[FFmpeg-devel] H.264 + PAFF: BBC HD recording shows extreme interlacing artefacts

[Jeff Downs]

On Mon, 5 Nov 2007, Reinhard Nissl wrote:

> > To me, when FIELD_PICTURE is true, the patch says:
> > 
> > top_field_first = MIN(top_poc, bottom_poc) == bottom_poc
> > 
> > or, more simply:
> > 
> > top_field_first = bottom_poc <= top_poc
> > 
> > Or do I miss something?
> 
> I've hacked in this equation and it works opposite to mine. What I have
> seen in my previous tests was that for field pictures poc pointed to the
> second decoded field at that source location and hence the
> top_field_first had to be set when poc pointed to the bottom field.

Are you dealing with field pictures all the time? The above was analyzed 
in context of field pictures only (ie, I removed the variability of 
FIELD_PICTURE from your original).  What you had for !FIELD_PICTURE looked 
like it would be plausible.

poc does not necessarily have to be the second decoded field's poc.
It should be the minimum of the two field pocs, regardless of decoding 
order, once a field pair is complete.

As an aside, are you testing this directly with ffmpeg, or libxine as well 
(like in your other post).  Reason I ask is that the same problem 
occurring there might be impacting things here, too.

	-Jeff