Help - Search - Members - Calendar
Full Version: Versions
Hydrogenaudio Forums > Lossy Audio Compression > MP3 > MP3 - General
cookie
Hi,
I'm a bit confused.
Dibrom has 3.90.2, mitiok has 3.91, lame.sourceforge.net says 3.90 is latest, www.sourceforge.net/projects/lame says 3.91 is latest.
Errrrr, what ?
CiTay
LAME 3.91 stable is the latest & official version. Dibrom will soon compile a 3.91 "fast version".

3.90.2 was a semi-official, bugfixed 3.90 "fast version", but the bugfixes are all included in 3.91 now. The reason why 3.90.2 is still online is that Dibrom hasn't compiled a fast version of the official 3.91 yet.
Fallen Guru
Sorry for the newbie question but what are the differences between the official LAME release, the mitiok release and the Dibrom release? Is it just encoding speed or also features / presets / tuned details?

Up until now I've just been using the latest released official LAME sources and compiled them on one of the Linux boxes... any reason to switch?

Regards

Fallen Guru
john33
The most recent stable release is 3.92, and that is the recommended version. The 3.93 alphas have had a somewhat chequered history and are best avoided at the moment. As far as whose compile to use? Really makes damn all difference!wink.gif
Dibrom
QUOTE
Originally posted by Fallen Guru
Sorry for the newbie question but what are the differences between the official LAME release, the mitiok release and the Dibrom release? Is it just encoding speed or also features / presets / tuned details?

Up until now I've just been using the latest released official LAME sources and compiled them on one of the Linux boxes... any reason to switch?

Regards

Fallen Guru


The differences between my compiles and mitiok's is that my compiles use a few flags which result in a faster executable (and require a 686 equivalent or newer processor), though the encoded files average out to be 5kbps larger with --alt-presets. There should be no real quality issues, however there has been one occasion where I've noticed a very slight difference in quality on the infamous fatboy clip. In this case, my compile sounded a slight bit better, resulting in a little bit less pre-echo. Most probably this was because of the slightly higher bitrate. All of the --alt-preset testing in the past has been done with my versions of the compiles.

So in summary, there should be no real difference in quality. However, if you're extra paranoid, use my compile (3.90.2).

Btw, if anyone wants to make a newer compile in the same fashion I did, I've included in the makefile I used with the 3.90.2 binary along with instructions on how the compile was made. Compiler used was ICL 4.5.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.
Invision Power Board © 2001-2009 Invision Power Services, Inc.