[MEncoder-users] interlaced/non-interlaced

Toerless Eckert Toerless.Eckert at Informatik.Uni-Erlangen.de
Wed Oct 6 01:20:45 CEST 2010


On Fri, Oct 01, 2010 at 03:43:46PM -0500, Rolf Ernst wrote:
> Mencoder has a sweet set of additional filters and IMHO far more  
> advanced input processing (i.e. to actually read a dvd instead of VOBs,  
> etc.) I know opinions differ here.

I still have encoding errors with mencoder and because i only archive
some content and can not visually check all of it, it's always a hard
bargain trying to see whether newer versions improve. So i even try
to build freaking static binaries so i can stick to an old, fairly well
known version of mencoder while the rest of my system (gentoo) continuously
updates.

So i wouldn't mind a better maintained software for a change although it's
not even clear to me whether the encoding bug i see are not also ffmpeg
created bugs ;-))

> The reason why deinterlacing at encoding time is advantageous is that  
> the codec has a much easier time compressing deinterlaced than  
> interlaced material.

Sure, but should that be my main concern ? I've got CPU time to spend
that's not the problem. The problem is the result at a given bitrate. 

> At the same bitrate you are likely to have poorer  
> results during interlaced encoding. (x264 has made great strides in this  
> area but there is no magic bullet).

Right. But i do not want to yet move from mpeg4/part2 to part10 for most
of my encodings (except HD) because of decoder compatibility issues in the
devices i use.

But what's wrong with me asking to be able to get real interlaced
content automatically encoded interlaced and non-interlaced encoding
encoded non-interlaced.  why can ffmpeg and/or mencoder not support
this automatic selection ??? At least as long as nobody proves me with
better arguments (DATA) that deinterlacing of real interlaced content
and then encodingdoes create a better result.

Cheers
    Toerless


More information about the MEncoder-users mailing list