Submit Hint Search The Forums LinksStatsPollsHeadlinesRSS
14,000 hints and counting!

Click here to return to the 'Ripping speed' hint
The following comments are owned by whoever posted them. This site is not responsible for what they say.
Ripping speed
Authored by: Thistledown on Dec 14, '01 11:16:02AM

Ok. I ran a test with two tracks this morning. I used ExactAudioCopy on a pc to extract 2 wav files from a CD. I transfered the wav files to my mac and here are the results.

800Mhz Pentium III PC running Windows 2000
Track 1
2.75x speed
1:31 time

Track 2
2.3x speed
2:17 time

Track 1
3.8x speed
1:01 time

Track 2
3.8x speed
1:22 time

550 Mhz TiBook running OS 10.1.1
LAME 3.89 (beta 1, 12/03/01) installed via fink
Track 1
.85x speed
4:34 time

Track 2
.80x speed
6:25 time

Clearly the Windows version is much faster which just irks me to no end. There are some obvious optimazations in the 3.90 MMX version. I'm surprised that a December build is as slow as it is on my mac. If anyone discovers a new build or optimization for LAME on OS X, let me know.

[ Reply to This | # ]
Ripping speed
Authored by: marmoset on Dec 18, '01 04:00:51PM

IIRC, the x86 versions of LAME use inline hand-coded assembly
in some performance-critical sections, LAME for other CPU
architectures falls back on routines written in C.
Unless someone goes through and does an equivalent
amount of optimization for the PPC version, using LAME
to compare the platforms' speed is not going to be a fair

[ Reply to This | # ]