What does Audio Normalizer do?
Tubelexity normalizes local audio with two real processing modes. Peak normalization measures the highest decoded sample and applies gain toward a dBFS target. Loudness normalization uses FFmpeg's loudnorm filter for integrated LUFS with a true-peak ceiling. This adjusts playback level; it is not mastering and cannot repair clipping, noise or poor recording quality.
How to use Audio Normalizer
- 1Choose a supported audio file and wait for local peak analysis.
- 2Select Peak normalization or LUFS loudness normalization.
- 3Choose a target, clipping prevention and peak ceiling.
- 4Process locally, review the before/after target report, preview and download.
Supported inputs and outputs
Inputs
MP3, WAV, M4A/AAC, OGG, WebM audio
Outputs
MP3, PCM WAV, AAC in M4A
Practical uses
- Matching voice-over peak levels
- Preparing podcast loudness
- Adding safe playback headroom
- Reducing inconsistent delivery volume
Browser and media limitations
- Normalization is not mastering and does not improve recording quality by itself.
- Already clipped samples cannot be restored.
- LUFS processing uses FFmpeg loudnorm and may take longer than fixed peak gain.
- Peak analysis requires browser-decodable audio; unsupported previews can still be converted but peak mode is disabled.
- Exports are re-encoded.
How local processing works
Audio processing runs in this browser with lazy-loaded FFmpeg.wasm. The source and output are not uploaded to Tubelexity or a processing server.
Frequently asked questions
What is peak normalization?
It applies a fixed gain so the highest measured sample reaches the selected dBFS target.
What is LUFS normalization?
It targets perceived integrated loudness with FFmpeg loudnorm and a true-peak ceiling.
Which LUFS target should I use?
Common presets include -14, -16, -18 and -23 LUFS; requirements depend on the destination.
Can normalization fix clipped audio?
No. It can lower level but cannot reconstruct samples already flattened by clipping.
Does this master my audio?
No. Mastering also involves tonal, dynamic and contextual decisions beyond normalization.
Is audio uploaded?
No. Analysis and processing remain in your browser.