[Mplayer-users] odivx <-> divx4 confusion

[Arpi]

Hi,

> Hi,
> I'm confused... why can I choose -vc odivx for DIV3 movies?
> And why is it much faster than DivX4Linux then? 
> 
> Is this a fallback to DivX4Linux, just with YV12 output, or 
> did you modify OpenDivX decore? 

RTFM DOCS/Divx4Linux.

if you compiled mplayer with divx4linux, then old opendivx
decore is disabled (because of variable/function name conflicts),
and the closed-source divx4 library is used, but with the opendivx
interface. (it is API-compatible with opendivx, as it's developed
from opendivx)

if you want to use the new API of divx4linux library (which is
designed for win32 interface) then you can do it using -vc divx4
the lib will do colrospace conversion and blitting, what is
required for win32, but we don't need it so it's pure double
slowdown (once the yv12->yuy2 in codec, and once as we need to
blit 30% more bytes to the video card)

so, -vc odivx and -divx4 are the same decoder core, but with different
API backend.

if compiled mplayer without divx4linux, then:
-vc divx4 - does not exists
-vc odivx - use the opensource opendivx decore v0.48alpha (it does NOT
decodes DIV3 then!)

yes i know it's hard to understand, and it's confusing...
but i had to disable opendivx decore, because it conflicts with
divx4linux library. and if divx4 has opendivx-compatible api too, why
don't we allow using it? :)
and it's even faster, at least if your card (or -vo null) supports YV12.
for RGB-only cards (or -vo x11) -vc divx4 may be faster than -vc odivx.

when i finished libvo2, and direct rendering will be possible, then
things will change. -vc divx4 has support for direct rendering, so
it will be faster than -vc odivx then, even for YV12.


A'rpi / Astral & ESP-team

--
mailto:arpi at thot.banki.hu
http://esp-team.scene.hu

_______________________________________________
Mplayer-users mailing list
Mplayer-users at lists.sourceforge.net
http://lists.sourceforge.net/lists/listinfo/mplayer-users