[mythtv] to gigem: is multirec on dvb-s supposed to work in current master?

David Engel david at istwok.net
Wed Jul 13 14:43:48 UTC 2016


On Tue, Jul 05, 2016 at 10:27:26AM +0200, Piotr Oniszczuk wrote:
> 
> > Wiadomość napisana przez Piotr Oniszczuk <piotr.oniszczuk at gmail.com> w dniu 29.06.2016, o godz. 09:52:
> > 
> > 
> >> Wiadomość napisana przez David Engel <david at istwok.net> w dniu 28.06.2016, o godz. 20:47:
> >> 
> >>> 
> >>> be fixed.
> >> 
> >> Nevermind.  It's fixed now.
> >> 
> >> David
> > 
> > David,
> > Perfect. This issue is now fixed!
> > 
> > I have however another Q: with recent changes every channel change is the same in contact: within the same mplex vs. between different mplexes.
> > 
> > Before - on my DVB-S tuners:
> > - channel change within the same molex was around 1-2sec and without ~0.5sec black screen
> > - channel change between different mplexes was 3-5sec with short black screen.
> > 
> > Now:
> > - every channel change has 3-5sec (or maybe even little longer) and always black screen.
> > 
> > If this is work-in-progress - forget this report.
> > 
> > If not - let me know how can I help with debugging/gathering more useful info for You info to improve situation.
> > 
> > br
> > 
> > 
> David,
> 
> I suspect current master issue (with longer channel change times within the same mplex) is result of regression in generation list of available inputs at channel change.
> 
> Namelly:
> 
> if I have chan1 and chan2 at mplex1 and using 4 virt.tuners (id17..20) for livetv then I see following:
> 
> -old code:
> starting livetv with chan1: tuner17 is selected
> changing livetv from chan1 to chan2: livetv still on tuner17
> changing livetv from chan2 to chan1: livetv still on tuner17
> 
> -new code:
> starting livetv with chan1: tuner17 is selected
> changing livetv from chan1 to chan2: livetv is on tuner18
> changing livetv from chan2 to chan1: livetv is on tuner17
> 
> It looks like in new code livetv - at channel change time - excludes current tuner from candidates to tune to new channel.
> 
> This hypothesis may explain:
> 
> - why in new code channel change time within the same mplex is the same long as between mplexes (as there is full tune on new tuner instead of retune on the same tuner with the same tuning parameters. Such retune are skipped by mythtv when old and new freq. are the same giving speedup for channel change within the same mplex)
> 
> - issue with his channels change I described in other replay in this thread.

I believe these issues are fixed.  I'm sorry it took so long.  It was
actually done last week, but the changes were mixed in with other
things I was working on that weren't ready yet.

David
-- 
David Engel
david at istwok.net


More information about the mythtv-dev mailing list