The audio levels for playbacks are generally low. When I export my piece to mp3 to share it with someone, I’d like the generated file to be amplified as much as possible (so the loudest part of the piece is just below the maximum).
In my current workflow, I either put the mp3 in Audacity and choose Effects - Amplify (the default value that Audacity picks is the maximum value without oversteering it), or to use ffmpeg with:
$ ffmpeg -i audio.mp3 -af volumedetect -f null /dev/null
# ...
[Parsed_volumedetect_0 @ 0x600003a74000] max_volume: -29.0 dB
# ...
$ ffmpeg -i audio.mp3 -af "volume=29" output.mp3