[MPlayer-dev-eng] [PATCH] bigendian fix for hwac3

[Ulion]

2007/9/15, Reimar Döffinger <Reimar.Doeffinger at stud.uni-karlsruhe.de>:
> Hello,
> On Sat, Sep 15, 2007 at 06:04:14PM +0800, Ulion wrote:
> > Since this patch applied, I'd like to introduce a little fix for
> > data-type-dependent bits when construct iec61937 ac3 header (vlc
> > already did this years ago). That's
> > the 8 to 10 bits of buf16[2], it should be the bsmod value from input
> > ac3 buffer.
> >
> > To verify that, I checked iec61937 from some guy. It did say that bits
> > should be bsmod from ac3 frame.
> >
> > So I made another patch here.
>
> I already saw the last time you posted it.
> What effect does this change have?

I don't known it exactly, but the iec said that and vlc did that for
years. I can send you the iec about these.

> And could please some other people test all those patches? Is there only
> one single person who cares enough about this to test if it breaks
> things or not??

I'm really care about the ac3/dts passthrough function because I has a
macbook and a creative external decoder and a 7.1 sound system. I
still prefer to use mplayer instead of vlc or qt, so I made these
patches. To test bigendian, I even bought an old iMac G5, that's a
real cool machine:)
The main patch for mac osx support ac3/dts passthrough I posted last
month, I'm still fixing it to be better on my iMac G5 and will be post
soon.

-- 
Ulion