Information about regional channels in Freesat BAT

[Tony Houghton]

I'm writing some DVB software which I want to make as easy as possible
to set up for Freesat and Freeview. I'm having some trouble with
automatically allocating LCNs (channel numbers) appropriate for the
user's configured region.

What I do know (mostly by browsing the source of something called
deadpenguin) is that in Freesat's BAT on pid 3841 there are descriptors
with a tag of D4 which contain a mapping of region codes to region
names, and in D3 descriptors there's a mapping between service_id,
region and LCN. A region code of 65535/-1 means it applies to any
region.

The trouble is with BBC 1, BBC 2 and Channel 4 (and S4C). The BBC 2
channels all have a region code of 0 and the only way I can find to
allocate them is matching the country in the region name to the service
name suffix. Similarly for BBC 1, but only if the region is London
(region code 1) or not England - other English regions use the expected
codes.

Then there's Channel 4. Again, I think it's OK for most English regions
while London and non-England are a problem, compounded by them all
having the same service name. However, I have spotted a pattern. If I
look up the service with LCN 104 and region 65535 and no other region,
the only two results are what I'm told is the correct one for London
(service_id 9211) and S4C, easily distinguished by their service names.

Doing all this seems a bit of a bodge, is there some other information
in the SI that I've missed?

--
TH * http://www.realh.co.uk

[Brian Gaff]

That number is interesting as its the biggest number able to be held in two
bytes.

Brian

--
--
From the sofa of Brian Gaff -
bri...@blueyonder.co.uk
Blind user, so no pictures please!
"Tony Houghton" <h...@realh.co.uk> wrote in message
news:slrnk5jj...@realh.co.uk...

[Tony Houghton]

In <k3cjlk$lv0$1...@news.albasani.net>,

Brian Gaff <bri...@blueyonder.co.uk> wrote:

> That number is interesting as its the biggest number able to be held in two
> bytes.

Yes, it's 65535 or -1 depending on whether you treat it as unsigned or
signed. That would be why they chose it to mean all regions. A value of
zero obviously has some significance to the BBC, but I'm not sure what.

[Graham.]

We could have done with some intelegent development at the start with
Freeview, then we would have something more elegent than the wanted
channels assigned LCNs in the 800s, just because there were some
weaker out of area ones detected earlier in the scan.

--
Graham.
%Profound_observation%

[ne...@address.invalid]

On Wed, 19 Sep 2012 15:47:18 +0100, Martin <m...@address.invalid> wrote:

>It's the amount of real savings they made by moving to Salford.

NO, that's the -1

[PeterC]

On Thu, 20 Sep 2012 13:23:36 +0100, Martin wrote:

> or even -32767

x10^3
--
Peter.
The gods will stay away
whilst religions hold sway

[Johny B Good]

On Thu, 20 Sep 2012 13:23:36 +0100, Martin <m...@address.invalid> wrote:

>or even -32767

Actually, I'm pretty certain that should be -32768

With zero being one of the 65536 possibilities that leaves a maximum
value of +32767 for the opposite end of the Two's Complement signed
integer numbering range.
--
Regards, J B Good

[Johny B Good]

BTW, I'm not disputing that 65535 absolute can be interpreted to mean
-1 in a 16 bit Two's Complent signed integer system, just that the
maximum negative value is not 32767 but 32768.

[Max Demian]

"Martin" <m...@address.invalid> wrote in message
news:hktq58lkbp34khohh...@4ax.com...

> On Sat, 22 Sep 2012 02:48:30 +0100, Johny B Good

> <johnny...@invalid.ntlworld.com> wrote:
>>On Thu, 20 Sep 2012 13:23:36 +0100, Martin <m...@address.invalid> wrote:

>>>or even -32767
>>
>> Actually, I'm pretty certain that should be -32768

Reminds me of the story of the computer-controlled airliner that flipped
itself upside down when it reached 32768 feet as it thought it was beneath
the Earth.

--
Max Demian

[John Legon]

The information you need will presumably be found in the D-Book of the
DTG, which is available to members only. The following draft document
seems to be based partly on the D-Book, however, and although there are
obvious differences when compared to Freeview in the UK, it might just
provide the clue you need:

http://freeview.reactivework.co.nz/images/uploads/pdfs/DTT_v2.0_tranmission_rules_.pdf

There's no BAT, and regional variations are apparently indicated by the
service availability descriptors in the SDT.

[Tony Houghton]

In <m8OdnQndf8S6ZcLN...@brightview.co.uk>,

John Legon <jo...@nospam.demon.co.uk> wrote:

> The information you need will presumably be found in the D-Book of the
> DTG, which is available to members only. The following draft document
> seems to be based partly on the D-Book, however, and although there are
> obvious differences when compared to Freeview in the UK, it might just
> provide the clue you need:
>
> http://freeview.reactivework.co.nz/images/uploads/pdfs/DTT_v2.0_tranmission_rules_.pdf
>
> There's no BAT, and regional variations are apparently indicated by the
> service availability descriptors in the SDT.

There's a lot in there that could be useful to me, so thanks. But my
original post was about Freesat, which has several differences,
especially in handling multiple regions.

[John Legon]

Tony Houghton wrote:
> In <m8OdnQndf8S6ZcLN...@brightview.co.uk>,
> John Legon <jo...@nospam.demon.co.uk> wrote:
>
>> The information you need will presumably be found in the D-Book of the
>> DTG, which is available to members only. The following draft document
>> seems to be based partly on the D-Book, however, and although there are
>> obvious differences when compared to Freeview in the UK, it might just
>> provide the clue you need:
>>
>> http://freeview.reactivework.co.nz/images/uploads/pdfs/DTT_v2.0_tranmission_rules_.pdf
>>
>> There's no BAT, and regional variations are apparently indicated by the
>> service availability descriptors in the SDT.
>
> There's a lot in there that could be useful to me, so thanks. But my
> original post was about Freesat,

So it was, though you did mention Freeview at the beginning. :)

> which has several differences,
> especially in handling multiple regions.

I wouldn't be surprised if the info you're looking for is carried in one
of the data streams from the Freesat transponder at 28.5 E, together
with the EPG. With Freeview, of course, the EPG is unencrypted and is
in the mux with the programme streams.

[Tony Houghton]

In <NcednSdZJpJ9QP3N...@brightview.co.uk>,

John Legon <jo...@nospam.demon.co.uk> wrote:

> Tony Houghton wrote:
>> In <m8OdnQndf8S6ZcLN...@brightview.co.uk>,
>> John Legon <jo...@nospam.demon.co.uk> wrote:
>>
>>> The information you need will presumably be found in the D-Book of the
>>> DTG, which is available to members only. The following draft document
>>> seems to be based partly on the D-Book, however, and although there are
>>> obvious differences when compared to Freeview in the UK, it might just
>>> provide the clue you need:
>>>
>>> http://freeview.reactivework.co.nz/images/uploads/pdfs/DTT_v2.0_tranmission_rules_.pdf
>>>
>>> There's no BAT, and regional variations are apparently indicated by the
>>> service availability descriptors in the SDT.
>>
>> There's a lot in there that could be useful to me, so thanks. But my
>> original post was about Freesat,
>
> So it was, though you did mention Freeview at the beginning. :)

Oh yes, reading it back I see I didn't make it clear that the problem
with regions was specific to Freesat.

> > which has several differences,
>> especially in handling multiple regions.
>
> I wouldn't be surprised if the info you're looking for is carried in one
> of the data streams from the Freesat transponder at 28.5 E, together
> with the EPG. With Freeview, of course, the EPG is unencrypted and is
> in the mux with the programme streams.

Possibly. I heard that this special Freesat transponder transmits the
EPG so densely that you can get a whole week's worth in something like
30 seconds instead of the similar number of minutes it takes on the
other transponders. Someone told me the PID(s) that's on are different
again from the already non-standard PIDs Freesat uses on other
transponders. Unfortunately I forgot to make a note of these special EPG
PIDs, but I do know the others and already have code for decoding the
EPG. I can probably find the PIDs with dvbsnoop. I presume the
transponder is the one that carries pseudo-channels with freesat in the
name, like "FreesatHome" and "FREESAT EPG" and is (currently) on
11427830 horizontal with ts_id 2315?

[John Legon]

Tony Houghton wrote:
> In <NcednSdZJpJ9QP3N...@brightview.co.uk>,
> John Legon <jo...@nospam.demon.co.uk> wrote:

>> I wouldn't be surprised if the info you're looking for is carried in one
>> of the data streams from the Freesat transponder at 28.5 E, together
>> with the EPG. With Freeview, of course, the EPG is unencrypted and is
>> in the mux with the programme streams.
>
> Possibly. I heard that this special Freesat transponder transmits the
> EPG so densely that you can get a whole week's worth in something like
> 30 seconds instead of the similar number of minutes it takes on the
> other transponders.

Yes, the EPG is compressed, and also encrypted in the sense that the
compression uses non-standard Huffman tables. I find that recording the
streams for about one minute gives me the full EPG.

> Someone told me the PID(s) that's on are different
> again from the already non-standard PIDs Freesat uses on other
> transponders. Unfortunately I forgot to make a note of these special EPG
> PIDs, but I do know the others and already have code for decoding the
> EPG. I can probably find the PIDs with dvbsnoop. I presume the
> transponder is the one that carries pseudo-channels with freesat in the
> name, like "FreesatHome" and "FREESAT EPG" and is (currently) on
> 11427830 horizontal with ts_id 2315?

I don't know about the ts_id and don't need it for my FTA satellite box,
but the transponder is correct with symbol rate '27500'. The EPG PIDs
are 3002 (services) and 3003 (events).