Ryan Bair a écrit :
> Everything I've tried has some really nasty artifacts, the CBR 64
> samples sound far superior while usually being roughly the same size
> as VBR 8. I've tried playing around with the sample rate, raising up
> the bitrate floor as high as 64kb/s, and changing the quality to 0, 1,
> and 2 but the quality remains poor.
>
> Is there anything else I should try? Obviously I'm not shooting for
> something perfect, but I think it is reasonable to get quality at
> least equal to CBR 64 out of VBR 8.
First, you should try using the command-line executable, in order to
remove potential problem that could come from a mis-use of the lib.
(you've already done this).
Then, you might check using both 3.97 and 3.98, in order to isolate
potential regressions. Af at this point you have some samples on which
you find that -V8 is noticeably lower quality than -b 64, then a very
good thing would be to provide us a short test sample. A 10-20s cut of
the original uncompressed wav file, uploaded somewhere on the web, and
explicitely mentionning us the settings you used would be ideal.
Regards,
--
Gabriel Bouvigne
www.mp3-tech.org
personal page:
http://gabriel.mp3-tech.org-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference
Don't miss this year's exciting event. There's still time to save $100.
Use priority code J8TL2D2.
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone_______________________________________________
Lame-dev mailing list
[hidden email]
https://lists.sourceforge.net/lists/listinfo/lame-dev
