Well, I just purchased a Nova-T card and, after installing a new aerial,
it all seems to be working fine.
Which is more than can be said for Sudbury transmitter. After talking
with the aerial installer who had been surprised that the Sudbury
transmitter referred to itself as Tacolneston the Strange Behaviour all
made sense.
I haven't read the DVB specs down to the last bit, but I'm guessing that
every multiplex not only provides an index of what it sends but also an
index of other multiplexes that are available. Great, this makes a lot
of sense and means you only have to find one multiplex and scan
discovers all the others for you.
Unless you live in Ipswich.
In which case the local transmitter (Sudbury) carries what must be a
bit-for-bit copy of the other local transmitter (Tacolneston) except, of
course, on different frequencies.
End result? You add one of the Sudbury multiplex frequencies to
initial.h, run scan, it produces details of that multiplexes channels
but on the wrong frequencies (well, right for Tacolneston) then goes
wondering of on a wild goose chase looking at the wrong set of
frequencies (for Sudbury, right for Tacolneston).
hmm, sorry -- here you have no choice but want to ask your provider to
transmit correct network information tables, they seem to be wrong...Being a DVB newbie this had taken me most of the evening to figure out.
By the way, the scan program also seems to occasionally produce the
wrong apid for some channels. Its sometimes one too high. So, for
example, it claims BBC 1 is uses vpid 600 (correct) and apid 602 (wrong,
it should be 601).
All this from linuxtv-dvb-1.0.0-pre2.
use the more recent source tree from CVS, there scan supports multiple
audio PIDs for multilingual services. Old version reported the one that
occured first (or last?) in the stream...