[MEncoder-users] Re: Re: Re: new doom9 codec comparission (submission)

Michael Niedermayer michaelni at gmx.at
Mon Dec 19 03:15:31 CET 2005


On Sun, Dec 18, 2005 at 08:33:00PM -0500, Rich Felker wrote:
> On Mon, Dec 19, 2005 at 12:09:22AM +0100, Doom9 Feedback Hotline wrote:
> > >Of course if you're using different postprocessing the output will be
> > >different.... What Michael is asking for is a comparison of the
> > >decoders, not the postprocessing.
> > 
> > You have an answer to everything, don't you? But unless you grew up in a 
> > country where they speak German, my German is better than yours and he 
> > asked to use the same decoder and same postprocessing settings. Which is 
> > exactly what I did by selecting one decoder to decode them all.
> Indeed I agree, you must use the same postprocessing for each mpeg4
> codec for the tests. (Personally I would use none at all, but that's
> not really relevant.)
> However, Michael was asking if there are any bugs in lavc's decoding
> of xvid files, not if ffmpeg's postproc code is the same as xvid's.

hmm, doom9 said:
> >and experience during this comparison, I actually find the result of
> >decoding xvid decoded by xvid more visuallypleasing than when xvid is
> >decoded by lavc (and that's in the avi container where the auto idct
> >selection works just fine).XviD stands a lot less to lose by being decoded

and my reply was just to show this to be false, as such comments are
making lavc look bad to who reads this
the whole thing really is just a silly missunderstanding, due to everyone
saying one thing and meaning something slightly different

libavcodec isnt equal ffdshow, libavcodec and xvid decode most mpeg4
files binary identically if lavc either automatically or manually chooses
the xvid mmx idct. (yes neither xvid nor lavc will decode xvid well if
mmx is unavailable)
the files i found where decoding differs generally are broken
in one way or another, and lavc is and has always been better at decoding
these as a large number of people on non-windows OS / non-x86 depend on it/
send us bugreports ...

that decoding with ffdshow on windows looks worse then xvid on windows is
a different issue which has nothing to do with the decoders itself, and
i would dare to guess the pp code isnt the cause either, both pp
implemantations, implement the same algorithm IIRC
also note iam also always interrested in bugreports which would show
examples where my pp code is worse, if noone sends us bugreports we
cant know about the issues and cant fix them ...

encoding then OTOH is a totaly different issue, xvid seems to have won
against lavc this time, iam not saying the comparission was perfect
(not knowing about the decoder before, no double blind tests?)
but i dont think a perfect comparission would have changed the results

its IMHO also better to improve lavc / the used paramaters instead of
simply claiming it to be better, this somehow reminds me of thouse who claim
that the linux kernel is supperor to windows in every aspect, it surely
is in many aspects but not all, and before a issue can be fixed its needed
to accept that there is an issue, now strangely this reminds me of gcc and
how they deal with asm bugs too ...



More information about the MEncoder-users mailing list