Computed replaygain with FLAC, and with WINAMP and both gives almost same results (usually with less than 0.02db difference). But Foobar results are more than 10% off the ¿correct? value. Is there a way to calculate value as FLAC does?
foobar2000 has been using libebur128 since version 1.1.6 (~2011). I think FLAC and Winamp use the original ReplayGain (by 2Bdecided).
You can try replacing the newer foo_rgscan.dll with an older version (e.g. foobar2000 1.1.5), but this might cause issues... not sure.
Or just use a portable 1.1.5 "installation" for your original ReplayGain scanning.
You can try replacing the newer foo_rgscan.dll with an older version (e.g. foobar2000 1.1.5), ...
That did the trick. Any idea why they changed calculation method?
It's more accurate according to the 1.1.6 changelog:
ReplayGain scanner now uses libebur128 for improved accuracy.
If you search this forum (use Google) you can find some discussion on whether EBU R128 was indeed an improvement over original ReplayGain. YMMV depending on material.