[MPlayer-users] Transcoding from one aspect ratio to another
Nix
niks1024 at gmail.com
Thu Nov 4 22:33:16 CET 2010
On 2010.11.04. 03:24, Rolf Ernst wrote:
> On 11/3/2010 7:56 PM, JD wrote:
>> On 11/03/2010 05:44 PM, Rolf Ernst wrote:
>> Hey guys - let us not belabor this enquiry.
>> I accept that what I want cannot be done without one of
>> 1. Black bars on top and bottom
>> 2. Zoomed in until each frame fills 800 vertical pixels of the screen,
>> and then cropped so that only 1280 horizontal pixesl of the zoomed
>> frames remain.
>> Neither of these is what I want, so let's not continue to beat a dead
>> horse :)
> This is simply not so but by all means, I shall not force you to
> accept the facts.
Since my weekend started this evening you are welcome to prove it. No
seriously, take a square (for sake of example you can assume that it has
an image mapped on it) that's being displayed with ratio of 2:1 (in
other words the image mapped on it is in fact stored in an vertically
stretched state for improved vertical resolution). Now, if you crop
away, say, 20% from right side, do you really think that it should be
still be displayed with DAR of 2:1? Obviously, if you retained PAR, then
all is fine but DAR would still be different. I don't know about
mencoder but generally for muxing you need to specify either DAR or the
intended viewing resolution which more or less obviously same thing.
P.S. I just found this cute little line in mkvmerge manpage:
"Matroska™ files contain two values that set the display properties that
a player should scale the image on playback to: display width and
display height."
P.P.S. Regarding your other post about anamorphic being non-square
pixels, I'm not so sure about your definition. Mostly because anamorphic
video was invented like before 1930's when digital computers didn't even
exist, therefore at the very least your definition that ties anamorphic
with PAR is at the very least badly defined, in my humble and worthless
opinion obviously.
Cheers,
nix
More information about the MPlayer-users
mailing list