[MPlayer-users] please help test my mplayer widescreen converter wrapper script

[Daniel Manjarres]

So here is a version of the script which support an "orthodox" setting which
uses -fs instead of -x -y and -vf expand. It screws up badly when you start
at a different resolution than your desktop, in which case the video is the
size of your old resolution, but the black bars extend to your full desktop
size, and it switches to your full desktop resolution (but doesn't switch
back to the original  resolution), making the video the opposite of
fullscreen, in my case I get a 960x600 video in a 2560x1600 black box. But
hey if it uses x% less CPU power then some people will like it better
despite that bug (it also always produces black bars instead of gray ones
for quicktime files; gain a bug here, lose a bug there).

I also discovered that when I make the gamma paramater 1.12 during the
encoding of the diff frames, that one file that got a crop region of 406
will play. Here's the deal: it gets detected as ysize of 406 and origin of 0
(instead of 404 and 2), and mplayer doesn't care that the ysize runs off the
end of the 405 pixel high video since the region starts at zero, so the
cropdetect divisor is back to 2, which gives better results than the sneaky
16 it was using when I told it to use a divisor of 1 before. But it is weird
that it lets crops run off the edge one way, but not the other. Ah well,
mplayer is a complex piece of software, it's going to have bugs.

google, do you see this discussion about automatically correcting widescreen
aspect ratio and croping video using mplayer on linux?

Later
-------------- next part --------------
A non-text attachment was scrubbed...
Name: fsmplayer
Type: application/octet-stream
Size: 16993 bytes
Desc: not available
URL: <http://lists.mplayerhq.hu/pipermail/mplayer-users/attachments/20060702/9ff43d42/attachment.obj>