[mythtv] packed image format and new deinterlacers

[Isaac Richards]

On Saturday 16 September 2006 1:10 pm, Markus Schulz wrote:
> Am Samstag, 16. September 2006 18:44 schrieb Isaac Richards:
> > On Saturday 16 September 2006 12:27 pm, Markus Schulz wrote:
> > > Currently i'm searching for the source where the fields will weaved
> > > together into the AVFrame (and later VideoFrame). Still not found
> > > yet.
> >
> > That's how they're decoded (and recorded, in the case of software
> > encoding).
>
> yes, but decoded from mpg2 stream will be seperate fields. And i look
> for the code where two fields will be build together to one (full
> sized) frame. But currently i can't find this code. Perhaps you can
> point me to this source?

Kind of difficult to do, because it doesn't exist in myth.

> I've found DecodeMPEG2Video and avcodec_decode_video calls (libmpg2 and
> ffmpeg). The resulting AVFrame from this calls contains a complete
> frame (two combined fields), therefore inside this calls the combine
> should go, but where?

Look in the libraries?  Regardless, though, depending on the source format 
being interlaced mpeg2 won't really work for everything.

Isaac