DS Forums

 
 

COM multiplex mode change at switched transmitters


Reply
Thread Tools Search this Thread
Old 12-10-2011, 14:25
kruador
Forum Member
 
Join Date: Feb 2011
Location: Reading
Posts: 449

Digital UK's postcode checker has been updated to show a 'COM mux FEC change' event at many transmitters, occurring in the next couple of months. Not all multiplexes are changing at all transmitters. The new mode is shown as 64QAM 3/4 8K - requiring greater C/N ratio for reliable reception.

I'm speculating that the reason for this is to change the guard interval to a longer duration, to make wide-area SFNs possible. If some transmitters don't change, it can't be to increase capacity - in which case the most likely guard interval is 1/8, quadrupling the distance between members of the SFN. The new capacity would be 24.88Mbps compared to the current mode capacity of 24.13Mbps.

At many sites, the COM muxes will be restricted more by co-channel interference of uncoordinated broadcasts than by SFN operation.
kruador is offline   Reply With Quote
Please sign in or register to remove this advertisement.
Old 12-10-2011, 14:54
Ray Cathode
Forum Member
 
Join Date: Jun 2005
Location: Guildford / Crystal Palace
Posts: 13,114
Good find well done. I have an alternative explanation. This mode (8k, 64QAM, 3/4FEC, 1/32GI) has been previously tested at Preseli and Caldbeck with the stated object of increasing capacity. With the change in FEC the capacity of the transport stream increases from 24128kbps to 27144kbps, by the equivalent of two shopping channels. I heard no complaints at all during the tests, which might have been worse interference due to the reduction in error correction. Assuming one channel sells for £10 million per annum, that is a good incentive to increase capacity. Why it has been kept so secret is a surprise considering the trials were widely advertised on the Freeview website for example. At the very least this should not be brought in without an Ofcon consultation and they have been silent as usual. But this could help the BBC if it spreads to PSB muxes.

If we take Mux C as an example with 9 SD MPEG-2 streams in DVB-T, that could rise to 12 SD streams at DSO and 14 SD streams with this FEC change. DVB-T2 would generate even more, perhaps 40% more with MPEG-4.

I take the point that this is not being extended to all transmitters, so it could be a full scale trial before being rolled out everywhere.
Ray Cathode is offline Follow this poster on Twitter   Reply With Quote
Old 12-10-2011, 15:30
Ray Cathode
Forum Member
 
Join Date: Jun 2005
Location: Guildford / Crystal Palace
Posts: 13,114
It appears to be a trial on Arq B only at most DSO'd transmitters where full power has been achieved.

But there are exceptions and Oxford is already running with the new parameters on both Arq A & Arq B according to DUK.
Ray Cathode is offline Follow this poster on Twitter   Reply With Quote
Old 12-10-2011, 16:16
reslfj
Forum Member
 
Join Date: Jun 2007
Posts: 1,820
It appears to be a trial on Arq B only at most DSO'd transmitters where full power has been achieved.
The capacity and robustness can be found in this table from the DVB-T standard
The UK post DSO code rate is 2/3 and this new COM code rate is 3/4.

While the required C/N in a Ricean channel (LOS) is increased by 1.6 dB (17.3 -> 18.9 dB) , the C/N must be 2.7 dB higher (20.3 -> 23 dB) for a Rayleigh channel (indirect, reflections, indoor) - almost twice the power.
The COM muxes already operates at half the ERP of the PSB muxes from many masts.

This will reduce the coverage - not least in urban areas, in lofts and with indoor aerials, where 'line of sight' is not possible.

In Denmark the pay-TV muxes operated by Boxer/Teracom uses the code rate 3/4 , but the Danmarks Radio muxes uses the code rate 2/3. The Boxer license says 97% coverage - while DR try to get ~ 99%.

The Boxer muxes broadcasts about 10 MPEG-4 SD channels - in these 22.39 Mbps.

In Sweden even the MUX1 - Sveriges Television public service - operates with GI=1/4 and CR=3/4 in some SFN areas, while in other areas where GI=1/8 is possible a FEC CR of 2/3 is used. (both have a bit rate just over 22 Mbps)

Lars
reslfj is offline   Reply With Quote
Old 12-10-2011, 17:28
Mike_1101
Forum Member
 
Join Date: Jan 2004
Location: Lancs
Posts: 7,928
Good find well done. I have an alternative explanation. This mode (8k, 64QAM, 3/4FEC, 1/32GI) has been previously tested at Preseli and Caldbeck with the stated object of increasing capacity. With the change in FEC the capacity of the transport stream increases from 24128kbps to 27144kbps, by the equivalent of two shopping channels. I heard no complaints at all during the tests, which might have been worse interference due to the reduction in error correction. Assuming one channel sells for £10 million per annum, that is a good incentive to increase capacity. Why it has been kept so secret is a surprise considering the trials were widely advertised on the Freeview website for example. At the very least this should not be brought in without an Ofcon consultation and they have been silent as usual. But this could help the BBC if it spreads to PSB muxes.

If we take Mux C as an example with 9 SD MPEG-2 streams in DVB-T, that could rise to 12 SD streams at DSO and 14 SD streams with this FEC change. DVB-T2 would generate even more, perhaps 40% more with MPEG-4.

I take the point that this is not being extended to all transmitters, so it could be a full scale trial before being rolled out everywhere.
How much more of this garbage do we need - does anybody watch them?
Mike_1101 is offline   Reply With Quote
Old 12-10-2011, 17:48
Ray Cathode
Forum Member
 
Join Date: Jun 2005
Location: Guildford / Crystal Palace
Posts: 13,114
It looks as though the DUK database was being altered today while we were watching it and Ofcon has apparently altered the broadcast technical code without any public notice. It seems that all three COM muxes after DSO will be operating in the new mode 64QAM 3/4FEC.

This is at odds with the current code http://stakeholders.ofcom.org.uk/bin...ce/dttt_uk.pdf It's amazing how Ofcon can get away with this. Surely the public should be allowed to comment on changes that allow commercial operators to get another 3Mbps at the expense of some viewers with marginable reception. Maybe DUK are wrong.
Ray Cathode is offline Follow this poster on Twitter   Reply With Quote
Old 12-10-2011, 18:12
Paul237
Forum Member
 
Join Date: Nov 2010
Location: Midlands, UK
Posts: 4,964
Is this why the Wrekin transmitter is on the Digital UK site as having a "final step" on 22 November 2011?

It says no retune is required.
Paul237 is offline   Reply With Quote
Old 12-10-2011, 18:39
Ray Cathode
Forum Member
 
Join Date: Jun 2005
Location: Guildford / Crystal Palace
Posts: 13,114
Is this why the Wrekin transmitter is on the Digital UK site as having a "final step" on 22 November 2011?

It says no retune is required.
It says COM MUX FEC Change on 22/11/11 and applies to Arq B only. These have been updated on other transmitters throughout the day to all 3 COM muxes.

Waltham is getting Arq B changed on 22/11/11

Sutton C is getting Arq B changed on 1/11/11.
Ray Cathode is offline Follow this poster on Twitter   Reply With Quote
Old 12-10-2011, 18:40
kruador
Forum Member
 
Join Date: Feb 2011
Location: Reading
Posts: 449
Running through the transmitters that carry the COM multiplexes:

Caldbeck: SDN and ArqA in 2/3, ArqB already in 3/4
Selkirk: All COMs change to 3/4 on 23 November 2011

Winter Hill: ArqB on 3 November 2011. SDN and ArqA untouched.
Lancaster: ArqB on 8 November 2011.
Pendle Forest: ArqB on 25 October 2011.
Saddleworth: ArqB on 27 October 2011.
Storeton: ArqB on 31 October.

Carmel: ArqB already implemented
Kilvey Hill: ArqB ditto
Llanddona: ditto
Moel-Y-Parc: ditto
Preseli: ditto
Aberdare: ditto
Blaenplwyf: ditto
Pontypool: ditto
Wenvoe: ditto

Beacon Hill: 27/10/11, all COMs
Caradon Hill: 2/11/11, ArqA and ArqB
Huntshaw Cross: 1/11/11, all COMs
Plympton: 10/11/11, all COMs
Redruth: 3/11/11, all COMs
Stockland Hill: 25/10/11, all COMs

I'll do West, STV North, STV Central, Anglia, Central, Yorkshire and the regions yet to switch later. Postcodes from ukfree.tv transmitter pages.

Doing it at Oxford is bloody stupid when that transmitter's COM muxes are at one-quarter power until after Hemel Hempstead switches over! The final power level for the COM muxes is half that of the PSBs.
kruador is offline   Reply With Quote
Old 12-10-2011, 18:57
Paul237
Forum Member
 
Join Date: Nov 2010
Location: Midlands, UK
Posts: 4,964
It says COM MUX FEC Change on 22/11/11 and applies to Arq B only. These have been updated on other transmitters throughout the day to all 3 COM muxes.

Waltham is getting Arq B changed on 22/11/11

Sutton C is getting Arq B changed on 1/11/11.
Interesting. My downstairs TV is Sutton and my upstairs TV is Wrekin. I can get all 6 multiplexes on both, but COM4 seems to be the weakest from the Wrekin (at least the COM4 channels have broken up once or twice since the final power up on 28 Sept).
Paul237 is offline   Reply With Quote
Old 12-10-2011, 18:59
lnp
Forum Member
 
Join Date: Jul 2000
Posts: 12,229
The change from Belmont on 18th November is made before the ARQ A & B power up so there will be no way of comparing the before/after results.

As for Ofcon, another change that furthers the interests of big business over consumers. No surprise there then! Where will this end, how many more times will the error correction be lowered just so Arqiva can fit more crap in.
lnp is offline   Reply With Quote
Old 12-10-2011, 19:02
chrisy
Forum Member
 
Join Date: May 2002
Location: Beds (Sandy Heath TX)
Posts: 8,852
Sandy says 28th October for this change - on Arq B only. Arq A isn't changing (according to the db), despite not being live until 23rd November, and neither is SDN.
chrisy is offline Follow this poster on Twitter   Reply With Quote
Old 12-10-2011, 19:04
chrisy
Forum Member
 
Join Date: May 2002
Location: Beds (Sandy Heath TX)
Posts: 8,852
Waltham is getting Arq B changed on 22/11/11
Says 25th October on my postcode!
chrisy is offline Follow this poster on Twitter   Reply With Quote
Old 12-10-2011, 19:07
Paul237
Forum Member
 
Join Date: Nov 2010
Location: Midlands, UK
Posts: 4,964
The change from Belmont on 18th November is made before the ARQ A & B power up so there will be no way of comparing the before/after results.

As for Ofcon, another change that furthers the interests of big business over consumers. No surprise there then! Where will this end, how many more times will the error correction be lowered just so Arqiva can fit more crap in.
Agreed. More shopping channels aren't needed or wanted. I've never watched one for more than 30 seconds, usually with a raised eyebrow throughout.

I wouldn't mind so much, but there are a few channels that I like watching on that multiplex. It'd be a shame to lose them all, just because they want to add a few more nonsense channels.
Paul237 is offline   Reply With Quote
Old 12-10-2011, 19:26
Ray Cathode
Forum Member
 
Join Date: Jun 2005
Location: Guildford / Crystal Palace
Posts: 13,114
The data is still changing as it has all day. At present Crystal Palace, Guildford & Reigate plus Midhurst & Hannington are down with all 3 COM muxes after DSO with FEC3/4.
Ray Cathode is offline Follow this poster on Twitter   Reply With Quote
Old 12-10-2011, 20:40
a516
Forum Member
 
Join Date: Dec 2008
Location: East Midlands
Posts: 5,204
Slightly reserved by these developments:

Had this been done at DSO stage 2 when power was boosted, many might not have noticed. People in fringe zones who didn't get the commercial multiplexes prior to switchover* would probably have not been able to get a stable reception due to this change after switchover.

Now that this is being done now, people in fringe zones who didn't get the commercial multiplexes prior to switchover may have gained reception due to the boosted signal at switchover, but the subsequent FEC change could just tip them over the cliff resulting in a loss of signal. I just hope once Freesat has more UK spot beam, more Freeview FTA broadcasters join Freesat to give people an alternative in such situations.

Admittedly, it won't be loads of households, but there will be official calculations estimating how many might be affected. But individual households just become numbers to the commercial operators...

If you're in a good signal zone, no worries though.

I suppose a change in the FEC at Waltham Arq B would make reception more difficult in areas where Bolehill interferes. As I have previously reported on this forum, Bolehill can either cancel out Arqiva B or affect the signal quality. At one relative's house, the signal quality of Arqiva B is currently just 34%, compared to 90-100% on the other full-powered multiplexes from Waltham.
a516 is offline   Reply With Quote
Old 12-10-2011, 21:22
SimonBlackham
Forum Member
 
Join Date: Oct 2004
Location: NewForest (Free-View/Sat+19E)
Posts: 7,719
For us all COM MUXes shown on the postcode checker* show a migration to 3/4 by 2013 - except for Mendip that only shows this for ARQ-B on 8th Nov 2011 and the other Mendip COM MUXes still on 2/3 @ 2013 !!

Mistake??



*Mendip, Rowridge HP, VP & Stockland Hill.
SimonBlackham is offline   Reply With Quote
Old 12-10-2011, 22:01
jcsager
Forum Member
 
Join Date: Dec 2008
Posts: 144
The data is still changing as it has all day. At present Crystal Palace, Guildford & Reigate plus Midhurst & Hannington are down with all 3 COM muxes after DSO with FEC3/4.
Postcode checker shows Sudbury goes to FEC 3/4 for ArqA and ArqB on 22nd Nov 2011. SDN appears to stay at 2/3 even after powerup next year, as do all the PSB muxes.
jcsager is offline   Reply With Quote
Old 12-10-2011, 23:51
plane spotter
Inactive Member
 
Join Date: Apr 2010
Posts: 884
The capacity and robustness can be found in this table from the DVB-T standard
The UK post DSO code rate is 2/3 and this new COM code rate is 3/4.

While the required C/N in a Ricean channel (LOS) is increased by 1.6 dB (17.3 -> 18.9 dB) , the C/N must be 2.7 dB higher (20.3 -> 23 dB) for a Rayleigh channel (indirect, reflections, indoor) - almost twice the power.
The COM muxes already operates at half the ERP of the PSB muxes from many masts.

This will reduce the coverage - not least in urban areas, in lofts and with indoor aerials, where 'line of sight' is not possible.

In Denmark the pay-TV muxes operated by Boxer/Teracom uses the code rate 3/4 , but the Danmarks Radio muxes uses the code rate 2/3. The Boxer license says 97% coverage - while DR try to get ~ 99%.

The Boxer muxes broadcasts about 10 MPEG-4 SD channels - in these 22.39 Mbps.

In Sweden even the MUX1 - Sveriges Television public service - operates with GI=1/4 and CR=3/4 in some SFN areas, while in other areas where GI=1/8 is possible a FEC CR of 2/3 is used. (both have a bit rate just over 22 Mbps)

Lars
You will be utterly amazed that I agree with your conclusions on this.

To sacrifice s/n this way is inevitably going to cost the public ,those near the cliff edge a load of money to upgrade their antenna’s
For what, to add a couple of crap shopping channels that a fraction of 1% of the viewers will watch now an again.
What next extra gambling channels and a further deterioration in code rate ?

With all the re tunes ,this is what drives viewers to Sky ,as
their commercials stating they do it better really sinks home with these tactics .

What should have been done is the code rate and other parameters affecting how robust the signal is be set from the start unless it is found it needs to be changed to improve coverage ,due to coverage issues.

Also as you raised the subject ,what is the spin for the reasons why the coms muxes are operating at lower power than the PSB muxes?
plane spotter is offline   Reply With Quote
Old 13-10-2011, 00:06
Muzer
Forum Member
 
Join Date: Feb 2004
Location: Andover, Hampshire, UK
Posts: 3,624
Am I the only one excited by the prospect of six new channels, on top of whatever else is coming? Surely at least one of them might be something worthwhile for some of us.
Muzer is offline   Reply With Quote
Old 13-10-2011, 07:06
xtaz
Forum Member
 
Join Date: Mar 2004
Location: Bristol
Posts: 269
Am I the only one excited by the prospect of six new channels, on top of whatever else is coming? Surely at least one of them might be something worthwhile for some of us.
If anything decent came of it then I'd be excited too. Unfortunately I suspect these new channels will end up just being some more shopping channels or something similar. I could happily not bother tuning in the 3 COM MUX's and not really miss much. Only E4 and Film4 are the ones I care about. Shame really. If anything this extra bandwidth should be used to restore the resolution of the channels back to what it was last year (720x576). Last month or the month before they dropped them all down to the bare minimum 544x520 and now it looks like YouTube vision. Especially when you have a large TV. Mine's 46" and it looks hideous.
xtaz is offline   Reply With Quote
Old 13-10-2011, 07:56
plane spotter
Inactive Member
 
Join Date: Apr 2010
Posts: 884
If anything decent came of it then I'd be excited too. Unfortunately I suspect these new channels will end up just being some more shopping channels or something similar. I could happily not bother tuning in the 3 COM MUX's and not really miss much. Only E4 and Film4 are the ones I care about. Shame really. If anything this extra bandwidth should be used to restore the resolution of the channels back to what it was last year (720x576). Last month or the month before they dropped them all down to the bare minimum 544x520 and now it looks like YouTube vision. Especially when you have a large TV. Mine's 46" and it looks hideous.
You have summed it all up very well.

Sort of shows the greed of the mux operator .
plane spotter is offline   Reply With Quote
Old 13-10-2011, 15:46
kruador
Forum Member
 
Join Date: Feb 2011
Location: Reading
Posts: 449
Also as you raised the subject ,what is the spin for the reasons why the coms muxes are operating at lower power than the PSB muxes?
If you read the BBC presentation on switchover from 2000, page 9 mentions that if a conversion ratio (i.e. reduction of digital power compared to analogue) of 10 dB is used, "the digital transmissions would match the coverage of the analogue transmissions almost exactly". This refers to Crystal Palace's coverage area. The question is not why are the COM muxes operating at lower power, it's why are the PSBs operating at higher power, from some sites.

The Chester '97 agreement that allowed DTT in the first place allowed a conversion ratio of 7 dB for existing analogue allocations, though we never used this before it was essentially superseded by GE06.

Sites with equal PSB and COM power, -7 dB from analogue unless noted:

Sandy Heath (differs by only 0.25 dB, GE06 permits 200 kW for both, PSBs 0.46 dB below GE06 limit)
Sudbury (-4 dB compared to analogue)
Tacolneston (-4 dB compared to analogue)

Bromsgrove (-8.45 dB)
Lark Stoke
Nottingham
Sutton Coldfield
Brierley Hill
Malvern

Winter Hill
Lancaster
Pendle Forest
Saddleworth
Storeton

Crystal Palace
Guildford
Hemel Hempstead
Reigate

Bluebell Hill (-1.76 dB)
Heathfield
Rowridge (on VP) (-4 dB)
Tunbridge Wells (-4 dB)
Whitehawk Hill (-4 dB)

Black Hill
Rosneath (HP & VP)

Fenham

Kilvey Hill
Aberdare
Pontypool

Mendip
Bristol Kings Weston

Plympton

Chesterfield (-7 dB except PSB3 -4 dB compared to analogue, co-channel Waltham SDN)
Emley Moor
Idle
Keighley
Sheffield

Sites with PSBs +3dB compared to COM, PSBs -7 dB from analogue unless noted:

Caldbeck
Selkirk

Oxford
Ridge Hill
Fenton
The Wrekin
Waltham

Dover (PSBs -1 dB, COMs -4 dB)
Hannington
Hastings (PSBs 0 dB, COMs -3 dB)
Midhurst

Craigkelly
Darvel

Angus
Durris
Eitshal
Keelylang Hill
Knockmore
Rosemarkie
Rumster Forest

Bilsdale
Chatton
Pontop Pike

Divis
Limavady

Carmel
Llanddona
Moel-Y-Parc
Preseli
Wenvoe

Bristol Ilchester Crescent (PSBs -4 dB, COMs -7 dB)

Beacon Hill
Caradon Hill
Huntshaw Cross
Redruth
Stockland Hill

Oliver's Mount (PSB +3 dB, COM 0 dB)

Weirdos: (anything where there isn't a standard ratio between all PSBs and all COMs)

Rowridge HP: PSB conversion ratio -4 dB, COM -10 dB
Torosay: PSBs + SDN -7 dB, ArqA/B -10 dB
Bressay: PSBs + SDN -7 dB, ArqA/B -10 dB

Brougher Mountain: PSBs -7 dB, COMs -17 dB
Blaen-Plwyf: PSBs -4 dB, COMs -10 dB

Belmont: PSB -5.2 dB, SDN -10 dB, ArqA/B -7 dB (SDN is in-group, ArqA and B are out-of-group)
kruador is offline   Reply With Quote
Old 14-10-2011, 15:34
Ray Cathode
Forum Member
 
Join Date: Jun 2005
Location: Guildford / Crystal Palace
Posts: 13,114
It's still not clear what's going on as the DUK database is varying across the country. I presume they'll finish updating it next week. It certainly applies to Arq B, but not necessarily to the other muxes. The mux licencees wouldn't like the idea, but it would be a good way to get local tv capacity from each transmitter at high power. plane spotter?
Ray Cathode is offline Follow this poster on Twitter   Reply With Quote
Old 14-10-2011, 15:45
John259
Forum Member
 
Join Date: Nov 2003
Location: Norwich, Norfolk, UK
Posts: 14,282
Please excuse what is probably a silly question, but is all receiving equipment compatible with these new modes?
John259 is offline   Reply With Quote
 
Reply




 
Forum Jump


All times are GMT. The time now is 00:18.