dvb-t retuning in southern England

As almost everyone living in southern England knows, the old analogue TV signals are being switched off area by area. Each time this happens, the DVB-T channels are re-arranged and so TVs etc. need to be “re-tuned”.

This happened to us (near Oxford) recently and we managed to get the TVs and Humax DVR sorted fairly easily, but getting Kaffeine to work with the TV card in my desktop PC has stumped me. I’m running openSuSE 11.4 64 bit, KDE4 and it worked fine until the switchover.

As far as I can see, the key files are:

kaffeine’s private config file at ~/.kde4/share/apps/kaffeine/config.dvb

[device]
deviceId=P113171331461f936
frontendName=Zarlink MT352 DVB-T
configCount=1
[config]
type=2
name=Terrestrial
scanSource=uk-Oxford
timeout=1500

and the out of date scan data from package dvb at /usr/share/dvb/dvb-t/uk-Oxford (file timestamp 18 Feb 2011)

# UK, Oxford
# Auto-generated from http://www.dtg.org.uk/retailer/dtt_channels.html
# and http://www.ofcom.org.uk/static/reception_advice/index.asp.html
# T freq bw fec_hi fec_lo mod transmission-mode guard-interval hierarchy
T 578000000 8MHz 3/4 NONE QAM16 2k 1/32 NONE
T 850000000 8MHz 2/3 NONE QAM64 2k 1/32 NONE
T 713833000 8MHz 2/3 NONE QAM64 2k 1/32 NONE
T 721833000 8MHz 3/4 NONE QAM16 2k 1/32 NONE
T 690000000 8MHz 3/4 NONE QAM16 2k 1/32 NONE
T 538000000 8MHz 3/4 NONE QAM16 2k 1/32 NONE

Kaffeine just shows a handful of channels - I think this is just one multiplex. It also shows the initial scan file as dated “08/09/2011”, whether this means 8 September or 9 August is irrelevant as both are prior to the final transmitter changes.

I tried updating the uk-Oxford file using w_scan, saving over the original:

T 706000000 8MHz AUTO AUTO     AUTO AUTO AUTO AUTO
T 730167000 8MHz  2/3 NONE    QAM64   8k 1/32 NONE      # Oxon & Bucks
T 785833000 8MHz  2/3 NONE    QAM64   8k 1/32 NONE      # Oxon & Bucks
T 802000000 8MHz  2/3 NONE    QAM64   8k 1/32 NONE      # Oxon & Bucks
T 777833000 8MHz  3/4 NONE    QAM64   8k 1/32 NONE      # Oxon & Bucks
T 746000000 8MHz  3/4 NONE    QAM64   8k 1/32 NONE      # Oxon & Bucks

but this behaves just the same. It even shows the same ambiguous date, which it isn’t getting from either of these two files!

There are no other files named uk-Oxford on the machine

Surely someone else must have encountered this issue - any pointers gratefully received!

I suspect the ‘channels’ are actually the multiplexes as we had six multiplexes to retune in Yorkshire. A quick look at http://www.dvb.org/technology/fact_sheets/WP01_DVB-T Hierarchical.pdf accessed from Hierarchical modulation - Wikipedia, the free encyclopedia suggests that you have a combination of HD and other multiplexes.

Apologies for my sloppy language - yes, the tuning is actually to multiplexes. These have been re-arranged, and use new frequencies (no longer matching what is in the uk-Oxford file).

Although there is updated channel info at Digital UK - Home, this gives only the short channel numbers such as 53+, which is the BBC A (primary) Mux). I can’t convert these to frequencies - can anyone help?

Thanks
Richard

OK, I never did manage to map channels to frequencies, but I have most of a solution now.

I created a new scanfile.dvb with header data

# My version
# RPM 20 Oct 11
[date]
2011-10-20
[dvb-t/uk-Oxford_RPM]

Then ran w_scan with the -x flag rather than -k: (*)

w_scan  -c GB -x >> scanfile.dvb

Then I started kaffeine, select “Television” → “Configure Television”. The main tab shows today’s date for the scan file, a big clue that it’s got the right one. Next, I selected the tab for the DVB card [Device 1], and changed the source to “uk-Oxford_RPM”. Exit the setup and scan for channels, it finally found them.

(*) w_scan on-line help states -k as kaffeine format, but this doesn’t work. -x is described as “(dvb-)scan format” but does work.

I now have some duplicate channels in my multiplex list, if I could map to the short names (see previous post) the digital-uk site would allow me to determine which line(s) these are. Nearly there!

For the record, the file is now:


# My version
# RPM 20 Oct 11
[date]
2011-10-20
[dvb-t/uk-Oxford_RPM]
#------------------------------------------------------------------------------
# file automatically generated by w_scan
# (http://wirbel.htpc-forum.de/w_scan/index2.html)
#! <w_scan> 20110702 1 0 OFDM GB </w_scan>
#------------------------------------------------------------------------------
# location and provider: <add description here>
# date (yyyy-mm-dd)    : 2011-10-20
# provided by (opt)    : <your name or email here>
#
# T[2] <freq> <bw> <fec_hi> <fec_lo> <mod> <tm> <guard> <hi> # comment]
#------------------------------------------------------------------------------
T 674166670 8MHz  3/4 NONE    QAM16   2k 1/32 NONE
T 706000000 8MHz  3/4 NONE    QAM16   2k 1/32 NONE
T 650166670 8MHz  2/3 NONE    QAM64   2k 1/32 NONE
T 626166670 8MHz  2/3 NONE    QAM64   2k 1/32 NONE
T 658166670 8MHz  3/4 NONE    QAM16   2k 1/32 NONE
T 634166670 8MHz  3/4 NONE    QAM16   2k 1/32 NONE
T 730167000 8MHz  2/3 NONE    QAM64   8k 1/32 NONE      # Oxon & Bucks
T 746000000 8MHz  3/4 NONE    QAM64   8k 1/32 NONE      # Oxon & Bucks
T 785833000 8MHz  2/3 NONE    QAM64   8k 1/32 NONE      # Oxon & Bucks
T 802000000 8MHz  2/3 NONE    QAM64   8k 1/32 NONE      # Oxon & Bucks
T 777833000 8MHz  3/4 NONE    QAM64   8k 1/32 NONE      # Oxon & Bucks

More when I work out which these bad lines are…

Thanks for that feedback which may well be helpful to others having to deal with this.