What happens, if you replaygain twice ?

If you have a mpc file and you don't know, if it is replaygained, eg. because of missing logfile.

Is it bad to replaygain --auto twice ?

Or does nothing worse happen ?

The replaygain values are stored in the header of the mpc file. The audio data itself is not modified. So, you can replaygain files as often as you want without decreasing quality.

yes,

but I have tried it at a test_mpc.

The second log_file looked slightly different than the first.
It is obvious, that the second replaygaining sees effects of the first replaygain.

my question is, if the "second_generation_replaygained_mpc" is played, does it have different volumes than the "first_generation_replaygained_mpc" ?

I just tried the following:

replaygain --auto 1.mpc
cp 1.mpc 2.mpc
replaygain --auto 2.mpc
diff 1.mpc 2.mpc

Diff showed that both files are identical. So, nothing to worry about.

the difference in log fiels has to do with rounding in the displayed values, the stored values are not changed.

i spoke with one of the developers about this very issue

yea, I checked it too.

the mpcs are bit_identical, comparing the 1.replaygained mpc with 2.replaygained mpc.

The logfile of 2. trial shows indirectly, that replaygain saw, that it was already replaygained. So there is a difference between logfiles.
That was irritating me and so I asked.

Yes, the replaygain values are identical, only the display/logfile is always 0.01 dB off. This has irritated a bunch of people already... it's an old bug that needs to be fixed. (Frank Klemm, are you reading this? )

The situation is even worse with mp3gain.
The analysis log is not identical at all and you don't get a change log after 2. run.

after first run - analysis:


after second run - analysis:


after first run - change:

always run "--smart --auto" <- no more problem

I was talking mp3gain....not mpcgain