[MPlayer-dev-eng] mp-g2 pre14

[D Richard Felker III]

On Sun, Apr 20, 2003 at 10:46:26PM -0400, D Richard Felker III wrote:
> On Mon, Apr 21, 2003 at 03:17:52AM +0200, Arpi wrote:
> > > The callback idea is just extremely ugly, but it may be the
> > > only way...
> > it seems :(
> > or we have to disallow increasing fps in filters.
> 
> Hey, Arpi... Billy Biggs just suggested this to me on irc: how about
> we replace the architecture completely and use a pull model for the
> video filter chain, rather than push? I.e.:

Hmm, the more I think about this design, the more I like it. It
actually eliminates some horrible hacks in the current vf_ivtc design.
Right now, I delay the video by one frame in order to be able to look
at the following frame for making the decision whether to merge. Not
only does this result in slight a/v desync when playing (altho not in
mencoder); it also means that the final frame of the movie will be
lost. All these problems just fall away naturally when the push model
is replaced with a pull model.

As for DR and slices, I don't see any particular problems with this
design. Each vf can just pass the slice callback and/or DR buffer back
to the previous filter. The only area where I see a little bit of
confusion is negotiation of config and image format/dimensions, since
that normally starts at the codec end. There shouldn't be
insurmountible problems, but we should at least think through how to
handle stuff. I'm sure it can't be much worse than the negotiation
system already is, though...

Rich