Mailing List archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[vdr] Re: Problems with New AC3-over-DVB-S patch





Andreas Schultz wrote:

> Wolfgang Fiesel wrote:
> > Hi,
> >
> >
> > boesch@agrocom.com wrote:
> >
> >
> >>Hi list,
> >>with the latest patch from Wolfgang Fiesel AC3 out works fine but the video is
> >>stuttering. There is the same behaviour with the latast Nav3-patch which
> >>includes the AC3 patch. If I put AC3 to the soundcard with Alsa everything works
> >>fine.
> >>
> >>
> >
> > Does it happened on DVD or also on VDR playback ?
> > What happens if you miss the AC1=1 switch, same stuttering?.
> >
> > You are right, I've seen similar artefacts during DVD playback with NAVx patch. At
> > first I thought about the libcss and libdvdread stuffs. But using DVD under vdr0.99
> > is running perfectly with the same libs.
> > But the same behaviour happens without AC3 switch. Additional I've made the
> > experience skipping back and forward repairs the video sequence. Thus it seems the
> > main problem is not the AC3 handling but it correspondes maybe with wrong PTS time
> > stamps of AV synchronisation.
>
> The PTS time stamps are probaly to blame. I was already wondering about
> that, when io saw your patch. I had/have the same problem with AC3toPCM
> converted frames. It is apparently possible to get the AV syncing
> mechanism on the DVB card confused, resulting in described stuttering.
>
> For digital out the syncing might not be that important, did you try to
> just send the frames without PTS?
>

No really.
But If I compare the complete Audio frame extraction from vdr0.99x there is always a PTS
info available for AC3to PCM (start of LPCM frame, divided into 4*1536 byte packets). And
here the synchronisation is perfect(same as for direct AC3 output)
Now in the new NAV version there are nearly the same construction. But nevertheless
something is wrong here because of video stuttering.
As you mentioned before I'll try to deliver audio (PCM and AC3) without any PTS info to
see the difference. Perhaps extraction of PTS corresponds to a misleaded frame.

Wolfi

>
> Andreas
>
> > I'll try to investigate this more in detail.
> >
> > Wolfi

--
==============================================================
Dipl.-Ing. Wolfgang Fiesel                   EMAIL: fie@iis.fhg.de
Group Manager Audio Systems 2
Department: STUDIO                         PHONE: +49 9131 776-377
Am Weichselgarten 3                          FAX:   +49 9131 776-399
D-91058 Erlangen, Germany              WWW:   http://www.iis.fhg.de





Home | Main Index | Thread Index