I noticed some small loudness differences of replaygain normalized tracks when they're played through RadioBOSS or foobar2000. It seems to me that RadioBOSS somewhat adds approx. 1 dB to replaygain controlled tracks (well, it actually looks more complicated).
Honestly, I spent a bunch of time to calibrate our mixing console with audio interface to perfectly hit a 0 LU on Orban Loudness Meter (or -23 LUFS). Unfortunately, when tracks (containing RG info) are being played through RadioBOSS, it's not uncommon to see values 0.7-1.2 LU.
See, I tried to explain it to myself, that this is may be a hardware-related issue or that simply I did something wrong. Now, I have a proof that this is caused by RadioBOSS - see attached screenshots. Sure I can fine tune 'preamp for tracks with replaygain' from RadioBOSS's settings, but comparing peak values from RadioBOSS and f2k makes me feel like it's a bug.
FYI: measured track was Trick Me by Kelis. Its EBU R 128 compliant replaygain metadata (track gain) says that peak values should hit -13.49 dBFS. In RadioBOSS track was played 2 times (firstly through WASAPI, and secondly through DirectStreaming - both ended up with almost identical results). F2k utilized DirectStreaming. All values were measured with Orban Loudness Meter.
I also did a verification test: after removing RG info from track I attenuated its level by 13.5 dB in wave editor. It was played correctly in both applications.
Dmitry, may I ask you to inspect replaygain implementation?
Honestly, I spent a bunch of time to calibrate our mixing console with audio interface to perfectly hit a 0 LU on Orban Loudness Meter (or -23 LUFS). Unfortunately, when tracks (containing RG info) are being played through RadioBOSS, it's not uncommon to see values 0.7-1.2 LU.
See, I tried to explain it to myself, that this is may be a hardware-related issue or that simply I did something wrong. Now, I have a proof that this is caused by RadioBOSS - see attached screenshots. Sure I can fine tune 'preamp for tracks with replaygain' from RadioBOSS's settings, but comparing peak values from RadioBOSS and f2k makes me feel like it's a bug.
FYI: measured track was Trick Me by Kelis. Its EBU R 128 compliant replaygain metadata (track gain) says that peak values should hit -13.49 dBFS. In RadioBOSS track was played 2 times (firstly through WASAPI, and secondly through DirectStreaming - both ended up with almost identical results). F2k utilized DirectStreaming. All values were measured with Orban Loudness Meter.
I also did a verification test: after removing RG info from track I attenuated its level by 13.5 dB in wave editor. It was played correctly in both applications.
Dmitry, may I ask you to inspect replaygain implementation?
