[FFmpeg-devel] [PATCH]lavfi/telecine: Mark telecined frames as interlaced

[Carl Eugen Hoyos]

Am So., 12. Apr. 2020 um 15:16 Uhr schrieb Paul B Mahol <onemda at gmail.com>:
>
> On 4/12/20, Carl Eugen Hoyos <ceffmpeg at gmail.com> wrote:
> > Am So., 12. Apr. 2020 um 15:00 Uhr schrieb Paul B Mahol <onemda at gmail.com>:
> >>
> >> On 4/12/20, Carl Eugen Hoyos <ceffmpeg at gmail.com> wrote:
> >> > Am So., 12. Apr. 2020 um 11:35 Uhr schrieb Paul B Mahol
> >> > <onemda at gmail.com>:
> >> >>
> >> >> On 4/12/20, Carl Eugen Hoyos <ceffmpeg at gmail.com> wrote:
> >> >> > Am So., 12. Apr. 2020 um 10:38 Uhr schrieb Paul B Mahol
> >> >> > <onemda at gmail.com>:
> >> >> >>
> >> >> >> On 4/11/20, Paul B Mahol <onemda at gmail.com> wrote:
> >> >> >> > On 4/11/20, Carl Eugen Hoyos <ceffmpeg at gmail.com> wrote:
> >> >> >> >> Am Sa., 11. Apr. 2020 um 15:10 Uhr schrieb Paul B Mahol
> >> >> >> >> <onemda at gmail.com>:
> >> >> >> >>>
> >> >> >> >>> On 4/11/20, Carl Eugen Hoyos <ceffmpeg at gmail.com> wrote:
> >> >> >> >>> > Am So., 5. Apr. 2020 um 02:05 Uhr schrieb Carl Eugen Hoyos
> >> >> >> >>> > <ceffmpeg at gmail.com>:
> >> >> >> >>> >>
> >> >> >> >>> >> Am So., 5. Apr. 2020 um 01:02 Uhr schrieb Carl Eugen Hoyos
> >> >> >> >>> >> <ceffmpeg at gmail.com>:
> >> >> >> >>> >> >
> >> >> >> >>> >> > Am Sa., 4. Apr. 2020 um 00:44 Uhr schrieb Carl Eugen Hoyos
> >> >> >> >>> >> > <ceffmpeg at gmail.com>:
> >> >> >> >>> >> > >
> >> >> >> >>> >> > > Am Sa., 4. Apr. 2020 um 00:40 Uhr schrieb James Almer
> >> >> >> >>> >> > > <jamrial at gmail.com>:
> >> >> >> >>> >> > > >
> >> >> >> >>> >> > > > On 4/3/2020 6:37 PM, Carl Eugen Hoyos wrote:
> >> >> >> >>> >> > > > > Am Fr., 3. Apr. 2020 um 23:19 Uhr schrieb Carl Eugen
> >> >> >> >>> >> > > > > Hoyos
> >> >> >> >>> >> > > > > <ceffmpeg at gmail.com>:
> >> >> >> >>> >> > > > >
> >> >> >> >>> >> > > > >> Attached patch marks actually telecined frames as
> >> >> >> >>> >> > > > >> interlaced,
> >> >> >> >>> >> > > > >> other frames as progressive.
> >> >> >> >>> >> > > > >
> >> >> >> >>> >> > > > > New patch with changes to fate attached.
> >> >> >> >>> >> > > > >
> >> >> >> >>> >> > > > > Please comment, Carl Eugen
> >> >> >> >>> >> > > >
> >> >> >> >>> >> > > > Those yadif tests look wrong. The patch shouldn't
> >> >> >> >>> >> > > > affect
> >> >> >> >>> >> > > > them.
> >> >> >> >>> >> > >
> >> >> >> >>> >> > > Clearly, thank you!
> >> >> >> >>> >> > >
> >> >> >> >>> >> > > New patch attached, it should now only change the
> >> >> >> >>> >> > > telecined
> >> >> >> >>> >> > > frames and leave the other frames as they are, the
> >> >> >> >>> >> > > setfield
> >> >> >> >>> >> > > filter can be used to force a progressive setting for
> >> >> >> >>> >> > > them.
> >> >> >> >>> >> >
> >> >> >> >>> >> > New patch attached that also sets top_field_first
> >> >> >> >>> >>
> >> >> >> >>> >> Which had the effect that fate is correct again, new patch
> >> >> >> >>> >> attached.
> >> >> >> >>> >
> >> >> >> >>> > Patch applied.
> >> >> >> >>> >
> >> >> >> >>>
> >> >> >> >>> This was never approved by me.
> >> >> >> >>
> >> >> >> >> You reviewed it on irc and correctly pointed out the missing
> >> >> >> >> bits.
> >> >> >> >
> >> >> >> > Lies, I was against that idea from start.
> >> >> >> >
> >> >> >> >>
> >> >> >> >>> So revert it ASAP!
> >> >> >> >>
> >> >> >> >> What should be changed about it?
> >> >> >> >
> >> >> >> > Return of code as it was before this pointless change.
> >> >> >> > I see no good out of it.
> >> >> >>
> >> >> >> I gonna revert this ASAP!
> >> >> >
> >> >> > Could you explain why it is wrong to mark interlaced frames
> >> >> > as interlaced?
> >> >>
> >> >> The frames are not interlaced.
> >> >
> >> > Using the usual 3:2 telecine, the filter outputs two progressive
> >> > frames, followed by three interlaced frames, the patch should
> >> > mark the interlaced frames as interlaced and I believe it does.
> >>
> >> You are very ignorant or very stupid or both.
> >
> > Apparently yes because ...
> >
> >> Interlaced frames are frames produced by interlacing.
> >> Telecine is not interlacing.
> >
> > ... to the best of my knowledge, the telecine process outputs
> > interlaced (and non-interlaced) frames, so I do not understand
> > your argumentation, please elaborate.
>
> Interlacing usually destroys half of data, telecine never does that.

There are cameras that output interlaced content, they do not
destroy any data (the "missing" data never existed).

I don't think your definition is ideal, a more useful definition is that
the fields of one frame originate from different points in time.

> Claiming frames are interlaced will just confuse confused users more.

I was more thinking of encoders, they will be less confused with the
patch.

> >> >> I thought you knew that interlacing destroys half of data.
> >> >> Telecine does not destroys data.
> >> >
> >> > Telecine duplicates some data, leading to interlaced frames.
> >> > A (perfect) detecine process can remove the duplicated data
> >> > (and the interlaced frames).

Carl Eugen