The question of the track gain vs. album gain has been discussed in other threads such as:
Plex currently uses the album gain as the track gain. This makes an annoying experience in terms of volume gaps when switching from recent to old recordings and from regular albums to compilations. This is my empirical “finding” after listening to hundreds of songs on the Samsung TV app and with Plexamp (mobile and desktop). I pledge you to reintroduce the support for different album and track gain values and add an option to prefer album gain or track gain, as Winamp, Banshee (on Linux) and Lollypop (on Linux) did or do.
I’ve been using the (ReplayGain) track gain value for years and to me it works perfectly. I’ve adopted Plexamp for one year as my main player and the main irritant is the expectation of “gain problems” when songs from different epochs are played, a problem that was solved for years for me using the ReplayGain track gain with other players.
Please consider at least that the album gain is not a universal solution and if you are not moving away from it, please add a support and setting to prefer track gains over album gains, at least in Plexamp. I am willing to work for free on this feature. This is a must-have :). Let me know if I can help. Thanks!
With hopefully more people trying Plexamp I think the big thing they will notice is the lack of separate track and album gain.
PLEASE, look at revisiting this again and ideally implementing replaygain.
I have used Plexamp’s current loudness levelling, and it just grates - I am forever lowering and highering volume level’s when playing in a non album order.
Come on, you know it makes sense 
1 Like
Please implement support for the ReplayGain functionality. It can be disabled by default if you want. I just want the option. Some tracks are way too quiet and trying to turn them up just makes the next track obscenely loud. Playing these tracks in something that supports the ReplayGain tags it all sounds nice and level which is what I expected when getting a Plex pass that offers “loudness leveling”.
1 Like
Years back @OttoKerner mentioned that this decision was backed by solid research. But it was only tested on 24 songs with 38 subjects, where according to the paper 80% of the subjects
preferred album normalization over track normalization. Moreover the paper was only recommending album normalization to be the default, not that it should be disabled in anyway.
Personally, I’ve used replaygain for ever before getting plex media server, and I never had to expect getting shocked by loud tracks like I do now. These facts are clear. The devs offered the option to turn sweet fades off (which imho never turned out to be sweet), it’s not unreasonable to let users have the option to use track RG info.
1 Like
Something needs to change. I added music to my server installed Plexamp and played music. The volume difference between tracks when shuffling is unacceptable.
“We must get rid of the shuffle option.”
All joking aside, I had assumed Plex would have worked out volume issues long ago, and the default settings would give a better experience.
Is Plex actually using any replay-gain value? I have encoded albums with replay-gain tags, but I have not found any values in the info xml.
They are not using ReplayGain at all from what I can tell. They use their own album-only info from the “Analyze” option under “Manage Library”. This really should be a configurable option.
This is driving me insane.. I have large compilations which the source has plenty of variation track to track, so I use replaygain to do a track level analysis, and playing back in foobar2000 set to honour track gain and it does a pretty good job at equalising volume levels..
Plexamp however, well, it’s shockingly bad, I’m walking along enjoying one track and set the volume to match, then the next track may be whisper quiet.. looking at the track level gains, one is -1db, (the quiet one), the other -10db (the loud one).. so quite some difference, and foobar deals with it as expected..
C’mon Plexamp, I love the app/plex ecosystem and am fully bought in, but this is very very annoying and very easy to fix.. I get you think you can do a better job than replay gain, and claim more advanced algorithms, but its clear you need to just look at these obvious edge cases and get these sorted.. having 9db of track gain difference is not some nuance of an advanced algorithm, it’s a sledge hammer to the nuts kind of obvious..
I’m fine with you doing your own algorithm, but surely for compilations you could allow per track vs per album analysis?
