I've been using TimeMachine to record late night music radio programmes.
I'm pretty sure I didn't do anything different, but one recording is
significantly quieter than the others and the quiet one shows up as
of type 'Unknown (model/x.stl-binary)' in Nemo file manager, whereas
the good ones are of type 'Binary (application/octet-stream)'.
'file' says they are all Sony Wave64 RIFF data, WAVE 64 audio, stereo 48k
sox (-n stat) seems to show significant differences:
Good Quiet
Samples read: 676507136 666574336
Length (seconds): 7046.949333 6943.482667
Scaled by: 2147483647.0 2147483647.0
Maximum amplitude: 0.999969 0.999969
Minimum amplitude: -0.920116 -0.999969
Midline amplitude: 0.039927 0.000000
Mean norm: 0.154645 0.051673
Mean amplitude: -0.000065 -0.000022
RMS amplitude: 0.199364 0.066460
Maximum delta: 1.421344 1.754726
Minimum delta: 0.000000 0.000000
Mean delta: 0.090457 0.034210
RMS delta: 0.134921 0.048614
Rough frequency: 5170 5588
Volume adjustment: 1.000 1.000
Any idea what could have gone wrong?
What would be the right way to calculate the value to use with sox -v
to get the volume level somewhere near that of the 'Good' recording?
--
Thanks, John.