[MEncoder-users] avi sync questions

[Wes Morgan]

I've got a D-Link DSM-520 that I've been using for several years to play 
videos that are both downloaded and personally ripped. Periodic firmware 
updates are released, and there is a specific difference between version 
1.02 and higher releases... Any video I ripped with mencoder loses audio 
sync when played. I submitted a sample video to D-Link and their response 
was:

     We have checked your file. It seems that the timestamp of the MPEG4
     video is different from that of the AVI header and the AVI timestamp
     is the correct one.

     This make us to be in the delimma. Some others files are in the other
     side. This is very difficult for a streaming player. We can not jump
     randomly like the local player. We may add a sync key to reset the
     timestamp. However, I believe that this will be a strange thing for
     the end-users.

The person who sent me this response is very active in a forum thread 
about the player and is obviously someone who does or has worked on the 
firmware development team.

Does the claim make any sense? Is there any way I can prove or disprove 
that these two timestamps disagree? If so, is there any way to fix them?

The commandline for all of the encodes was basically:

mencoder -oac copy -ovc xvid -ffourcc XVID -ofps 24000/1001 \
     -passlogfile BAND_BROTHERS_D1.log -vf crop=712:480:4:0,harddup \
     -aid 128 -msglevel all=-1:statusline=5 -xvidencopts \
     autoaspect:chroma_opt:keyframe_boost=15:max_bframes=0:quant_type=mpeg\
     :quant_intra_matrix=/usr/home/morganw/.transcode/intra.matrix:\
     quant_intra_matrix=/usr/home/morganw/.transcode/inter.matrix:\
     turbo:threads=1:bitrate=1500:pass=1 -channels 6 \
     -o /dev/null -dvd-device BAND_BROTHERS_D1 dvd://1

With obvious differences for crop, aid, log file etc. I've tried various 
things like extracting the streams and using ffmpeg to remux them into an 
avi, using avidemux2, I think I even tried virtualdub in windows. Nothing 
seemed to fix the issue.

So, does anyone have an idea how to determine where the real problem is, 
with actual evidence?


Regards,


Wes