Loudness Normalizer
Loudness Normalizer measures integrated loudness (ITU-R BS.1770 LUFS) and true peak in your browser and sets audio to -16 LUFS for podcasts, -14 for Spotify, -23 for broadcast or the ACX range.
At a glance
- Accepts
- MP3, M4A, AAC, WAV, FLAC, OGG and Opus audio, or the sound in a video file
- Outputs
- MP3 (192 kbps), WAV (16-bit) or M4A when the browser has an AAC encoder; the sample rate is kept where the format allows
- Where it runs
- Entirely in your browser, on your device
- What leaves your device
- Nothing - the file never leaves your device
- Limits
- Changes the overall level only (no compression); a limiter can leave the result slightly under target, shown in the after numbers
- Account
- Not required - there is no sign-up
- Price
- Free, with no watermark and no usage cap
- Made by
- Visrut
- Last updated
- 13 October 2026
Measure how loud a recording really is - integrated loudness in LUFS, true peak and RMS - and bring it to the level a platform expects: -16 LUFS for podcasts, -14 for Spotify, -23 for broadcast, or the ACX range for audiobooks. The measurement follows ITU-R BS.1770, and the before and after numbers are shown so you can check the result.
How to use it
- Drop in an audio file. It is measured as it loads.
- Pick a target, or Custom to type your own LUFS and true-peak limit.
- Press Normalize, then compare the before and after loudness and peaks and download the file.
Honest limits
- It changes the overall level only. It doesn't compress or even out quiet and loud passages, so a recording with very uneven speakers stays uneven.
- The after numbers are measured on the processed audio before MP3 or M4A encoding, which can move peaks by a fraction of a dB. WAV is saved exactly as measured.
- Surround files are measured with BS.1770's channel weights assuming the usual 5.1 order (L, R, C, LFE, Ls, Rs).
- Everything runs in your browser tab; the file is never uploaded.
- Long files are read and processed a piece at a time, so the length of the recording isn't the limit. The finished file is held in memory until you save it: about 86 MB per hour as MP3, and about 690 MB per hour of 48 kHz stereo as WAV, which a phone may not manage.
- MP3 and M4A are re-encoded at 192 kbps, which is a lossy step (like any edit to an MP3). WAV is uncompressed 16-bit.
- M4A is offered only when the browser has an AAC encoder (most desktop Chrome, Edge and Safari do; Firefox does not).
Sources (checked 13 October 2026)
- ITU-R BS.1770 - Algorithms to measure audio programme loudness and true-peak audio level
- EBU R 128 - Loudness normalisation and permitted maximum level (-23 LUFS, -1 dBTP)
- Apple Podcasts - Audio requirements (-16 LKFS +/- 1 dB, true peak -1 dBFS)
- Spotify for Artists - Loudness normalization (-14 LUFS, true peak below -1 dB)
- ACX - Audio submission requirements (-23 to -18 dB RMS, peaks below -3 dB, noise floor below -60 dB RMS)
Frequently asked questions
How is loudness measured?
With the method in ITU-R BS.1770, which EBU R 128 and the streaming services build on: the audio is K-weighted (a filter that roughly follows how loud we hear different pitches), measured in 400 ms blocks, and blocks of silence or near-silence are gated out so pauses don't drag the number down. The result is integrated loudness in LUFS (LKFS is the same unit).
What loudness should a podcast be?
Apple Podcasts recommends about -16 LKFS, give or take 1 dB, with true peaks no higher than -1 dBFS. That is the Podcast preset.
Why -14 LUFS for Spotify and YouTube?
Spotify states that it plays tracks at -14 LUFS by default and suggests keeping true peaks below -1 dBTP. YouTube does not publish its target; -14 LUFS is the level it is commonly measured at, so treat that part as a rule of thumb.
Is there an Apple Music preset?
No. Apple doesn't publish the level Sound Check normalises music to, so we don't present a figure as official. Use Custom if you have a target in mind.
How does the audiobook (ACX) preset work?
ACX, the audiobook service behind Audible, asks for each file to measure between -23 and -18 dB RMS with peaks below -3 dB. The preset sets the plain RMS level to -20 dB and holds peaks under -3 dB. ACX also requires a noise floor under -60 dB RMS, which this tool doesn't change.
What is true peak, and why is it different from the sample peak?
The sample peak is the largest sample value. When the samples are turned back into a smooth wave, it can rise between two samples, so the real peak is a little higher. True peak estimates that by oversampling four times, the way BS.1770 describes. Both are shown.
Why did the result land slightly under the target?
When reaching the target would push peaks past the limit, a limiter turns those peaks down, which takes a little loudness with it. The after numbers are measured on the result so you can see exactly where it landed.