i found soundkonverter in the program library and installed it
- the install ran to "94%" -- and did not complete -- but -- the program seems to run OK
-it does set replay gain tags but seems to be missing the 0db reference point
i use Audacious for play back
Audacious does respect the gain tags if you select that option in their preference dialog
too, Audacious allows you to select 16,24, 32, or 32 float as input to ALSA. i'm using AMD on-chip DAC (ATI/Radeon)
I have my .flac files as 24 bit / 48khz . the default on Audacity is to feed 16 bit data to ALSA.
when i switched this to 32float it cleaned up the sound, noticeably. i chose 32/float because some of my .wav files use that format. you would not want to convert 24 bit to 16 but in converting to 32/float yoy don't lose anything
there seems to be an ongoing argument over this in the digital music works, some folks even arguing for 192khz sampling.
Monte doesn't buy into this -- see http://www.xiph.org/video/
and goes into a discussion of "aliasing" -- which I didn't understand -- but -- acording to Monte -- too high a sample rate can actually be detrimental-- 41khz 'sufficient' he says, while the 48khz rate provides a little better 'headspace'
i've tried this a little -- and i still think 32/float .wav format sounds a little better than 24 bit .flac
Home assembled box using ASUS M5A88-M motherboard and AMD Phenom II X4 3.4GHz cpu