[MPlayer-dev-eng] [PATCH] Multiple problems with tfields filter

[Uoti Urpala]

On Mon, 2010-02-01 at 12:22 -0500, John Lindgren wrote:
> Uoti Urpala wrote:
> > It has worked better than that if -correct-pts is active since my commit
> > that enabled video filters to add frames back in 2006 (svn r18922),
> > though the filter doesn't implement exact pts calculation for the added
> > frames (just adds that 0.02 seconds to last frame pts). Since he
> > modified the value added to pts I think the behavior is probably exactly
> > right for his particular videos. In git behavior even without
> > -correct-pts is better now in the sense that you can at least frame step
> > over the added frames individually, though in that case they don't have
> > any individual pts generation (in svn it'll jump over both frames at
> > once).
> 
> 1.001 / 60 at least corresponds to NTSC, whereas 0.02 seems completely
> arbitrary.  With my patch, pts is calculated correctly (at least for
> NTSC content) even without -correct-pts.

It may be calculated correctly, but without -correct-pts it makes little
difference what the code calculates since the pts from the filter will
be ignored in most cases.

>   (Side note: Using -correct-pts
> and tfields, both before and after my patch, produces stunning screen
> corruption, so I don't know what you mean by it working "better".)

Is this really related to tfields? I suspect not. The internal mpeg
demuxer (which is most likely what you were testing with as you
mentioned DVD source) does not work with -correct-pts. Using a different
demuxer (either a file format other than raw mpeg, or "-demuxer lavf"
for raw mpeg content) should allow testing it.