I did some informal timings on my PII 266 laptop. Dont take them
too seriously, but it looks like LAME is faster than BladeEnc:
Encoding a 4:39 song:
Playing_time/Encoding_time
BladeEnc v0.72 .52
LAME 3.03 with GPSYCHO .68
LAME 3.03 with ISO model (identical results to BladeEnc) .78
On an SGI, I've actually been encoding faster than real time, around 1.5x.
I think this is a great demonstration of the power of open source! (I
have made no speed improvements at all. This work is due to Mike
Cheng, the 8hz-mp3 group, Chris Matrakidis, James Droppo and probably
many others.)
Mark
--
MP3 ENCODER mailing list
