Re: [xine-devel] libac3

[Guenter B. <bar...@st...>]

Hi Siggi,

On Wed, 6 Dec 2000, Siggi Langauf wrote:

> I guess, that's the only chance we have. Does the new lobac3 code work in
> any other context? Does it work with oms? Or are there general flaws
> "current" libac3?

looking at the livid mailing lists I really get the impression not many
people got oms or a standalone ac3dec compiled and working? ;>

> I agree that we need a new interface, but it shouldn't generally be
> 5.1. The audio driver should be able to request stereo, 5.1 or undecoded
> AC3 streams, so we can use optimal decoding algoritms for any supported
> hardware. Decoding 5.1 just to downmix the 5 channels to stereo again is
> nonsense, and decoding the AC3 signal doesn't make sense if you want to
> use an external AC3 decoder connected to your sound card's digital out...

> I agree that we need a central point for downmixing/resampling, but if we
> can avoid those steps (or at least downmixing), that's even better.
> Would it make sense to do the downsampling in libac3?
> Alternatively, we could implement some kind of audio filter for this,
> which would be shared by all the audio drivers...

we talked about that last thursday, but now again for everybody: I think
this is a good idea, I really like that API design.

cheers,

   Guenter


-- 
   xine - a free video player

   http://xine.sourceforge.net

NEW: xine mailing list:
http://lists.sourceforge.net/mailman/listinfo/xine-user