[FFmpeg-user] (de)-interlacing question

[Toerless Eckert]

On Thu, Nov 24, 2016 at 01:01:43PM +0000, Andy Furniss wrote:
> Depends what you want/need to do. Personally I wouldn't de-interlace
> anything I wanted to keep, but then I wouldn't recode either - I mean
> gigs are far smaller than they used to be.

I am reevaluating after 8 years non-deinterlacing ;-)

Re-encoding still makes sense to me, archives are growing, and
you want to keep disks online, have backup, and replace after hmm.. 5..7 years ?

Re-encoding already re-encoded material is a different matter though.
trying to re-encode my interlaced mpeg4 part2 into part10 shows that
there's not much to be gained because my part 2 encoding already have
enough artefacts that re-encoding them causes rather high bitrate or
lowers quality (increases artefacts further). 

> If you must recode then it's possible to code h264 as mbaff - though
> care is needed WRT field order so you don't end up trashing.

Interesting technical option. I did not know about it. Have to check if
my older HW-devices for h264 support it.

But deinterlacing can also serve to improve quality because it can
be done non-real-time.

> yadif=1 for field rate seems mostly good enough. mcdeint can be better,
> but takes ages. Some of the others I find on SD that's going to get
> scaled on playback, look a bit crap on diagonals.

Recipe from doom9 forum that seems to work well:

yadif=1:0,mcdeint=0:0:10,framestep=2

> Depending on what GPU/TV you have you could in theory get a nice
> de-interlace on playback. Intels motion-adaptive vaapi looked OK when I
> tested it some time ago. It's even possible, though tricky, to get some
> TVs to deint for you, if they automagically deint when in an interlaced
> mode.

Wouldn't want to switch back to NTSC/PAL resolution on output these
days just to use display device deinterlacing. Messes up any type of
GUI or windowed playback. But yes, thats how i deal with deinterlacing
many years.

Playing with KODI 16, it looked as if the "software" deinterlce was better
than the DXVA of my nvidia card. Aka: encoder deinterlacing would
also give me more persistent playback quality if i have a range of devices.

Toerless

> >*sigh*
> >
> >>>I thought it might have gotten a lot easier through all the
> >>>experience collected with motion estimation. Aka: work in the
> >>>DCT domain, interpolate motion vectors and residual error - or
> >>>something like that.
> >>
> >>AIUI encoders get it easy in comparison to interpolation. An
> >>encoder has the ground truth for reference, so even if it can't
> >>find good motion vectors it can correct the difference with the
> >>residual or intra code a block.
> >
> >Use ground truth from 50p recordings to create 25p reference streams
> >to train a neural network. Nnedi already seems to use a neural
> >network for deinterlacing. Would guess it's using a similar
> >approach.
> 
> IIRC it just scales up fields - albeit nicely.
> 
> I've never seen a paper that uses neural networks - which doesn't mean
> there isn't one.
> 
> 
> 
> _______________________________________________
> ffmpeg-user mailing list
> ffmpeg-user at ffmpeg.org
> http://ffmpeg.org/mailman/listinfo/ffmpeg-user
> 
> To unsubscribe, visit link above, or email
> ffmpeg-user-request at ffmpeg.org with subject "unsubscribe".

-- 
---
tte at cs.fau.de