Ofcom reports on DAB with low cost transmission for small stations (Merged)
Phil Dodd
Posts: 3,975 Forum Member
✭✭✭
A somewhat obtuse article at the bottom of page 47 of The Times, Saturday 3 August, 2013. Unfortunately, because of the Times paywall, there isn't an internet link to those of us without a subscription, so I hope that the Times won't object to me quoting a bit of what they say.
The article, written by Nic Fildes, is headed “Salvation for Partridge despite the falcons”, and features our Al in thoughtful pose, wearing his 1970s headphones.
The gist of the article is that Rashid Mustapha, a sort of James Bond of radio engineering, was motivated to try and help small radio stations escape FM and go digital. Rashid is Senior Associate, Spectrum Policy Group at Ofcom, with plenty of hands-on engineering experience prior to getting there. The high cost is the problem ( as we all know ) , Nic Fildes explains in his Times article.
Rashid devised a cheap way of encoding information into a digital signal ( the evil codeword DAB isn't actually used in the article but the equally evil word “multiplex” is ), using a Raspberry Pi, which being currently at £31.20 including VAT would certainly cut the cost of encoding.
Nic Fildes stresses in his article that Rashid in particular and Ofcom in general are concerned that the cost of participation in digital is prohibitive for so many UK broadcasters. It will be great if something can be achieved to break the logjam to get more channels on board in a sustainable way, so that those who do make the move are still there a few months later. Good luck to them on that !
The falcons mentioned in Nic's story were occupying a tower that Rashid had to wait to use to conduct experiments with his Raspberry Pi encoded digital transmissions – but it worked in the end !
No doubt Rashid would give more info to anyone who is interested - meanwhile us listeners will wait in hope !
Re the Raspberry Pi, it always amuses me that books along the lines of “Raspberry Pi for dummies” in a certain well-known high-street newsagents cost as much as the actual computer itself...
The article, written by Nic Fildes, is headed “Salvation for Partridge despite the falcons”, and features our Al in thoughtful pose, wearing his 1970s headphones.
The gist of the article is that Rashid Mustapha, a sort of James Bond of radio engineering, was motivated to try and help small radio stations escape FM and go digital. Rashid is Senior Associate, Spectrum Policy Group at Ofcom, with plenty of hands-on engineering experience prior to getting there. The high cost is the problem ( as we all know ) , Nic Fildes explains in his Times article.
Rashid devised a cheap way of encoding information into a digital signal ( the evil codeword DAB isn't actually used in the article but the equally evil word “multiplex” is ), using a Raspberry Pi, which being currently at £31.20 including VAT would certainly cut the cost of encoding.
Nic Fildes stresses in his article that Rashid in particular and Ofcom in general are concerned that the cost of participation in digital is prohibitive for so many UK broadcasters. It will be great if something can be achieved to break the logjam to get more channels on board in a sustainable way, so that those who do make the move are still there a few months later. Good luck to them on that !
The falcons mentioned in Nic's story were occupying a tower that Rashid had to wait to use to conduct experiments with his Raspberry Pi encoded digital transmissions – but it worked in the end !
No doubt Rashid would give more info to anyone who is interested - meanwhile us listeners will wait in hope !
Re the Raspberry Pi, it always amuses me that books along the lines of “Raspberry Pi for dummies” in a certain well-known high-street newsagents cost as much as the actual computer itself...
0
Comments
The obvious issue is that DAB channels are about 1.7MHz wide (vs about 0.4 for FM) which means finding capacity for small scale stations will be tricky. Shame there wasn't a 1/4 and 1/2 width mode in the spec offering around 1/4 or 1/2 the capacity (i.e. 2-3 or 5-6 stations) in a 1/4 to 1/2 of the spectrum like there is on satellite.
e.g. in Nottingham you could see (for instance):-
12C - Local DAB
12D.1 - Erewash Valley and Hucknall - Trent Sound and Erewash Sound
12D.1 - Carlton Valley - Castle College radio (or whatever it is now)
12D.2-3 - City of Nottingham - Kemet / Dawn / Faza / URN / Fly FM
12D.4 - Vale of Belvoir - The Eye
13A - Regional DAB
Thanks for that detail Kev - a bit more reading to do from that website I can see - let's hope that it is what Rashid is trying out !
HA HA no doubt they'll try !
http://stakeholders.ofcom.org.uk/market-data-research/other/radio-research/Software-DAB-Research/
pdf here:
http://stakeholders.ofcom.org.uk/binaries/research/radio-research/Software-DAB-Research.pdf
Interesting fact about the ancient DAB standard.The final patents associated with DAB (Eureka 147) standard expired in January 2013 (in most of the world) making it the first free "open" digital radio broadcasting standard.
http://stakeholders.ofcom.org.uk/market-data-research/other/radio-research/Software-DAB-Research/?utm_source=updates&utm_medium=email&utm_campaign=lowcostDABresearch
http://stakeholders.ofcom.org.uk/binaries/research/radio-research/Software-DAB-Research.pdf
So for approx £8,600 per year (vs £10,000 for FM) you could have a whole multiplex covering somewhere around the size of Brighton from a single 100w transmitter (if you have a small commercial, community, and student station in your town then that's down to just over £2,200 per station or ~£4,000 for a two transmitter service).
And as guessed it builds on the CRC stuff, and leverage's SDR transmitters while taking advantage of the expiration in January of most of the patents related to DAB.
Suddenly leaving community and small scale ILR stations on FM after DSO might not be such an obvious move - perhaps even some of the larger scale small scale stations have a DAB future after all (e.g. The Bee in Preston sharing with Frequency 1350, Preston FM, and Chorley FM, and then using a different block in Blackburn with there community stations) - why wouldn't you want to be on a platform that puts you on parity with the "big boys" while reducing your running costs significantly?
The document is also a DAB Pirate Primer 101.
The section 6.2 Mp2 Comparisons makes interesting reading.
They did test DAB+ in Brighton
For instance a small-scale DAB multiplex could be used in the Shetland Isles for SIBC, without the need for them to transmit across the whole of northern Scotland.
Maybe even William Rogers will start supporting DAB now.
This is good news. See more at:
http://brianlister.blogspot.co.uk/
This suggests that Ofcom is finally understanding that the government cannot announce a FM switch-off - a statement which is expected in a few months time - without there being a digital transmission option for the hundreds of smaller services who cannot afford (or who can't be accommodated on) DAB at the moment.
More background at:
http://brianlister.blogspot.co.uk/
1) use very expensive cavity filters to avoid interfering with other stations
2) use your amplifier at a fraction of its capacity, and make sure that all elements in the signal chain are absolutely 100% linear to avoid interfering with other stations
3) interfere with other stations
Now, which approach do you think pirate DAB stations will take?
I agree.
I said the same things over ten years ago...
https://groups.google.com/d/msg/alt.radio.digital/1OQAdsl7iAQ/9ioZmjFHerAJ
It's worrying that radios can't decode more than 128kbps when broadcasts use the most robust error correction mode DAB has (PL1). I knew about the 192kbps limit at PL3, but this paper clarifies that the limit is CU (capacity units which include error correction), not audio bitrate, hence the maximum bitrate these radios can decode falls as transmissions become more robust.
The elephant in the room is spectrum. A single local station on FM takes ~ 0.3MHz. A single local multiplex on DAB takes 1.7MHz. Like it or not, in many areas local stations are going to have to get together to make this work.
It'll be interesting to see how OfCom chooses to license this "interleaved spectrum". Stations on "interleaved" frequencies will be the first to burble and vanish when there's a lift on. It will make being a planner at OfCom an interesting job.
Cheers,
David.
The new Surrey mux which has William Rogers' UKRD as a shareholder is now due for launch December. The Brighton test used 10C which when used for Surrey might mean 10C can't be used again in Brighton? but MXR has released new frequencies 12A, 12C and 12D, also 5A was not used in Oxford
http://stakeholders.ofcom.org.uk/consultations/dab-coverage-planning/annex-b/
and 11A not currently used for d2 are possible frequencies for small scale muxes?
A shame this didn't happen a few years ago but better late than never.
The patents expiring, the Raspberry Pi, the rise of cloud services and the advancement of SDR have been a perfect storm for this! If anything DAB+ is less likly of this thanks to 1) the "ease" of setting up such a multiplex, 2) the low demand likely to be placed upon it, 3) the patents issue.
A statement which mere listeners such as myself whole-heartedly agree with - local stations have every chance of bringing the variety that we need on DAB listening, that has so rarely ( outside the BBC ) been achieved with national DAB stations. Just how many 80s hits stations do we want, or unfunny presenters ? Thank you everyone above who's explained what's going on - what this means for a brighter future for UK digital radio...
Interesting in the link that you posted at the start of this thread, Kev, that it doesn't seem much more of a hurdle ( from the Swiss demos ) to go to the various other types of encoding at a later stage if needed ? ( In my ignorance as a listener ) to start off on DAB then switch to DAB+ as everyone else changes ? Thanks for the link anyway - very interesting reading and viewing of the demos !
The tricky bit has always been the modulation - basically you take a bunch of streams (MP2, MP3, WMA, AAC) and wrap them all up - you could broadcast a whole DAB multiplex with MP3 streams if you wanted too - nothing would be able to decode them mind! (Hence why they can add DAB+ onto existing multiplexes).
Looking around to see if DAB SDR is now viable (I want to add the local DAB services to TVHeadend to enable recording so need them as multicast streams...) I submbled upon this
http://forums.stereotool.com/viewtopic.php?f=8&t=4859 - hmm, where's my DAB iTrip
Yes that's what I thought Phil. There is the potential for more choice and better quality with this development. I do wonder if it'll encourage the pirates though as other have suggested.
From the document So looks like the digital power module is an existing unit which will reduce costs, but dies not go into detail, says it could deliver up to 100w, which seems to indicate the Brighton tests were less than this?, but covered the town.
from section 8
From Section 5.2
The amplifier provided a nominal gain of 34dB and the linearity was such that it could provide a very clean signal up to ~0.5 watts RMS. Exceeding this level caused the COFDM modulation peaks to enter the compression region of the amplifier which resulted in spectral re-growth. This is common in digital transmission, and such products are generally removed from the output of the transmitter by expensive band-pass cavity filters. Although the power module was capable of producing much more power, an objective of the experiment was to keep the costs to a minimum. The additional band-pass filtering required for compliance with the ITU mask precluded the available higher power levels from being used, but the high gain and clean drive made it an excellent pre-driver for a further, more powerful stage.