Play several recordings in a row and the volume jumps: one interview is a whisper, the next blasts out, a music track is far louder than the voice before it. Turning the volume up and down by hand is tiring for listeners, and many platforms now adjust loudness themselves, often not the way you would like.
Normalizing fixes this at the source. This tool measures how loud each file actually sounds, then raises or lowers it to the loudness target you choose: -14 LUFS for streaming, -16 LUFS for podcasts or -23 LUFS for broadcast. Add one file or a whole batch, choose the target and an output format, and press Convert. The files are processed on your device.
LUFS, loudness and why peak normalization is not enough
Older normalizers simply raised a file until its loudest sample touched the maximum. That makes a file with one sharp peak, such as a cough or a drum hit, stay quiet overall, while a heavily compressed song ends up much louder at the same setting. Our ears respond to average energy over time, weighted towards the frequencies we hear best, not to single peaks.
LUFS, loudness units relative to full scale, measure exactly that. The measurement, defined in the ITU BS.1770 standard and used by EBU R128, averages the perceived loudness of the whole programme and ignores silent passages. Targets are negative numbers: -23 LUFS is the European broadcast standard, around -16 LUFS is the common recommendation for podcasts, and many music streaming services play tracks back at around -14 LUFS. One LU is one decibel, so moving from -20 to -16 LUFS means raising the level by 4 dB.
When to normalize loudness
Podcast episodes and interviews
Normalizing each episode to -16 LUFS keeps every release at the same listening level, even when guests were recorded on different microphones.
Music for streaming or a playlist
Matching tracks at -14 LUFS shows how they will sit next to each other on services that normalize, and stops one track from jumping out of a mix tape.
Broadcast and corporate delivery
Broadcasters and many video platforms in Europe ask for -23 LUFS. The broadcast target delivers that, with a safe true peak ceiling.
Two passes and one gain
Each file is read twice. The first pass measures its integrated loudness, true peak and loudness range with ffmpeg's loudnorm filter. The second pass applies a single gain to the whole file so that it lands on the target, which means quiet and loud passages keep the same relationship as before: the dynamics are not squashed. True peaks are kept below -1.5 dBTP, a little headroom that stops MP3 or AAC encoding from clipping. If reaching the target with one gain would push peaks past that ceiling, as can happen when raising a very dynamic recording to -14 LUFS, loudnorm falls back to gentle dynamic adjustment for that file. Files that contain only silence are processed in one pass.
Output format defaults to Same as input; a WebM recording, which this tool reads but does not write, becomes MP3. You can also choose MP3, WAV, M4A or FLAC. Compressed formats are written at 192 kbps, and the sample rate of the source is kept up to 48 kHz for compressed output and up to 192 kHz for WAV and FLAC. Processing uses the site's ffmpeg-based audio converter, a one-time download of about 31 MB, and results end in -normalized.
Bu araç nasıl kullanılır
- Drop one or more MP3, WAV, OGG, M4A, AAC, FLAC or WebM files on the box above, or click it to browse.
- Choose a Target loudness: -14 LUFS (streaming), -16 LUFS (podcasts) or -23 LUFS (broadcast).
- Choose an Output format, or keep Same as input.
- Press Convert and download each normalized file, or all of them with Download all (.zip).
Limitations
- Normalization sets the overall loudness of a whole file. It does not even out a quiet speaker and a loud one inside the same recording; that needs compression in an audio editor.
- Raising a quiet, noisy recording raises its noise too. Run Noise Reduction first.
- Saving to MP3 or M4A re-encodes the audio, which adds a small lossy step. Choose WAV or FLAC to avoid it.
- Video files are not accepted. Extract the sound first with MP4 to MP3.