[MPlayer-users] new version of detc filter! (inverse telecine)

D Richard Felker III dalias at aerifal.cx
Thu Mar 13 07:19:28 CET 2003


On Wed, Mar 12, 2003 at 10:38:52PM -0500, Robert Henney wrote:
> [Automatic answer: RTFM (read DOCS, FAQ), also read DOCS/bugreports.html]
> On Sun, Mar 09, 2003 at 01:07:12PM -0500, D Richard Felker III wrote:
> > [Automatic answer: RTFM (read DOCS, FAQ), also read DOCS/bugreports.html]
> > I just committed it. Please send any comments to the list here.
> 
> interlacing is still very visible with ADV Films' Samurai-X.  this certainly 
> may be due to a poorly done telecine process which is beyond the hope of
> any filter to correct, but I wanted to mention it just in case.

Please upload a few of the worst clips to mphq and I'll see if I can
make it work.

> btw, this new version of the detc filter appears to use a huge amount more 
> cpu than the old one.  enough that when the filter's on, it's a bit beyond
> the capability of my primary multimedia machine.  not really a nitpick,
> but rather just an observation.

Yes, it's known. I didn't intend for it to be used in realtime, since
every 4th frame will still be shown twice as long, looking really bad
during pans. And the cpu time used isn't really significant with
mencoder since the actual encoding will take much more time.

However, there are some very obvious mmx optimizations that can be
done to make it much faster than the old version. I just haven't had
time to write the code yet.

Rich



More information about the MPlayer-users mailing list