotify and YouTube, around −16 LUFS for Apple Music). If your upload is louder than the target it gets turned down; if quieter, it gets boosted. Normalizing to the target yourself means you hear in your own mastering decisions what listeners actually get — without a platform applying gain on top of peaks that were too close to zero.
What should my true peak be?
Keep it at or below −1 dBTP (−2 dBTP is extra safe for loud material). Inter-sample peaks above 0 dBFS frequently cause audible clipping after the platform re-encodes your track to AAC/MP3, because lossy codecs can push peaks several tenths of a decibel higher. The tool's normalization step automatically limits against the −1 dBTP ceiling when needed and tells you if it had to.
16-bit or 24-bit WAV? Which one should I deliver?
The tool exports 16-bit PCM, which is perfect for checking loudness before you commit — a normalized 16-bit file at −1 LUFS has about 90 dB of usable dynamic range, far beyond anything a platform's re-encoding will touch. If your master benefits from extra precision (very dense material, heavy limiting), generate the final delivery in your DAW at 24-bit using the same numbers this tool reports: target integrated loudness and true peak ceiling are identical either way.
Can I use this to match the loudness of a reference track?
Yes — that is one of the best uses. Load the commercial release you want to sound like, note its integrated LUFS, then load your own mix and normalize toward the same number (or apply the reported gain in your DAW). Matching perceived loudness level is the single fastest way to make an unmastered track sit competitively next to a finished one before investing in full mastering.