Normalization is the process of adjusting a recording's volume to hit a consistent target level. It's why a well-produced playlist or podcast feed doesn't send you lunging for the volume knob between tracks. There are two kinds, and knowing the difference clears up most of the confusion around it.
Peak normalization finds the single loudest instant in your audio and raises (or lowers) the entire file so that peak sits at a chosen ceiling — often 0 dB, the maximum before distortion, or a hair below it. Everything moves up or down by the same amount, so the balance of the audio is untouched; it's just made as loud as it can go without clipping. This is the quick, safe way to make a quiet recording usably loud.
The problem with peak normalization is that one brief spike can stop a whole track from getting louder, even if it feels quiet overall. Loudness normalization measures perceived loudness instead — how loud the audio actually sounds to a person over time — using a unit called LUFS (Loudness Units Full Scale). It then adjusts the file to hit a target loudness, which is how streaming platforms keep every song and episode at a comparable volume.
| Where | Typical target |
|---|---|
| Music streaming (Spotify, YouTube, etc.) | around −14 LUFS |
| Podcasts (Apple and common guidance) | around −16 LUFS |
| Broadcast TV / radio (EBU R128) | −23 LUFS |
You don't have to hit these exactly — platforms adjust playback themselves — but aiming near the right target means your audio plays back at a sensible, consistent level rather than noticeably louder or quieter than everything around it.
These are often confused. Normalization moves the entire file up or down by a fixed amount — it changes the overall level but keeps the dynamics (the difference between loud and quiet parts) intact. Compression actively reduces that difference, making loud parts quieter and quiet parts relatively louder, which evens out a performance. Normalization is a single overall adjustment; compression reshapes the audio moment to moment. For simply making a clip a consistent, appropriate volume, normalization is what you want.
In the AudioTrim editor, the Normalize button does this in one click, and you can fine-tune afterward with the volume control.
No — it only changes level, not the underlying audio data. The one side effect is that raising a quiet file also raises its background noise.
Sort of — it makes audio louder to a defined target rather than by an arbitrary amount, which is what keeps multiple files consistent with each other.
Around −16 LUFS is the widely used target for podcasts. Music streaming tends to sit near −14 LUFS.
For a straightforward one-click normalize, no — you can do it in the browser. Dedicated loudness metering to hit an exact LUFS number is a job for specialized tools, but most people don't need that level of precision.
Related: What bitrate should you use? · Why audio clicks when you cut it · All guides