[MPlayer-dev-eng] New inverse-telecine filter

[Zoltan Hidvegi]

> > But what I'm saying is that if it did give interlaced chroma, than it
> > would destroy the chroma of progressive frames, so encoding movies
> > from TV would look bad.  There is no correct way to subsample mixed
> > interlaced/progressive 4:2:2 into 4:2:0 without knowing what is
> > interlaced and what is progressive.
> 
>   I'm glad you phrased it this way because I think this is the subtext I
> was reading in your emails, and this is what I disagree with.  How does
> giving interlaced chroma all the time destroy the chroma of progressive
> frames?  The process of converting from interlaced 4:2:0 to progressive
> 4:2:0 once you have determined that two consecutive fields are from the
> same progressive frame should be a reasonably straightforward filter.  I
> understand that you cannot just merge them like you can with the luma,
> but it should not be very difficult to do something intelligent.

You are right, once you have 4:2:0, but to get 4:2:0 from 4:2:2, you
have to scale the chroma, and scaling interlaced content is not the
same as scaling progressive content.  Just try -vf scale on interlaced
content, it'll be a mess, because -vf scale is designed for
progressive scaling.  But the reverse is too, if you had an interlaced
scaler, like il=d:d,scale,il=i:i (which is not really correct, since
it would slightly shift the chroma), it will mess up your progressive
frames, it'll probably add some interlace artifacts to the
non-interlaced frames.

Zoli