Is it possible to normalize audio based on AVERAGE instead of PEAK?

[sebastian_2840]

I have multiple voice recordings from many different sources and consequently at many different sound volumes. I want to bring them all to the same audio level.

I know I can use the Audio Gain menu to "Normalize all peaks" to a certain db level, but the problem is that only considers the peak of each individual clip.

So for example if I have some video that for 99% of its duration peaks at -10 dB, but then at some point for one second has one audio spike that goes up to -2 dB. When I normalize everything to, for example, 0 dB, that clip will increase its gain by +2 dB. Which makes that one peak go to 0dB indeed... but then the rest of the clip will still peak at -8dB, which is way below the 0 dB that I want.

Is there some way to normalize based on average volume instead of max peak? It can also involve Adobe Audition if needed.

[R Neil Haugen]

Have you tried the Auto Match in the Sound panel?

[PaulMurphy]

Measuring your audio's "loudness" rather than peak levels gives you a more accurate way to normalize your audio.

In the Essential Sound panel, you can match all your dialogue to -23 LUFS (Loudness Units Full Scale) by using the Auto Match button. Here’s a guide on how to do this.

-23 LUFS is a good standard for broadcast mixing, but it may be too low for platforms like YouTube. For YouTube, I recommend using Loudness Normalization during your export. Here’s a video on how to do this specifically for YouTube.

[MyerPj]

Thanks Paul, good tutorial on YouTube audio! 🙂