On date Sunday 2010-04-18 16:10:35 -0700, S.N. Hemanth Meenakshisundaram
encoded:
> On 04/13/2010 12:05 AM, Stefano Sabatini wrote:
> >On date Monday 2010-04-12 22:52:44 -0700, S.N. Hemanth Meenakshisundaram
> >encoded:
> >[...]
> >>I had one (dumb?) question about yadif :
> >>
> >>Since the PicRef structure in libavfilter has no 'fields' member,
> >>how do I find out if the image data in the buffer is
> >>MP_IMGFIELD_ORDERED, MP_IMGFIELD_TOP_FIRST etc. Is it indicated in
> >>the PixelFormat?
> >Ehm no, I believe we need to add it to the AVFilterPicRef struct.
> >
> Hi,
>
> I have a couple more questions :
>
> 1. It looks like we output two frames for every input frame/field if
> mode is 1 or 3, else we output a single frame. Is this right?
>From the manual page of mplayer:
yadif=[mode[:field_dominance]]
Yet another deinterlacing filter
<mode>
0: Output 1 frame for each frame.
1: Output 1 frame for each field.
2: Like 0 but skips spatial interlacing check.
3: Like 1 but skips spatial interlacing check.
So yes you look right.
> 2. The mplayer code has a correct_pts variable that seems to allow
> doubling of frame rate. Here is it safe to assume this is always
> allowed/true?
I suppose so, that will also depend on the mode, from what I can
understand you'll need to duplicate the frame rate only with mode 1/3,
this will be done manipulating the output pts.
> 3. If I add a 'fields' member to AVFilterPicRef struct, how do I
> make sure it will be set to the right value (i.e. TOP_FIRST etc)?
It is a problem of the calling application to set this, for example an
application requests a picref with avfilter_get_video_buffer() and
sets the field accordingly. In the case of ffmpeg.c/ffplay.c this
depends on the options set in the codec context or by the user
(e.g. -top_field 1).
> Thanks
Regards.
_______________________________________________
FFmpeg-soc mailing list
[email protected]
https://lists.mplayerhq.hu/mailman/listinfo/ffmpeg-soc
