Shawn Riley schrieb:
> >Also, Fraunhofer use intensity stereo at low bitrates. Maybe you could
> >compare LAME with FhG -no-is.
> � No can do. The Fraunhofer ACM codec only
says if the output is in Stereo or Mono. Winamp displays these "Stereo" files
created by FhG ACM as being "Joint Stereo", but I'm not sure if it really is JS
or if Winamp is incorrectly displaying IS as JS.
> � What is IS? Is it like JS
at low frequencies with panned mono highs, or is it every frequency band in
panned mono?
IS means intensity stereo, some higher frequencies encoded in mono with some
panning information
it is one of the joint stereo tools you can enable turning on joint stereo
the other one is mid/side stereo (the only one implemented in LAME)
> � I think it's quite disastrous that the standard Fraunhofer
encoder doesn't let you specify exactly how you want stereo files to be
treated. The encoder knows nothing about the system that the files will be
played back on, & can't really be sure if that mode will be best. Thanks heaps
to whoever added the -m option to Lame with all its parameters. More options =
better control (usually).
I'm not quite sure actually, but I think that mp3enc3.1 lets you control
more than that ACM
>
> � I just did some tests with v3.83. Mostly the
low sample rate stuff. 24/56/JS from v3.83 is better than both 24/64/JS
22/64/JS from v3.62! Well done, everyone! :-) The resampling works a charm too!
Just a couple of things though.
> 1- Compressed a 4'30" song to MP3 using the
command line lame.exe -v -V8 -B40 -mm -c -h -p --lowpass 5.3 --resample 12
--noshort in.wav out.mp3
> � a- The average bitrate was reported as 19 kBit/sec
by the frame analyser at the end of encoding (which was correct), but when I
loaded it in Winamp, it thought the file was 38 kBit/sec. It also reported the
song as being 2'15" (half of what it actually was). Winamp bug or Lame bug?
Winamp didn't stop when it got to 2'15" though. And it DID actually play the
whole song, so that's a good sign...
it looks like Winamp doesn't expect a MPEG2.5 VBR file?
> � b- Using the same options, except with
resampling to 11025 Hz, the file averaged 27 kBit/sec. That puzzles me too.
Shouldn't it be slightly smaller in size than the 12 kHz sample?
no, because of how the freq bins get filled and the masking works
> 2- Winamp
refused to play back an 11025 Hz - 96 kBit/sec - Stereo MP3. The
--strictly-enforce-ISO option didn't work. A(nother) bug in Winamp, perhaps?
this is because FhG always tolds that MPEG2.5 files can only go up
to 64 kbits
> 3- When using CDex to convert a 12 kHz - 64 kBit/sec - Stereo MP3 to a Layer-3
WAV, the Fraunhofer ACM wouldn't load it. The WAV header was correct. Then I
checked the compression options for the Fraunhofer ACM. 12 kHz was ABSENT! Does
FhG have something against MP3 with 12kHz?
>
> Shawn
Robert
--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )
