On Wed, Nov 06, 2002 at 10:56:04PM +0200, Ville Syrj�l� wrote:
> 
> I'm sorry but I don't see it. All I see is some pincushion distortion

That's it!  "Pincushion distortion".  That is the term I could not
think of.

> which is to be expected with TVs. 

Why do you say this?  Mplayer does not cause the same distortion.
Notice with Mplayer, the left edge of the display is very straight and
sharp -- right out to (and beyond even -- as you say, "overscanned")
the left edge of the display tube.

> That's to be expected also.

OK.

> I think that with mplayer you just don't see it due to the content being
> designed for overscan displays.

OK, but why is the GTK output not being overscanned the same way (like
it is along the top)?  I guess this is my issue.  If the output was
simply being overscanned beyond the viewable area of the display, I
would reduce the size of the toplevel GTK widget (which I have at
720x486 right now) and center it so that it displayed inside the
viewable area, but minus the pincushion effect.

This pincushion effect is identical to when I was using fbset and
matroxset to enable TV-Output of the framebuffer.  However when I was
doing that, both GTK and Mplayer's output were pincushioned exactly
like the GTK (and only GTK) output is with the matrox-crtc2 mode now.

I guess I don't understand why using the matrox-crtc2 mode of directfb
now corrects the Mplayer output but not the GTK output.

> When using a TV one really shouldn't draw anything vital near the edges.
> European teletext pages only use something like 80% of the total screen
> width.

That is fair enough, and I am willing to follow a similar restriction
if I can get the GTK output to overscan like the Mplayer output is
currently.

b.

-- 
Brian J. Murrell

Attachment: msg01026/pgp00000.pgp
Description: PGP signature

Reply via email to