For those that don't want to jump to the site:
LAME v1.00 engine 3.30 (don't know where this came from)
"Attenuated right channel on all encoded streams (linear distortion)
The most configurable encoder
Lame is the second encoder capable to produce VBR streams. At the actual stage
of it's development there are some serious problems with this software:
- the VBR setting doesn't really affect quality. If you set VBR0 you
get about the same quality as in VBR. This -of course- is also visible in the
average bitrate of the produced mp3 streams. The same applies to the quality
setting. The encoder is faster then BladeEnc but also the quality is slightly
poorer"
> This page reviews a number of mp3 rippers, encoders etc. The author(s)
> seem relatively unbiased. They have recently rated the LAME encoder.
> LAME scored well but not very well. Blade encoder ousted
> LAME on a few technical points (specifically frequency response, and
> a right channel bias.)
I simply don't grok why people think bladenc has a "magical" improved quality
over the ISO code. While johann has done some good work on optimizing for
speed, nothing much else has been done to the code.
Other points regarding USE!
- i'm sure lame is faster than bladeenc.
- didn't specify which options used to encode.
- I'm not sure on what the "detailed ratings" meant.
- freq throughput? is higher better?
- distortion - is lower better?
- supported platforms 3 points maximum. lame should score 3.
- wtf is space exigence?
who's measuring these? are they an average of a whole bunch of
different files? is it just the data for one file? Are the encoded and original
files just compared bit for bit by a program or a listener?
If you're going to set up some sort of database comparing encoders, the meaning
of different terms and how they were derived should be made clear.
later
mike
(who doesn't believe most benchmarks anyway)
--
MP3 ENCODER mailing list ( http://geek.rcc.se/mp3encoder/ )
