[MEncoder-users] "Too many video packets in the buffer"

Andrew Berg bahamutzero8825 at gmail.com
Sat Apr 24 15:35:15 CEST 2010


On 4/20/2010 8:23 AM, belcampo wrote:
> -fps 50 -ofps 25
> and
> filmdint=io=50:25
> which doesn't make sense to me at least.
I thought filmdint would be more reliable than framestep=2, but I was
wrong. Using framestep=2 and -demuxer lavf (which I was sure I had tried
before in vain), I was able to make it work, but oddly enough, only for
my ultrafast test preset and not the much slower set of options I
normally use.

In sync:
mencoder 2010.02.02.m2ts -msglevel all=1:statusline=5 -demuxer lavf -aid
1 -fps 50 -ofps 25 -endpos 180 -of avi -oac copy -ovc x264 -x264encopts
level_idc=41:crf=30:force_cfr:turbo=1:frameref=1:scenecut=0:nf:nocabac:bframes=0:partitions=none:no8x8dct:me=dia:subq=0:aq_mode=0
-vf yadif=1,framestep=2,hqdn3d=0:0:8:8,harddup -o test_fs2.avi

Not in sync:
mencoder 2010.02.02.m2ts -msglevel all=1:statusline=5 -demuxer lavf -aid
1 -fps 50 -ofps 25 -endpos 180 -of avi -oac copy -ovc x264 -x264encopts
level_idc=41:crf=19:force_cfr:keyint=300:keyint_min=30:frameref=4:mixed_refs:weightp=2:b_pyramid=normal:weight_b:bframes=4:b_adapt=2:me_range=24:direct_pred=auto:cabac:8x8dct:me=umh:trellis=1:subq=7:threads=auto:nointerlaced:nopsnr:nossim:deblock=-1,-1
-vf yadif=1,framestep=2,hqdn3d=0:0:8:8,harddup -o full_test_fs2.avi


I compared both tests to the original and the former is perfectly in
sync while the other is 50ms or so off (it's not a lot, but it is
noticeable). I can make links to the 3 minute clips (350MB and 535MB
respectively) if someone wants to take a look.


More information about the MEncoder-users mailing list