[mythtv] [mythtv-commits] Ticket #4163: DVB-C in mythtv does not work while it works with mplayer/czap

Markus Schulz msc at antzsystem.de
Tue Dec 4 01:16:36 UTC 2007


Am Sonntag, 2. Dezember 2007 schrieb Rudy Zijlstra:
> >  Replying to [comment:4 anonymous]:
> >  > IMHO this bug is neither minor nor a candidate for an unknown
> >  > milestone.
> >
> >  It is. Your cable provider is to blame. The transmits a NIT with
> > DVB-S delivery descriptor.
>
> We are talking here about - a likely very small - level 4 in Germany.
> This can be as small as 1 housing complex. A lot of these guys have

here (Cottbus/Germany) there are ~100k citizen where about the half of 
them using such a "bad provider" and the others get the KD feed. I live 
in the bad area ;).

> neither the knowledge, nor the money to change their incoming signal
> to conform to DVB-C. What they typically do is to point a dish to
> Astra, receive the TSes, demodulate them, and re-modulate them into
> QAM64 and put them on the cable on some frequency. This they do with
> a de-modulator - re-modulator for each TS they handle. I suspect the
> smaller level 4 to have a fully independent de-modulator / modulator
> setup for each TS.
>
> For a lot of them, the abbreviation "NIT" will, i suspect, get you a
> blank look. Followed by: But the STB's are taking this well, so why
> should i spent (significant) money on this? And they are typically
> running on a shoe-string for budget......
>
> And yes, the STB manufacturers would love to see them go away, as
> they are a major headache to support in Germany.

but workaround this in each of their STB. But perhaps this was one 
reason why there are less variety of dvb-c STB compared to dvb-s.
But is this really a german problem? There are even transponder lists 
(with more then one transponder we need for NIT scan) for other 
countries inside dvb-utils package.

> The only way i am aware to handle such "networks" is a slow frequency
> scan, and to detect the services in each of the found TS. STB
> manufacturers are not happy with it, as it is slow and difficult to
> get fool-proof (you are not certain to which sat they are pointing,
> so additional info you would want to catch as well (EPG) may be
> difficult to handle).

Perhaps we can use the dvb-utils scan data files for dvb-c? Or let the 
user choose a start-frequency, frequency-step (8Mhz in most cases) and 
modulation(QAM64+256). Transport stream id for found multiplex we can 
get from scan (Network-id from NIT). To know if the NIT is completely 
valid or not we can check if modulation of found transponder was 
plausible.
The scan duration will be long, but will give working channel/multiplex 
tables. 

>
> So you are not alone in being unhappy with them.
>
> I am still hoping for a better solution in Myth.

me too.

> >  DVB-C with correct service information works as intended. Please
> > try as workaround to import a chnnels.conf.

upon my last attempt it wont work, cause each imported transponder was 
scanned (automatically during channels.conf import) and again updated 
with invalid data (QPSK instead of QAM64, ...) from wrong NIT. But i 
will try it again with svn-HEAD soon.

-- 
Markus Schulz


More information about the mythtv-dev mailing list