Thus spake Robert Hegemann ([EMAIL PROTECTED]):
> By the way:
> LAME gets press coverage in Germany's respected c't computer magazine,
> which will appear on newsstands tomorrow, monday.
> I've heard, they've tested LAME against Xing, Fraunhofer and some other
> encoders at constant and variable bitrates, and LAME is always by the best.
> But I will read it tomorrow :)
I wrote that article ;)
The article was written because the earlier articles in c't about mp3
and audio compression left a lot to be desired -- they encoded sine
waves and white noise and came to the conclusion that all encoders were
of the same quality. I wrote a big rant letter to them and they told me
that if I could write a better article they would print it.
So I wrote an article about how audio compression works. They decided
to make a whole MP3 issue of c't and asked me to do an encoder test,
too.
I took a few "hard to encode" samples and had the contenders encode them
at 96 kbps. The most prominent sample was from a live CD of Herbert
Gr�nemeyer, basically lots of applause.
The conclusions where:
constant bitrate:
Fraunhofer, lame, Xing, (long pause) bladeenc
VBR:
lame+Xing, (pause) Fraunhofer
speech constant bitrate:
Fraunhofer, lame, (long pause) Xing, bladeenc
Lame is always close second place or even tied first place.
But even more important than the encoder test is another article (in the
same issue) they did after I suggested to do some double blind tests of
MP3 against audio CDs with musicians and so-called audiophiles and none
of them was able to detect the MP3 sound even at 128 kbps. That article
really is hilarious and I highly recommend reading it.
They gave me permission to post the unedited audio compression basics
article (in German) on http://www.fefe.de/, where I will put it later
today. I will ask them for permission to post the encoder test as well.
Felix
PS: This is my first email to this list, so I should also mention that I
submitted the RTP patch to lame (in case anyone wants to ask me
anything about it).
--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )
