[FFmpeg-devel] dealing with tables in DV codec

Balatoni Denes dbalatoni
Thu Sep 18 23:24:49 CEST 2008


Wednesday 10 September 2008 23:06-kor Michael Niedermayer ezt ?rta:
> balatoni denes tried to get a half optimized idct in and i nicely told
> him that it has to include all tricks and optimizations we could think of.
> We ended up with something that was clearly quite a bit faster than
> what he submitted first, that is between all the complaints about how
> little time he had and how evil we were for not just acceptiing it as
> is.

The Legend of the Warrior - now I am officially part of the ffmpeg 
mythology :)

> > What I'm aware of, though, is that Sun's medialib on SPARC seems to
> > perform better than what's currently in FFmpeg (for obvious reasons).
> our shiny idct isnt enabled by default as its not bitexact to any other
> ANSI C idct.
> That besides reminds me that rumors say the mlib idct is so inaccurate
> that its practically useless.

Yes, the inaccuracy was the sole reason for writing the new idct. 
Unfortunatelly it was not written using Walken's more accurate algorithm - 
which was posted after the simple idct was already mostly done.


More information about the ffmpeg-devel mailing list