|
From: Jan S. <ha...@st...> - 2011-09-12 09:55:05
|
> And I can not understand how the voice is changed. I use sox C:\sox-14-3-2\musteri.au --show-progress -n stat. And the output is: > > > ============================================================== > Input File : 'C:\sox-14-3-2\musteri.au' > Channels : 1 > Sample Rate : 8000 > Precision : 14-bit > Duration : 00:01:41.40 = 811200 samples ~ 7605 CDDA sectors > File Size : 811k > Bit Rate : 64.0k > Sample Encoding: 8-bit u-law > > In:100% 00:01:41.40 [00:00:00.00] Out:811k [ -====|====- ] Clip:0 > Samples read: 811200 > Length (seconds): 101.400000 > Scaled by: 2147483647.0 > Maximum amplitude: 0.378784 > Minimum amplitude: -0.378784 > Midline amplitude: 0.000000 > Mean norm: 0.009511 > Mean amplitude: 0.000011 > RMS amplitude: 0.049504 > Maximum delta: 0.196045 > Minimum delta: 0.000000 > Mean delta: 0.003645 > RMS delta: 0.017613 > Rough frequency: 453 > Volume adjustment: 2.640 > Done. > > ================================================================ > > > How I understand that normalize our file? On a scale from -1 to +1, the minimal/maximal amplitude present in your file is about +/- 0.378784; so the volume can be adjusted by a factor of 2.640 without introducing clipping. That's what normalize means. You need to read the man page, really. It explains everything in detail, and teaches you a thing or two about digital audio in general. Jan |