I have an MP3 file with the following characteristics:
Media Type: MPEG 2.0 Layer III
Audio: 64 KB/s, 22KHz (joint stereo)
Emphasis: none
CRC: No
Copyright: No
Original: Yes
Padding: Yes
Length: 6:07
If I play the file via my sound card using a command
like 'mpg321 file.mp3' it sounds normal.
If I process the file like so:
mpg321 -r 44100 -s --stereo -q file.mp3 > file.raw
sox -r 44100 -s -w -c 2 file.raw file.wav
The resulting file appears to be encoded at only half
the bitrate and therefore sounds like it is speeded up
by 2X.
I'm using the mpg321 code from CVS as of today and
mad-0.14.2b.
Sample MP3 file
Logged In: YES
user_id=330927
Hm, this shouldn't be a problem - I think. I'll have to
investigate, but the WAV file should have all relevant data
in the headers, so that different bitrates, etc won't
matter. Anyhow, post-0.3.0, since I'm concentrating on
network issues for 0.3.0.
Logged In: YES
user_id=330927
This is a fundamental problem with mpg321. It hasn't got
resampling capabilities, so it ignores the -r 44100 argument.
I am planning on enhancing libao, the audio output library
(which handles all output, even to stdout) to do resampling;
however, I've not put in the time to do it yet, and might
not until after mpg321 version 0.3.0 is released.
Logged In: YES
user_id=272596
This would be good to have so that mpg321 can be used
instead of mpg123 for asterisk (http://www.asterisk.org) ---
it has to downsample to 8Khz to play over phone interface.
The options passed are "-q -s --mono -r 8000 -b 2048" and
when you listen all you hear is strange gurgling noises.
Assume it's the resample problem... :)
Logged In: NO
This is a critical bug because instead giving an error or
resampling correctly cdr files created with --cdr are no
resampled and sound funny in the cdaudio player.