Hi,
I need to convert a larger amount of audio material pretty soon, for
archiving purposes (i.e. the material will be gone after that, so I
won't have a second try). The source material is in CD format (16bit,
44.1 kHz).
All things considered, I'm currently inclined to simply use
lame -k -h -b 320
because
- the VBR implementations in lame seem to be in a constant state of
flux, with no convergence or general consensus in sight what kind
of tweaking is good or bad (please tell me if I'm wrong here)
- I clearly hear differences between CBR 128 and CBR 192 on my
current equipment, so to be safe something equivalent to CBR >200
in audio quality would be needed, at least
- I don't want to hear differences on any future equipment I might
have, so my current equipment has only reduced use in determining
subjective sound quality
- my soundcard forces resampling to 48 kHz, so I can't monitor the
exact result of the decoded mp3 anyway
- storage space doesn't matter much at this point in time
(i.e. 20-25% compression is already quite acceptable)
- I can still re-encode it later on if it should become necessary
- CBR encoding currently is significantly faster than VBR
- I'm rather paranoid :)
At the moment I'm testing with the notlame incarnation of lame 3.70
under Linux.
Are there any significant changes in the CBR implementation or new
options in later versions that I should be aware of?
Which of the currently existing versions would you recommend, or should
I be fine with 3.70?
Are there any plans when the next "stable" version should / will be
released (I guess this would be 3.90)?
Does anyone disagree that the above command line is the best one can get
out of the mp3 format currently (without going beyond 320 kbs, that is)?
Thanks,
-- Niklas
--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )
