[MPlayer-users] divtc filter description

[Ville Saari]

On Mon, Feb 28, 2005 at 02:26:22AM +0300, Vladimir Mosgalin wrote:

> In man page, there are following words in description of divtc:
> 
> In order to use divtc  two  pass  with two pass video encoding, you must
> perform three passes: first divtc pass one and encoder pass one,  then
> divtc pass  two  and encoder pass two, and finally divtc pass two and
> encoder pass two.
> 
> I don't quite get it.. shouldn't it be other way around? First pass of
> divtc with second pass of encoder makes no sense to me.

Is that my cola or did someone else add that sentence to the manpage
afterwards?

Other way round is correct of course, but actually doing a redundant
encoder pass one for the first divtc pass is waste of cpu cycles.
I usually do it using this kind of dummy encoding:

mencoder -vf divtc=pass=1,crop=16:16:0:0 -ovc raw -nosound -o /dev/null

It runs very fast. Just remember to place softskip after divtc=pass2 in
the encoding passes. Using mplayer -nosound -vo null for the first divtc
pass may also work, but I haven't tried it.

It would be useful for this kind of tricks if mencoder had a null codec
that would just drop the data and produce empty frames. The frameno
should be pretty much like that, but it didn't work when I tried to
use it. Unfortunately I can't remember what kind of problem I had with it.

-- 
 Ville