MPEG-FAQ 4.0: So, the Xing decoder is cheating, right ?
MPEG-FAQ 4.0:
So, the Xing decoder is cheating, right ?
[ About what Xing is messing up again ... ]
Date: Mon, 3 Jan 1994 12:20:33 -0800 (PST)
From: Jared V Boone <jboone@patriot.wtfd.orst.edu>
Unfortunately, my program DOES NOT decode in real time. But then, Xing's
program cheats. It does not decode the entire file, but plays the lower
half of the subbands and only one channel of a stereo pair. My program
will decode the whole thing, but there's a price to be paid. Decoding
'together.mp2' takes approximately 797 seconds on a Intel 486DX2-66
Windows NT/Visual C++ PC, and 1152 seconds on a Intel 486DX2-66 NetBSD/GCC
V2.4 UNIX system. So I guess that's about 3-5 times slower than necessary
for real-time playback. I've got some tricks I want to try, but they'll
involve a lot of code modification. I also don't think they'll make THAT
much difference. We may be asking these processors to do more than they can.
I'll keep you posted...
Jared Boone (jboone@patriot.wtfd.orst.edu)