On Tue, 2005-02-15 at 14:19 -0500, Isaac Richards wrote: > On Tuesday 15 February 2005 02:05 pm, Brian J. Murrell wrote: > > See, that is totally wrong (or rather not as right as right could be) > > methodology. The real secret to picture perfect tv-out is to display to > > the TV _exactly_ as the original signal would have. Record the signal > > exactly as it comes and display it back exactly as you recorded it > > Don't post/pre-process it because of a crippled TV-Out method. > > > > So you record the signal field by field and then redisplay it clocked > > perfectly, field by field. That is the only way to do it. > > > > This is what DirectFB and the G400's TV-Out accomplishes. > > Unfortunately, Myth's use of QT makes this impossible because QT does > > not support DirectFB, or even SDL. SDL would be satisfactory because > > SDL supports drawing on a DirectFB surface. > > Guess I'll have to remove all those SDL visualization methods in mythmusic, > and the DirectFB videoOutput class in mythtv, since they're obviously > impossible.
Go back and reread what I said. I did not say Myth's use of DirectFB was impossible (indeed it is great) nor did I say Myth's use of SDL was impossible. What I said is that QT did not support writing on either an SDL surface nor a DirectFB surface, which is the disappointing part because that makes getting the OSD on DirectFB impossible. I'm really not sure how displaying the OSD on an X-server on a G400 "matroxset" mangled framebuffer is supposed to work with the DirectFB "layer 2" CRTC2 output anyhow since I thought the two were mutually exclusive (i.e. framebuffer/X11 needing the maven kernel module and DirectFB requiring no maven kernel module). I did try it and all I got was the video playing in a small square in the top right of the screen. b.
signature.asc
Description: This is a digitally signed message part
_______________________________________________ mythtv-users mailing list [email protected] http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-users
