[MPlayer-users] Slave command "panscan" doesn't work with frontends.

Reimar Döffinger Reimar.Doeffinger at gmx.de
Thu Aug 27 14:07:10 CEST 2009


On Thu, Aug 27, 2009 at 01:47:56PM +0200, Reimar Döffinger wrote:
> On Thu, Aug 27, 2009 at 07:27:09AM -0400, Eric Appleman wrote:
> > On 08/27/2009 04:44 AM, Reimar Döffinger wrote:
> > > On Thu, Aug 27, 2009 at 04:30:35AM -0400, Eric Appleman wrote:
> > >> SMPlayer: http://img269.imageshack.us/img269/5809/mazuiharuhi21overscan.png
> > >>
> > >> MPlayer: http://img268.imageshack.us/img268/6272/mplayeroverscan.png
> > >>
> > >> Why is this happening?
> > > 
> > > Frontends use -wid. MPlayer fully fills that given window, leaving full
> > > control of position and size to the frontend. Thus panscan handling is
> > > out of MPlayer's hands.
> >
> > I see. But according rvm, SMPlayer is forced to emulate panscan and that
> > the correct behavior I seek cannot easily be replicated.
> > 
> > In this case, the proper behavior is that of MPlayer, where the
> > subtitles are not cut off.
> > 
> > The issue at hand is not the panscan itself, but how the subtitles are
> > treated during a panscan.
> 
> Hm. Well, I intended to work on "fullscreen" with wid, where the meaning
> of fullscreen would be switching between aspect/panscan handling by
> application or by MPlayer.
> It should not be that hard, I just never got around to implement it, and
> it is also always a bit of a pain to test the -wid stuff.

MPlayer SVN with -vo gl and -vo xv now supports -wid with -fs where
MPlayer handles aspect and panscan on its own.
This probably will need some additional support from SMPlayer though.
I'd be happy to know if this feature is helpful and works well for
frontend developers.


More information about the MPlayer-users mailing list