On Tue, Dec 09, 2003 at 10:19:47AM +0000, jp wrote: > > IMHO scan is too clever for its own good. It does it the right way(tm) > but in practical terms not every transmitter is set up the right way. > > Heck some transmitters don't even include the NIT. > > The only safe approach is to try to scan each RF channel individually. > (using scan -c) > Sure this will take a while (there are ~46 RF channels used in the UK) > the first time, but once you establish what frequencies you > need you can rescan them to pick up new channels/revised pids etc.
Yes, this should be implemented as a fall back strategy for DVB-T and DVB-C. But because it usually takes more time it shouldn't be used unless really necessary. > The check "have I scanned this freq before?" combined with -n flag > should get you all channels.conf in a single pass assuming the > transmitter has a well-formed NIT. > > In boundary areas, like where Gavin is, where the best signal comes from > one transmitter for some multiplexes and some from another, scanning > each possible freq and picking those with the better signal (by reading > SNR from the card if that is supported on your card) is the most > practical approach. I think www.dvb.org has some interesting white papers about DVB-T, and also about mobile DVB-T reception, just in case someone wants to do some research. (Unfortunately I don't have time to read them all by myself ;-( ). And BTW, http://www.engr.sjsu.edu/rmorelos/ee252s03/WhyCOFDM.pdf is a nice paper which explains how COFDM works. Regards, Johannes -- Info: To unsubscribe send a mail to [EMAIL PROTECTED] with "unsubscribe linux-dvb" as subject.
