[MPlayer-dev-eng] Status of CoreAVC (last time) and using win32 codecs on x86_64
Michael Niedermayer
michaelni at gmx.at
Wed Jul 25 20:51:37 CEST 2007
Hi
On Wed, Jul 25, 2007 at 09:40:13AM -0400, Alexander Strange wrote:
>
> On Jul 25, 2007, at 6:20 AM, Michael Niedermayer wrote:
>
> > Hi
> >
> > On Tue, Jul 24, 2007 at 01:00:56PM +0200, e-t172 wrote:
> >> Arpi a écrit :
> >>> Hi,
> >>>
> >>>> On Mon, 23 Jul 2007 07:16:37 -0700
> >>>> Alan Nisota <alannisota at gmail.com> wrote:
> >>>>
> >>>>> Well, I continue to get a lot of mail about using CoreAVC with
> >>>>> mplayer,
> >>>>> so I thought I'd post a status update.
> >>>
> >>> WTF is coreavc/corecodec ???
> >>
> >> http://en.wikipedia.org/wiki/CoreAVC
> >>
> >> Right now CoreAVC is the only existing solution to play HD video with
> >> decent performance (unless you got your PC from the NASA :p )
> >> through a
> >> DirectShow filter. Other solutions based on DXVA (if your graphics
> >> card
> >> supports it) exist, but only proprietary, expansive, bloated software
> >> (PureDVD, Nero Digital) supports it.
> >>
> >> CoreAVC can be used with any DirectShow-compatible player, except
> >> mplayer which needs a few modifications (the patches we're talking
> >> about
> >> here).
> >>
> >> Yes, CoreAVC is proprietary. However there is no existing solution at
> >> this time to play HD (1080i) video in real time with a totally free
> >> software solution. I guess libavcodec could do that if ffh264 was
> >> multithreaded (for use with dual-core CPUs).
> >
> > see (Re: [FFmpeg-devel] [PATCH] h264 parallelized) on ffmpeg-dev for a
> > patch which adds slice based multithreading to libavcodec h264
>
> CoreAVC can do multithreaded decoding on single-slice h264.
well several people (like loren) have said since a long time how to
properly implement multithreaded decoding, if people choose to implement
something else and submit it so be it, i cant do anything about it ...
also i would guess that the various specs require a large minimum number of
slices for hd content, it would be quite broken design otherwise
> I assume it's parallelizing the motion compensation.
i dont assume anything as i do not know what they do and random assumtations
can lead to wrong conclusions
[...]
--
Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB
Good people do not need laws to tell them to act responsibly, while bad
people will find a way around the laws. -- Plato
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
URL: <http://lists.mplayerhq.hu/pipermail/mplayer-dev-eng/attachments/20070725/f35016de/attachment.pgp>
More information about the MPlayer-dev-eng
mailing list