[MPlayer-dev-eng] New inverse-telecine filter

D Richard Felker III dalias at aerifal.cx
Sat Dec 20 14:05:58 CET 2003


On Fri, Dec 05, 2003 at 04:08:10PM -0600, Zoltan Hidvegi wrote:
> > On Fri, Dec 05, 2003 at 12:00:28PM -0600, Zoltan Hidvegi wrote:
> > > > > > Also, the way hard-interlaced mode alternates a couple times per
> > > > > > second between using bottom and top field as a source of interpolation
> > > > > > creates a horrible low-frequency flicker.
> > > > > 
> > > > > Yes, but what else can I do, that's the only way to keep the output
> > > > > rate even for interlaced content.  Ive started to write a motion
> > > > 
> > > > Well then at the very least, you need a better filter.
> > > 
> > > Yes.  Any suggestions?  That's the same problem as tfields, that
> > > flickers too.
> > 
> > Yes, but it flickers at a steady, rapid 60 Hz, and your eyes filter
> > that out, just like on a real TV. On the other hand, 5 Hz flicker
> > looks horrible.
> 
> Yes, I know it is ugly, but the only filter I know that could
> eliminate that is motion-compensated deinterlacing, and you know that
> it's a quite complex task to write such a deinterlacer.  Or do you
> know any simple kernel filter that does not flicker, and still OK
> quality (pp=lb is not OK for me)?  You can to top-fiel all the time,
> but then the motion will not be smoth.  What your filter does?  Output
> at 60fps, since you can do variable rate in G2?

Hey, here's a cheap idea that might look decent. Spacially interpolate
your interlaced fields up to full frames (60 fps), then temporally
interpolate with a linear filter of some sort (perhaps quadratic
approximation in time). That is, sample your 24fps frames from the
60fps ones like:

60:  0-----1-----2-----3-----4-----5-----6-----7-----8-----9-----
24:  -00-------------11-------------22-------------33------------

Any ghosting should only be minor, and by interpolating from both top
and bottom fields, you'll avoid some (hopefully most) of the flicker.
Of course it'll take some processing time... :(

Rich




More information about the MPlayer-dev-eng mailing list