Daniel Kristjansson wrote:
On Tue, 2005-10-18 at 12:48 -0600, Chris Dos wrote:
Daniel Kristjansson wrote:
Hmm, what are your exact compile options? I have a P4 @2.8 Ghz and it
has no problem with software decoding. I'm using:
"-march=pentium4 -O3 -fomit-frame-pointer"
My compile options are: --cpu=athlon-xp --tune=athlon-xp --enable-mmx
Try mine, "-O3 -fomit-frame-pointer" makes about a 10%-20% difference.
Also, use march not the old cpu & tune options, and also you should
know that on many versions of gcc athlon-xp is slower than pentium4
on Athlons.
On on svn version 7524 now.
I'm using gcc 4.0. I compiled twice using the following options:
--extra-cflags="-O3 -march=k8 -fomit-frame-pointer "
--extra-cxxflags="-march=k8 -O3 -fomit-frame-pointer "
and
--extra-cflags="-O3 -march=pentium4 -fomit-frame-pointer " \
--extra-cxxflags="-march=pentium4 -O3 -fomit-frame-pointer " \
Using libmpeg I was dropping about 3 frames per second using no deinterlacing
and about 7 frams using deinterlacing. So a 2.6 to 2.8 ghz is probably the
sweet spot to do software mpeg2 decoding.
Well, based what you just said, I would say the A/V diverged code indeed
broken. :)
I improved it a little this morning..
After enabling XvMC, I'm still receiving dropped frames: The log messages are
totally differn't know:
NVP: Audio is 3.58185 frames ahead of video,
doubling video frame interval.
Frame H w/children: A F is being held for later discarding.
Frame H w/children: A F is being held for later discarding.
Frame H w/children: A F is being held for later discarding.
Frame H w/children: A F is being held for later discarding.
NVP: Audio is 4.69679 frames ahead of video,
doubling video frame interval.
Frame B w/children: D F is being held for later discarding.
Frame B w/children: D F is being held for later discarding.
Frame B w/children: D F is being held for later discarding.
Frame B w/children: D F is being held for later discarding.
AFD: positionMap[ 216 ] == 250758160.
It doesn't say it's diverging anymore. But the results seem to be the same.
Increasing the buffers to 8192 and 25112 also didn't make any difference.
Is there anything else I can try for you?
Chris
_______________________________________________
mythtv-dev mailing list
[email protected]
http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev
