seriously, does playback use ReplayGain correctly for you?

[winamper]

Try this:

Take any 2 songs that have very different loudness levels.

Use WACUP to scan them with ReplayGain and write the gain information to their tags.

Play the songs in WACUP.
Then play them in Foobar2000.

My experience:
The difference between the how the 2 players output the songs is stark.
With WACUP, I still got a very soft and a very loud song (in terms of volume levels). I actually couldn't tell the difference in the loudness before and after ReplayGain.
With Foobar2000 the played levels are more or less equal (for the songs).

The songs I used were these (I downloaded and converted them to .mp3 with with yt-dlp and FFMPEG. Both are "All I ask" by Adele, one is a cover by Bruno Mars):
1) Low volume levels song: https://www.youtube.com/watch?v=oCtYekG1qBc
2) High volume levels song: https://www.youtube.com/watch?v=sonzlE32YVg


I've tried this with multiple versions of WACUP. I'm currently using the latest public release. My settings look like this:




-----------------------



-----------------------

[dro]

Can you attach the images to the post directly as I'm being geo blocked by Imgur so I just get a load of "content not viewable in your region" images.

[MarkRH]

Heh.. didn't know Imgur was Geoblocking the UK.  That kinda sucks.  Anyway, I decided to take a screenshot of all the images and attach it.

[dro]

It's probably so they don't have to deal with the age verification crap that was so poorly implemented as a requirement a few months back.

As for the problem being reported I've still to have a look into it but if it's against the x64 build then either in_wave isn't applying it or the fallback handling in the core which is there for the x86 build (mainly to deal with in_mp3 related issues) might not be working under x64.

[dro]

The adele link is also blocked for me so I've instead been testing with a file from Clapcast which is wanting ~-9dB adjustment & that appears to work as expected especially when toggling the overall replaygain (rg) option on/off. With the low volume yt url I could get that & the rg values are being read from it once generated but it's not working as expected due to how I'm trying to determine what the actual peak value to apply needs to be vs getting a value above zero.

Re-working that to my ears is now making an obvious difference along with comparing to an old fb2k install I've still got where it seems to sound roughly the same in either player whether the rg processing mode is enabled or not. In either instances I can now hear a change with the option enabled or not but I haven't been able to replicate where a below zero rg value isn't being applied per the initial post about this.

Another thing based on it not working for either file, I'd also double-check what's reported as the rg values if they're present in the local library db since the values from that are used over the local file if the local library is enabled & the files are known to it. Depending on how / when the rg values were calculated on the files it's possible that the local library hasn't had those values read into it & so would ignore what's then been added into the file as that's how the local library as a cache is being used. That'd require a metadata refresh for the rg values to then be read into the local library db cache so it can then be used as expected.