[linux-dvb] scan command returning wrong FEC ?
linuxtv at simong.net
Mon Dec 17 11:35:03 CET 2007
Thanks for the information Robert.
Any tips on getting the full ts from my card. I am using a budget PCI card
(Terratec Cinergy 1200 DVB-T SAA7146)
And I used tzap -r to set up the dvr device. Then I just used cat to cat
to the test.ts file ?
Or any tips on sniffing PID 0x0015 directly from the card.
> From: "Simon G" <linuxtv at simong.net>
> Unfortunately neither is a full transport stream - both files only
> contained PIDs 0x0418 and 0x0419, so it's useless for analysis :(
> Anyway, here's how the system works:
> In case of an SFN (single-frequency network), the multiplex is generated
> one place and SFN MIPs (megaframe initialization packets) are inserted
> which contain a GPS time synchronization offset and the transmission
> parameters to use. The transmission towers will then transmit the signal
> GPS time synchronized using the specified transmission parameters. In case
> of an MFN (multi-frequency network) where only one transmission tower is
> used, the transmission parameters are configured there.
> In any case, the DVB-T signal contains TPS (transmission parameter
> signalling) which contains the modulation type, FEC rates, hierarchy
> guard interval and transmission mode. In effect, the receiver only needs
> know the carrier frequency and the channel bandwidth, and can autodetect
> all other parameters. So it does no harm if it is given a wrong FEC rate -
> at worst, it will take a little longer to lock the channel as the receiver
> first needs to receive the TPS, whereas it might be able to lock faster if
> it is given all the right parameters.
> The transmission parameters given in the NIT are really only
> informational - they SHOULD be correct but it doesn't harm if they aren't.
> What actually affects the reception is the parameters given in the SFN MIP
> (as it influences which parameters the transmission tower uses) and the
> (as they influence what the receiver will use). Now it'd be interesting to
> see what happens if the transmission tower puts incorrect values in the
> bits, but I think that's quite unlikely to ever happen...
> Robert Schlabbach
> e-mail: robert_s at gmx.net
> Berlin, Germany
> linux-dvb mailing list
> linux-dvb at linuxtv.org
More information about the linux-dvb