All posts by Mike Dimmick
Below are all of Mike Dimmick's postings, with the most recent are at the bottom of the page.
jb38, Keith Hook: The Sudbury transmitter transmits the Anglia region, and the probability of reliable reception is VERY HIGH - not 'not possible'. I assume you're referring to the probabilities for Sandy Heath, also part of the Anglia region.
Keith, are you missing any other channels or just those on Mux 2/D3&4?
| link to this comment |
SJ: 'Blurring' can be a sign that the receiver is on the verge of being unable to correct all the errors in the signal. Since it's just started happening, my guess is that this is due to too much signal, which is causing distortion. If you have a booster, you should remove it now, or turn it down if you lose channels when you remove it. If not, consider adding an attenuator.
| link to this comment |
Alan Matthews: When all is complete, on 27 June 2012 (the date of Dover's second switchover step), the power increases will be:
SDN by 5.5 dB (3.57x) to 50kW
ArqA by 12.2 dB (16.7x) to 50kW
ArqB by 13.6 dB (22.7x) to 50kW
These are all substantially greater increases than they have already had.
| link to this comment |
Briantist: No, it is not for the London 1 and London 2 multiplexes, unless Arqiva have supplied the wrong map. The website Ash linked to is specifically for the London 3 multiplex, which is licensed to a different operator from the other two (each is a different operator).
The linked map does appear to be the correct map, because it shows the Brookmans Park transmitter near Potters Bar. Neither London 1 nor London 2 use this transmitter.
I think you need to go back and figure out why your maps don't correspond to Arqiva's published coverage maps.
| link to this comment |
John Fisher:
Signal *strength* is the raw level of power coming in on the selected UHF channel, from *all* sources - the transmitter you want, other transmitters using the channel, signal distortion and noise.
Signal *quality* is a measure of how much interference or distortion there is. Any of either causes the box to read the wrong value for the carrier. The signal is encoded in a way that allows the box to determine the most likely value, but this decision can be wrong if the levels of interference, noise, or distortion are too high. To combat this, the signal carries redundant information to help guide the decision and correct the errors (called 'forward error correction', FEC). Signal quality is usually an indicator of how many errors were corrected.
Unfortunately on most boxes, both measures are useless for diagnostics. The 'strength' is usually a percentage relative to the maximum possible input from one multiplex alone with no interference, but as this differs from box to box, 50% (say) on one box may mean something different from 50% on another box. As a general rule you should stay well below 100%, as the combined power from all multiplexes (and remaining analogue channels) from all receivable transmitters would exceed the power on that one single test multiplex. The upper limit recommended by the Confederation of Aerial Installers for one multiplex (before switchover), which allows for a combination of five high-power analogue and six low-power digital signals, falls at about 50% on my box (Humax PVR-9200T). Some boxes apparently under-read if the signals are actually over the maximum they can handle.
Signal quality on many boxes (again, including mine) is simply 100% minus the relative number of uncorrectable errors over the last few seconds. It doesn't tell you how many errors were corrected, which means you can't find out how close to the point of break-up you are, and you can tell visually and audibly if it isn't capable of correcting all the errors, as it usually starts breaking up or giving audible 'pop' sounds at this point.
If the number of uncorrectable errors is low, you can get a more blocky or blurry picture without actual tearing (where a whole line is discarded). This is right at the top of the so-called 'cliff edge'. Otherwise, if the signal quality is good enough - sufficient signal, low enough noise and interference - all errors can be corrected and the decoding is perfect. Digital can tolerate up to about 20dB, or 100 times, more interference/noise than would start to cause problems on analogue, though the problems would be very minor at this point. However, you could still get an understandable picture and sound on analogue with noise levels greater than will cause complete failure on digital.
Note that you can still get a blocky or blurry picture with perfect reception, if the broadcaster is compressing the picture too much. We can only cram nine channels into the space that used to carry one by throwing a lot of information away. It's designed to throw away the least-noticeable information, and only transmit the changes between pictures, but if too much information is discarded you can see the joins between coding blocks, distinct colours where it should be smooth, or blurry detail that should be sharp. There's nothing at all you can do about this - the information was never transmitted, it wasn't lost in transit.
| link to this comment |
Ed: The analogue television service was planned carefully to allow four complete networks to be transmitted from every site in the country. (There are a very few one- or two-channel fill-in relays where a region has later been split, to allow an area to get its 'correct' local news.) The original digital signals were crammed into this plan, at whatever level would cause minimum interference to analogue services, and from very few sites: the 'main' transmitters (fed directly from studios) and a few relay sites that served a large chunk of the population, or parts that the advertisers considered important. There simply wasn't room for any more power increases.
The frequency plan wasn't just for this country, but also to allow for the transmissions from Ireland, France, the Netherlands and Belgium (being most affected by our transmissions).
There were various changes over the years since digital services started in 1998, particularly the equalization programme of 2000, bringing the coverage of each multiplex from a particular site to a similar level (originally the plans might have huge coverage on Mux 1 but very low coverage on Mux D), the mode change of late 2002 (effectively increasing coverage with no power increase, on the four multiplexes that changed mode), and power doubling of many multiplexes at many sites in 2003/4. These changes could be made as interference wasn't as bad as first thought, or few people were actually using theoretical coverage from one transmitter: some changes were required to analogue relays to prevent people from losing reception. There were diminishing returns on these changes, though, and analogue switch-off is really the only way to improve it further.
The switchover is to turn off the analogue signals, so that the digital signals (at least for the public-service broadcasters) can take over their frequencies and, largely, power levels. But it can't happen until analogue is switched off, because it will cause too much interference at other sites.
You get two weeks to sort out problems - BBC Two is switched off first, replaced by high-power digital BBC services. Then, two weeks later, the other four analogue channels are switched off and the remaining high-power digital services start. If you had digital before, you need to retune on both days, as the signals move to the much better frequencies vacated by the analogue transmissions.
| link to this comment |
Ben: Crystal Palace has NOT yet switched over. That happens next April. The Sudbury transmitter, which serves parts of Suffolk and Essex, has just switched - but some of the multiplexes are still on quite low power, because their new frequency allocations are still in use at Dover and/or Bluebell Hill, so they have had to stay on pre-switchover channels.
Digital UK's predictor does think that you'll get best reception of the PSB multiplexes - those run by the traditional analogue broadcasters, carrying their main channels - from Sudbury. The commercial multiplexes are expected to be problematic from there at the moment even with a roof-top aerial. This should improve as they move to clearer channels - you should retune on 27 June next year. The Arqiva B multiplex will move to UHF C63 on 16 November this year, but this change is not expected to improve your reception.
| link to this comment |
Jane Hore: Can you provide a full postcode please, so we can see the prediction for you?
Using Bing Maps to get a co-ordinate for Oulton Broad, I got a nearby postcode of NR33 8QH from nearby.org.uk. This shows very high probability of reliable reception, suggesting a problem with your system rather than any issue at the transmitter. Is analogue reception clear? If not, is it snowy or is there 'herringbone' patterning? If snowy or lacking colour, I'd suspect a broken aerial or cable - if herringboning, it could be too much signal and you should remove or turn down any boosters or amplifiers.
In answer to your final question, the transmitters are owned and operated by Arqiva. This is the successor company to NTL, which took over the IBA's transmitters in 1993; in 2007 Arqiva bought up the company that eventually ended up owning the BBC's transmitters after they were privatized. Tacolneston was originally a BBC transmitter, the IBA (for ITV and later C4) used the site as a tenant for colour TV. At other sites, e.g. Sandy Heath, the IBA was landlord and the BBC the tenant.
| link to this comment |
Barny Otoole: It's actually quite a tricky problem, because the island of Ireland uses a different grid reference system to that of Great Britain and the minor British Isles. It's based on a projection from a different point - a different way of displaying the curved surface of the Earth on a flat piece of paper. The maps are then displayed by Google Maps, which requires points in GPS co-ordinates, based on a third different projection and even a different model of the Earth's shape.
I suspect the part of Brian's software that gives the distance treats IJ163867 (which nearby.org.uk gives as the Irish Grid containing the centre of your postcode) as square IJ of the Ordnance Survey GB National Grid, which is actually just off the west coast of Norway. This would produce an error of 1,063km from the south-west corner of OSGB square NW to the south-west corner of 'IJ'. It should instead say I = Irish and use the OSNI transformation to GPS co-ordinates.
Digital UK's predictor - which I hope is based on the correct information, since it displays the OSNI logo - shows that you are 16km (10 miles) from Divis.
| link to this comment |
Wednesday 20 July 2011 10:38AM
carl: Did you have the BBC channels after the first stage of switchover? If not, your box may not be able to handle 8K mode. See whether it's on the list at http://www.digitaluk.co.uk/2kequipment . That's a list of known-bad equipment, there may be others that also don't work (or sometimes boxes are rebadged, the whole family of equipment doesn't work, but only one appears on the list).
If you hadn't retuned since mid-2008, you might also now be experiencing the 'Split NIT' problem for the first time. See http://www.digitaluk.co.uk/splitnit for equipment known to have a problem with this.
No changes have been made to Freesat, you should check the connection to the dish. That would point to a problem with the cabling, generally. Are you using a communal system, sharing an aerial and dish with your neighbours? If so, the communal system may need adjusting - check whether this is currently being done.