Mailing List archive

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

[linux-dvb] Re: Audio sync



Klaus Schmidinger writes:
 > Holger Waechtler wrote:
 > > 
 > > Marcus Metzler wrote:
 > > > Antti O Lammi writes:
 > > >  >
 > > >  > > Does anybody else have problem with audio sync, it is quite bad (nearly
 > > >  > > ridiculous) on Nexus DVB-T, not so bad on Nexus DVB-S.
 > > >  > > Yesterday I tested the version on CVS, no better than before.
 > > >  > > The driver I'm using now is linux-dvb.2002-12-08.tar.bz2 pack from
 > > >  > > vdr-project, no problem there.
 > > >  >
 > > >  > I have had the same problems since the LinkLayer CAM change for VDR (the
 > > >  > firmware changed). I can't use the new drivers/VDR at all because of this.
 > > >  > Still using VDR 1.1.19. ;-(
 > > >  >
 > > >
 > > > Does that also happen with our firmware (i.e. www.metzerbros.org/dvb)?
 > > 
 > > it's a bug (or missing feature?) in VDR, the PCR PID is not correctly
 > > set up. I think this will get fixed soon.
 > 
 > Well, I wouldn't know how!
 > 
 > Are you saying VDR should set VPID, APID _and_ PCR-PID when recording?

Yes.

 > Would it then receive extra TS packets with the PCR-PID?

Not with the old recording mechanism. But this is only an "AV7110 problem".


 > What should it do with those? Would they have to be used when
 > converting TS to PES in remux.c? What if PCR-PID == VPID?
 > 
 > So far for channels where PCR-PID == VPID things appear to work fine.
 > Since Oliver Endriss' driver fix live tv as well as replaying recordings work
 > great. So apprently in case PCR-PID == VPID there's nothing else VDR
 > would have to do, I guess.

But the current fix only works by chance. It sets a PID which is not
really set by the application. The real bug is the whole PID setting
mechanism which is still adjusted to the old firmware. There the
firmware scanned the PMT and set the PCR by itself. 
If you remember there also were syncing problems back then until I
introduced a AV PID reset which is triggered after the internal scan.
This also fixed a problem with recording you had back then (sometimes
the first packets were broken and sync messed up). I am wondering if 
this problem is also reappearing now!?

Anyway, now the application has to set the PCR.
Otherwise you will get lots of problems with other hardware which is
not so lenient. 

 
 > What's the reason to have an extra PCR-PID, anyway?

Some channels share the PCR. E.g. some Dutch transponders use the same
PCR for all of the channels on one transponder.
It probably depends on the kind of encoder they are using.


Ralph



-- 
Info:
To unsubscribe send a mail to listar@linuxtv.org with "unsubscribe linux-dvb" as subject.



Home | Main Index | Thread Index