[MPlayer-users] Re: [-] Re: [-] Re: [-] TV-recording with mencoder, "video-buffer full"

[Matthias Wieser]

Am Dienstag, 1. März 2005 02:42 schrieb D Richard Felker III:
> On Mon, Feb 28, 2005 at 10:26:39PM +0100, Matthias Wieser wrote:
> > Am Montag, 28. Februar 2005 17:01 schrieb D Richard Felker III:
> > > Bad suggestion, IMO. If you're going to use pp=fd, just capture a
> > > single field (height=288) instead. The quality will be comparable
> > > either way,
> >
> > My tests show the opposite.
> > (http://home.knuut.de/MWieser_/index.html)
>
> The opposite of what?

The opposite of "comparable quality" ;-)

> I didn't say pp=lb looks good (in fact it's very 
> bad). What I said is that pp=fd essentially throws away a whole field,
> and it has other flaws, so it should never be used.

I know it has flaws, but I have never seen ghost images and I think the 
other flwas are not as bad as when using pp=md, pp=lb, pp=ci,...

> > > since pp=fd essentially throws away one whole field
> >
> > ..only when it actually needs to deinterlace. Static scenes show much
> > better quality than pp=md, pp=ci, and so on.
>
> pp=md or perhaps kerndeint throws away the least for static scenes.

That's right. For that reason I have recommended to use kerndeint if the 
CPU is fast enough. But pp=md is nearly unusable because it produces 1x2 
pixel blocks whenever there is motion which looks very bad. 

> But both still have underirable side effects. It's impossible to fix
> the fact that interlacing is fundamentally mathematically flawed. All
> you can do is lessen the ugliness.

That's right, too (unfortunately!).

> > > (but also adds edge ghosting from the other field -- very ugly!)
> >
> > I have never seen those edge ghosting. Do you have screenshots?
>
> Yep, I'll make one right now:
>
> http://brightrain.aerifal.cx/~dalias/lavcdeint_sucks.jpg
>
> This shows 3 frames from telecined content, so you see how the 'ghost'
> of one appears in another.

Could it be that this effect is caused by ntsc? I have never seen anything 
like that (but I dont't watch Manga films).

> Also notice how bad the thin nearly 
> horizontal lines get aliased by the filter.

But probably less ugly than when using pp=md or pp=ci.
Have you done some comparisons with kerndeint?


> > ...and give a LOT less quality. If someone has a >2GHz Athlon,
> > kerndeint should be worth a try. It's imho the only deinterlace
> > included in Mplayer, that does not decrease the quality of static
> > scenes.
>
> The only remotely correct deinterlacer in MPlayer is vf_tfields. :)
> But it's only usable with MEncoder, and requires nasty hacks to trick
> to make it get the framerate right.

A year ago it worked quite well. But I had to use a higher bitrate to get 
reasonable quality. Now I do most TV recordings at 25fps.

Regards,
  Mat(waiting for TomsMoComp deinterlacer)thias