[RndTbl] Video Overlay cropped

John Lange john.lange at open-it.ca
Thu Oct 12 13:10:53 CDT 2006

On Thu, 2006-10-12 at 12:47 -0500, Trevor Cordes wrote:
> On 11 Oct, John Lange wrote:
> > Turns out it is a bug with the ATI drivers using Xvideo. Don't know if
> Oops... I had missed this earlier email of yours.  You had answered some
> of my questions already I see.
> > A "sort-of" work around was to use a different playback such as:
> > 
> > # mplayer -ov gl2 <filename>
> > 
> > This works great! Unfortunately that does not solve the mythtv problem
> > because myth appears to use its own internal playback mechanism which
> > only works with Xv. As far as I can tell there is no possible way to
> > adjust this...
> > 
> > BTW, to answer your question, downloaded video had the same problem so
> > its a playback problem, not a capture problem.
> This entire quote doesn't make sense.  You're saying that playing
> something with mplayer -ov gl2 works ok but playing a downloaded video
> (not a NUV file) does not?  What are you using to play a downloaded
> video?  Myth just uses mplayer internally I believe for non-NUV file
> playback.  Therefore a downloaded file, if it plays properly with
> mplayer outside of myth should play ok inside of myth too!

Back when I was using the binary ATI drivers; playback of both liveTV
(from the tuner) and from a file was borked using Xvideo playback no
matter what tool was used (Myth, mplayer, Xine).

However, using mplayer with the gl2 playback looked fantastic on both
liveTV and on files. Myth has no option for gl2 playback of LiveTV. For
playback of files Myth uses mplayer (as you stated). So yes, playback of
files works fine in and out of Myth as long as you use the -xv gl2
option with mplayer (or Xine if you configure it that way).

> I'm still curious to see if your NUV files are ok or not.  You know, you
> can play NUV files on XP now using some .NUV reader available somewhere
> (sourceforge?).  Works well but causes equal-width black bars to be to
> the left and right when I playback my Plextor-created DIVX/MPG4 NUV
> files.  But you could use that to test.

That would assume I have a windows machine which I do not.

> I'm probably barking up the wrong tree with the recording-flaw theory,
> but in situations like this you must eliminate things in every step of
> the process.

Agreed. I feel I have eliminated capture as the problem but will perform
one final test this evening to confirm.


More information about the Roundtable mailing list