[MEncoder-users] pullup useless?

[Laine Lee]

On 11/28/08 8:12 AM, "Daniel Gnoutcheff" <gnoutchd at union.edu> wrote:

> On Fri, 2008-11-28 at 14:16 +0300, Vladimir Mosgalin wrote:
>> Well, no one said that pullup is universal solution. If detc or ivtc
>> work for you, use them (though you should use ivtc=1 with mencoder, and
>> corresponding detc options). These filter have their advantages and
>> there are cases when they do work better, due to different strategy.
> 
> If that's the way things are, then that's cool. It's just that the docs
> gave me a different impression; when I read the man page entires for
> detc and ivtc, I came away with the message "don't use these, these
> suck, they're not being maintained, they're depreciated, use pullup or
> filmdint instead". So I thought the fact that detc & ivtc worked and
> pullup didn't was something to be really concerned about.
> 
> BTW thanks much for the tip about passing parameters to detc and ivtc. I
> guess I need to stop reading between the lines of the documentation and
> start reading the damn lines. :D
> 
> Thanks very much all,

I know a month has gone by, but I just thought I would suggest that you try
the only filter chain that has ever successfully accomplished pullup for me.
I'll grant you that I've not tried it with a variety of sources, but it
might be worth a try for yours.

-vf framestep=2,filmdint=fast=0,softskip,harddup

If you get good results for much of the output, but find certain passages in
which 1 of every 4 frames is a duplicate of the previous one (a way to
figure that out is to pause mplayer, then press "." to advance a frame at a
time), you may want to try adding "-noskip" to your command line.

Though nothing proved as successful as the above chain, I also had results
that seemed to be adjustable to some extent by using pullup with some
variation of the options available for tinterlace.

-- 
Laine Lee