[linux-dvb] DVB-T South Africa

Patrick Boettcher patrick.boettcher at desy.de
Mon May 12 14:22:20 CEST 2008


I'm now 100% sure that this is a DVB-H transport stream.

the appearing/disappearing TS-PIDs in the dvbtraffic are indicating that 
timeslicing is active.

The dvbsnoop clearly says that there are only MPE-sections which is 
another indicator for that.

I will commit my two little, proof-of-concept-like, tools soon and tell 
you where to find and how to try it.


On Mon, 12 May 2008, Rogan Dawes wrote:

> Patrick Boettcher wrote:
>>  Hi Rogan,
>>  your dvbtraffic output raises a question: What happens when you run it for
>>  several seconds ?
>>  Are the PIDs always the same? Especially the one with the higher bitrate?
>>  I'm asking because if that is the case, it could be that this is a DVB-H
>>  transmission.
>>  I have some tools (which I did not commit yet) which "scan", in a very
>>  basic way, for DVB-H services, maybe this could help you.
>>  Before that you can try to use dvbsnoop on PID 0x00 and 0x10 to see
>>  whether it signals a INT-section.
>>  I could also be a pure radio transmission, but in that case scan should
>>  detect those channels.
>>  Patrick.
> Hi Patrick,
> Actually, I think you may well be right. Our cell networks are trialling (or 
> even deploying) DVB-H, and the content is provided by MultiChoice. 
> Unfortunately (for me) that content is almost definitely encrypted.
> I guess I might have to retry w_scan to see if it picks up any other 
> frequencies that might have the real DVB-T signals on them. And maybe improve 
> my antennae - I am currently using a Technisat DigiFlex TT2, which is just 
> sitting on my desk.
> I am attaching the results of "dvbsnoop -s pidscan", as well as a longer 
> capture of dvbtraffic (using "dvbtraffic | tee dvbtraffic.txt", then Ctrl-C 
> after 6-7 seconds).
> Thanks for your help.
> Regards,
> Rogan

More information about the linux-dvb mailing list