What is Normalize Audio – Should You Normalize Audio?

Normalization in audio mixing terms relates to a process to affect the uniform volume of a track or tracks in your mix. Let’s identify what is normalize audio in mixing and music terms in general and more importantly should you normalize audio in your mix?

What is Normalize Audio

Let’s identify what is normalize audio in the context of mixing and audio production in general.

Normalization in audio is actually a term which carries a number of different connotations, but ultimately it boils down to two different types: peak normalization vs average normalization.

Let’s identify each one, starting with the more practical average normalization.

Average Normalization

Average normalization is more about the average volume of a song and applies more in the audio mastering process where the mastering engineer is trying to get all of the songs on an album to be the same perceived loudness. This gets into LUFS territory, so check out that article for more about how we quantify the perceived loudness of music.

This is the same kind of normalization and concept which you’ll see as a feature if you go into the settings of your music streaming platform of choice. In Spotify, there’s an option to turn this on in the Preferences:

When ticked, this automatically adjusts the volume of the songs which you play so that they’re all roughly the same perceived volume. This setting is useful as obviously different songs are mastered to different final volumes, not to mention music from past decades is on average a lot quieter than modern music (thanks Loudness War).

This setting is turned on by default as it keeps you from having to adjust the volume of your speakers/device when listening to music and one song is louder or quieter than the last.

Speaking which, check out my tutorial on what LUFS to master to for Spotify and other popular streaming platforms.

That’s all average normalization, but what about peak normalization?

Peak Normalization

Peak normalizing a track in your mix refers to uniformly turning the gain up or down to reach a common peak level. Typically this peak level is 0dB (the digital point of clipping).

Here are three clips of audio before and after normalization has been applied to bring each track’s peak to 0dB:

You’ll see that after normalization, the waveforms of all three tracks all look much more similar in terms of volume. The audio hasn’t been altered in any way like would happen with a compressor, the gain has simply been pulled up.

Because each track will only be brought up as much as it takes to get the loudest peak to 0dB, it stands to reason that the top track was turned up the most because it had the most headroom below clipping.

Most DAWs have a method built in for normalizing one or more tracks at once, but when should you normalize in mixing?

Should You Normalize Audio

Now the important question, SHOULD you normalize audio? While it’s an important aspect in the mastering process to ensure that the average perceived volume of each song in a collection (like an album) is comparable, normalizing audio in the mixing process is unnecessary.

In fact, it’s counterproductive in the sense that it generally defies gain staging, the process of keeping all of the tracks in your mix at optimal input levels for plugin processing and creating headroom.

Even if you’re able to set the target for any automated normalization your DAW is capable of to -10dB peaks, you’ll still need to go in and make some manual adjustments as the dynamics of the various tracks in your mix will be all over the place, anyway.

Anyone who recommends peak normalizing audio, or turning everything up to just below the point of clipping, typically does so with the argument that it creates louder and therefore better mixes and masters.

There are far more effective ways to make your music louder, not to mention safer ways as pushing all of your tracks to 0dB peaks doesn’t account for inter-sample peaks which will cause clipping during the DAC process (I go into detail on this in my overview of true peak limiting).

In summary, the verdict on whether you should normalize audio or not is there’s generally no good reason to do it. For what it’s worth, after mixing for roughly 15 years now, peak normalizing isn’t something I’ve ever done.

That said, average normalization has its place in mastering or simply used as a setting when you’re listening to music in your streaming service of choice.