Since there are 4 different Madplay 0.15.0b compilations
I decided to make a little test. Performed on WinME\Duron 650
using 10 MP3 tracks (192 kbps).
John33 (Compiled using MSVC) http://rarewares.hydrogenaudio.org/files/MADbundle.zip (http://rarewares.hydrogenaudio.org/files/MADbundle.zip)
------
4:17
Speek (Compiled using MSVC) http://home.wanadoo.nl/~w.speek/download/madplay.exe (http://home.wanadoo.nl/~w.speek/download/madplay.exe)
-----
8:20
Thop (Compiled using mingw32) http://www.huno.net/cstm/madplay.zip (http://www.huno.net/cstm/madplay.zip)
----
3:22 (2.95.3)
3:04 (2.95.3 Accuracy)
2:40 (2.95.3 Speed)
3:19 (3.2.3)
2:42 (3.2.3 Accuracy)
2:30 (3.2.3 Speed)
William (Compiled using mingw32 3.2.3) See link below
-------
3:23
2:48 (Accuracy)
2:32 (Speed)
Note 1: Thop made his own test using AthlonXP 2224 and 3.2.3 compilations
have been slower on his machine, so you better make your own test
to determine which compilation better for your PC.
Note 2: It's strange that accuracy compilations show faster results than
default. Does anybody have some thoughts about it ?
Question from me: should I use -d (No dither) option when decoding
44100 16bit Stereo MP3s and why ? Thanx
John33 (Compiled using ICL) .......
It's not ICL, it's also MSVC6.
I have made a simple web page for my MAD compiles:
http://www.williamng.ws/mad/ (http://www.williamng.ws/mad/)