[linux-dvb] Re: New Input file for dvbscan and new channels.conf for Bremen, Germany

[Johannes Stezenbach <js@linuxtv.org>]

Robert Schlabbach wrote:
> From: "Hans-Jürgen Ficker" <devnull@backmes.de>
> > Input file for dvbscan:
> > # DVB-T Bremen
> > # T freq bw fec_hi fec_lo mod transmission-mode guard-interval hierarchy
> > T 482000000 8MHz 1/2 NONE QAM16 8k 1/32 NONE
> > T 538000000 8MHz 1/2 NONE QAM16 8k 1/32 NONE
> > T 562000000 8MHz 1/2 NONE QAM16 8k 1/32 NONE
> > T 642000000 8MHz 1/2 NONE QAM16 8k 1/32 NONE
> > T 666000000 8MHz 1/2 NONE QAM16 8k 1/32 NONE
> > T 698000000 8MHz 1/2 NONE QAM16 8k 1/32 NONE
> 
> I don't believe this is correct, the fec_hi and guard-interval values are
> likely wrong. These parameters would yield 12.06Mbps bandwidth. From what I
> heard, the West German DVB-T transmitters are using FEC=2/3 and
> guard-interval 1/4, yielding 13.27Mbps bandwidth.
> 
> The file may work, but only because the DVB-T card detects the actual
> parameters and ignores the wrong ones from the file.
> 
> If possible, please use a tool which either parses the NIT in the received
> stream, or reads back the TPS parameters from the DVB-T card to get the
> _correct_ reception parameters.

It matches the values from teh channels.conf file, which are likely
created by dvbscan, so they are from the NIT.

http://web11.p15158686.pureserver.info/fileadmin/content/downloads/standorte_leistungen/gesamtuebersicht.pdf
(linked somewhere on http://www.dvb-t-nord.de/) says something different, though.

Johannes