[mythtv] DVB 3.4 report

John Pullan jmp at tarantella.com
Fri Jan 7 05:15:30 EST 2005


On Fri, 2005-01-07 at 16:33 +1100, Hamish Moffatt wrote:
> Hi,
> 
> Just did some testing with the current CVS code plus the DVB 3.4 patch.
> 
> This is great work! Much easier than configuring all the PIDs by hand,
> and the associated problems when they occasionally change. I'm
> particularly looking forward to getting AC3 running here.
> 
> Here are some notes:
> 
> 1. Cancelled my very first scan of a transport because I input the
> frequency incorrectly. The setup program segfaulted. I didn't get a
> backtrace, sorry.
> 
> [...]
> 2005-01-07 14:29:18.269 DVB#0 WARNING - Status: SIGNAL | NO LOCK!
> ScanWizardScanner::cancelScan
> 2005-01-07 14:29:21.490 SIParser: Stopping SIScanner
> Segmentation fault
> 

Hmm I'll try this tonight.

> 2. We have five television networks here (Australia, in the capital
> cities at least) and each one has their own multiplex in each location.
> Their NIT information doesn't include the other stations. Hence I had to
> add a channel five times with each station's multiplex information.
> Not much can be done about that.
> 
> Each time I chose 'full scan', which scanned the new network to obtain
> NIT, then scanned each of the existing multiplexes. This wasn't too
> painful, although a new type of scan for new multiplexes only might be
> useful.

The other way to do this is to add the multiplexes by habd via advanced
and then scan each one (can't remember if scan all multiplexes was in
3.4)
 
> 
> The linux-dvb distribution includes user-contributed data for its
> 'scan' application which contains multiplex information. Would it be
> useful to be able to import that information?
> 
We do have plans to do this, but are currently trying to get it into the
main tree. So we're concentrating on that for the moment. I was actually
thinking of having some fancy(ish) UI so you could drill down to the
appropriate area in the world.

So I'd go
Europe->UK->Northern England->Winter Hill
Where the last three choices would allow you to scan with varying
degrees of speed.

There was also some talk of a full scan. But I'm not sure how possible
this is. I think the cards may need to be seeded with some basic
information. Inversion,Bandwidth, maybe more to get it to work.

> 3. Four of the five multiplexes include just a single NIT in their
> stream, but one of them includes entries for half a dozen different
> frequencies - but only one of them exists in this area. Is this legal?
> 
> It means that the scan (and any subsequent full scan) takes a long
> while, while the scans of all the non-existent multiplexes timeout.
> Not much you can do about that either really.
> 
As far as I can tell, it is.

> 4. Some of our networks include multiple services with different service
> IDs but the same PIDs. I don't know why they bother, and I don't know if
> they assign the same channel number to each (since I don't have a full
> DVB set-top-box to check with.) I guess the solution here is to mark the
> duplicates as not visible(?). Here's an example output from scan:
> 
> hamish at tv:~$ scan -c
> using '/dev/dvb/adapter0/frontend0' and '/dev/dvb/adapter0/demux0'
> 0x0000 0x0530: pmt_pid 0x0300 Seven Network -- 7 Digital (running)
> 0x0000 0x0534: pmt_pid 0x0340 Network Seven -- 7 HD Digital (running)
> 0x0000 0x0531: pmt_pid 0x0020 (null) -- 7 Digital 1 (running)
> 0x0000 0x0532: pmt_pid 0x0320 (null) -- 7 Digital 2 (running)
> 0x0000 0x0533: pmt_pid 0x0330 (null) -- 7 Digital 3 (running)
> 0x0000 0x0536: pmt_pid 0x0360 Seven Network -- 7 Guide (running)
> dumping lists (6 services)
> 7 Digital                (0x0530) 01: PCR == V   V 0x0301 A 0x0302       TT 0x0304
> 7 Digital 1              (0x0531) 01: PCR == V   V 0x0301 A 0x0302       TT 0x0304
> 7 Digital 2              (0x0532) 01: PCR == V   V 0x0301 A 0x0302       TT 0x0304
> 7 Digital 3              (0x0533) 01: PCR == V   V 0x0301 A 0x0302       TT 0x0304
> 7 HD Digital             (0x0534) 01: PCR == V   V 0x0341 A 0x0342       TT 0x0304 AC3 0x0343
> 7 Guide                  (0x0536) 01: PCR == V   V 0x0361 A 0x0362
> Done.
> 
> I'll have to go through and enter XMLTV IDs anyway, as we don't have any
> useful DVB guide data here.
> 
> 5. Setup all complete now. Started mythfrontend; tuned to the starting
> channel just fine and quickly with live TV. However any channel change 
> causes the backend to die:
> 
This is a known problem. There's a patch from John Poet pending which
cures this. So either apply that, or revert hdtvrecorder.* dtvrecorder.*
back to Jan 1

<snip>
Thanks for the feedback


-- 
John Pullan <jmp at tarantella.com>



More information about the mythtv-dev mailing list