[mythtv] Fwd: Playback sometimes freezes on jump or FF / stream time position problem

John Freer john_freer at yahoo.com
Thu Mar 31 12:35:15 UTC 2005


I too found a similar problem last night.

I was on 0.3.2i off at-bleeding, I think, and my mythbackend logs
were full of some sort of mpeg errors (hundreds of lines) that went
something like [MPEG2 at 0x234095] Warning Missing KeyFrame at 11 23 
and about 10 other types of messages repeating.

I also had a compounding problem with a corrupted recordingmarkup
table that I needed to drop and replace (thanks issac).  However,
even after I fixed that I still had flickering and jumpyness in the
playback.

I isolated the problem to ivtv since the errors / jumpyness didn't
exhibit with my bttv/mpeg4 card.  The problem also went away when I
downgraded to 0.2.0, at-stable.

So, double check your recordedmarkup table and check the backend log
for the mpeg2@ errors.  If you find 'em, consider downgrading ivtv.

YMMV.
JF



--- Louie Ilievski <loudawg at comcast.net> wrote:
> > What version of the ivtv driver are you using?
> 
> I'm using 0.3.2p, as you are now.  This very well may be the
> problem, but I 
> didn't assume so since it never has been in the past (except when
> it's 
> obvious from DMA errors and such).  Since I upgraded Myth and ivtv
> that same 
> night, I didn't consider ivtv to be the culprit.  I'll try
> downgrading it 
> tomorrow (err later today...need sleep first).
> 
> I can confirm though that the timestamp problem still exists with
> the 0.2 
> series.  I've been getting occasional DMA errors with it as well,
> and this is 
> why I upgraded to 0.3.2p, after hearing about some serious
> improvements to 
> the DMA code by Chris.  And, besides this freezing, it does run
> very very 
> well.  Not a single error since I installed it.
> 
> Anyway, I'll try downgrading again and see.  I will also try a
> 2.6.8 kernel 
> like the first reply suggested.  I'm having a hard time seeing why
> the kernel 
> would affect Myth behavior this way, but hey, anything's possible. 
> Ideally I 
> would like to keep this ivtv driver because of how well it has been
> running.  
> If the older kernel works with it, then I'll keep it.
> > _______________________________________________
> mythtv-dev mailing list
> mythtv-dev at mythtv.org
> http://mythtv.org/cgi-bin/mailman/listinfo/mythtv-dev
> 


More information about the mythtv-dev mailing list