Emulating Windows 10's Loudness Equalization

Long-time lurker, first-time poster. :slight_smile:

My question is: How should I proceed to emulate/replicate the behavior of Windows 10’s automatically-change-the-playback-volume-in-real-time aka “loudness equalization” by means of Audacity and offline processing?

Bonus points if “the understanding of human hearing” is used to reduce perceived volume differences. :wink:

Any help would be greatly appreciated.

Unfortunately, Audacity has no built-in tool to measure loudness in Terms of R128 or ReplayGain. (At least, none that I knew.)

If you want to adjust .mp3 or AAC (.m4a) files to a common loudness, use the program Mp3gain or AACGain instead!

If you want to adjust lossless files or your own recordings, you could measure the loudness of a track using an external tool such as ffmpeg (*) and subsequently use Audacity’s “Amplify” effect to change the loudness to the desired value.

(*) The command line program named “ffmpeg”. It is not part of the FFmpeg library, which accompanies Audacity.

I couldn’t find any information about how Microsoft’s Loudness Equalization works…

ReplayGain, MP3Gain, and Apple Sound Check all work by pre-scanning the loudness of the file and then adjusting the volume at playback time (or by “permanently” adjusting the volume in the case of MP3Gain). These use a loudness perception algorithm and they make one volume adjustment before the file starts playing so the dynamics of the program are not messed-with (loud parts of the song remain relatively loud and quiet parts remain relatively quiet).

Microsoft’s Loudness Equalization doesn’t seem to be associated with particular files or any particular player so I assume it’s some kind of automatic volume control. (If that’s what it is, it will continuously adjust the volume during the program/song making quiet parts quiet louder and loud parts quieter.)

Audacity’s has a Compressor, Limiter, and Leveler effects can be used to make the loud parts quieter and/or the quiet parts louder. These work on amplitude and timing without any loudness perception.

And, there is an optional [u]ReplayGain plug-in[/u] (which isn’t really ReplayGain but it uses the same/similar loudness perception algorithm ).

Thank you very much; I didn’t know that. I assume it comes closest to what was asked for.

I will take a look at the ReplayGain plug-in. Thank you all!