dab digital radio DMB radio

Unbiased advice on DMB, DAB and DAB+?

Confusion by LuluP @ Flickr

It’s really great to meet radio people from around the world, and talk with them about their own plans for digital radio. But I’m often surprised about how much confusion surrounds DAB, particularly the (mis)information about DMB.

It’s important that radio people inform themselves properly and independently about the technology choices they are making. This isn’t as easy as it seems.

Follow the money

What isn’t widely understood is how money flows around the business of consumer electronics these days. You might think that a manufacturer makes money through retail margins – selling radios at a price higher than it cost to produce them. That’s certainly true, but the economics of the last decade or so have eroded retail margins to be incredibly slim. You don’t make much money simply by selling radios.

One of the issues that is new to radio people is IPR – Intellectual Property Rights. IPR represents some “cleverness” that a company (or group of companies) has thought up to make technology work better/cheaper or both. Legally, they “own” that idea or process, and they can choose to licence it to third parties. A modern consumer electronics device (like an MP3 player) may have IPR from twenty or thirty companies in it, and everyone of those companies is entitled to a licence fee. There has recently been a case where consignments of MP3 player have been seized because the manufacturers have not been paying the IPR for using the MP3 technology.

Here’s a specific example: To enable DAB+ or DMB requires an audio encoding technology called HE AAC, combined with a technology called SBR (Spectral Band Replication). These two technologies cost €1.60 and €0.15-€0.20 to licence per receiver respectively. So every DAB+ or DMB enabled receiver generates €1.80 to those companies who own the IPR rights. Multiply that across every radio sold in the world, and that’s a substantial amount of revenue. Put it another way, if you sell a DAB+ radio for €20, as a manufacturer you might get €0.50 from the retail margin, but as an IPR licensee you might make €1. Making radios is not as profitable as making the technology to go into radios.

But here’s an interesting thing. A DMB device also needs another technology called MPEG II Transport Stream. That adds another $0.50 per device. So a DMB device automatically has a higher cost, even if it only ever decodes audio. And there’s another $0.50 per device flowing to an IPR holder (or group) somewhere.

So who owns this IPR?

It’s not always clear who benefits from these extra licensing fees. But it does stand to reason that the companies who have IPR rights in a particular technology will be those companies most enthusiastic about promoting it and getting it widely adopted. And boy, there’s no wider adoption than radios. (100m radios in the UK alone – only 70m mobile phones). It’s a vast vast opportunity. If you’re an IPR holder, even if it’s only a total of €0.10 per device, you could be looking at millions and millions of Euro in licence revenues for decades to come – just by persuading someone to use your technology.

So now it makes sense why a technology company might fly people around the world, and make expensive “prototype” devices to encourage uptake of that technology in which they might have IPR rights -both declared and potentially hidden too. A few hundred thousand Euro in airfares, flights and prototypes might net millions of Euro return.

(It would be like the petro-rich countries encouraging the development and universal adoption of the internal combustion engine. Whatever the development and marketing costs were, they would quickly be dwarfed by the petro-dollars rolling in for decades and decades).

Take your technology advice carefully

Here’s a harsh statement – don’t trust technology suppliers to give you impartial advice. They might benefit substantially from your decision. They are selling you a solution, and your consumers will be paying for it with every device they buy, for ever.

So now you have an insight to the motivation of technology suppliers, who can give you impartial advice?

Well, the answer is that few people can give you genuinely impartial advice. But I would suggest that other broadcasters probably have objectives more similarly aligned to your own, and rarely have an IPR interest, so their advice might be far less prone to distortion. But of course, they don’t make any money from their advice, so sometimes it’s hard to justify spending a few hundred Euro one a flight and a hotel to discuss these things.

I’m disappointed that our colleagues in France have an expectation of Digital Radio that’s virtually identical to ours, but have been sold a completely different set of technologies to delivery it – technologies that will add about €0.75 in IPR to every single digital radio sold around the world. We can’t afford to make that mistake again.


Sometimes it appears that technology suppliers would prefer the simplicity of DAB to be obscured from broadcasters. It obviously helps them sell a solution if the solution looks hard.

DAB isn’t complicated, but you do have to know your options.

Everything in DAB starts with the multiplex, and the multiplex can support a mix of technologies all co-existing in the same set of spectrum and on the same infrastructure. The most prominent DAB applications are:

  • DAB – confusingly, the same name as the multiplex. DAB refers to the original method of broadcasting radio using the MPEG Layer II audio encoder, and this is now largely superseded by DAB+. You can add visuals to DAB using Slideshow at frame rates of up to 1fps.
  • DAB+ – the way to transmit Digital Radio. DAB+ is a direct upgrade of DAB. The great thing about DAB+ is it supports exactly the same data services as DAB, so there is a clear migration path for countries using DAB now (like the UK) into DAB+ without starting from scratch again.
  • DMB – the way to transmit mobile TV. DMB is substantially more complicated and expensive to transmit, and on the receiver, than DAB or DAB+. Unless you absolutely need to transmit TV (moving pictures with synchronous audio), you should not be considering DMB at all. The extra IPR load on the receiver of DMB is nasty, and makes the idea of a €50 radio with a colour screen virtually impossible.

I don’t think I can make it clearer than that. Don’t use DMB for radio, as it’s unnecessarily complex and expensive for radio, even radio with visuals. Use DAB+, as it was beautifully developed (largely by broadcasters with no IPR interests) to work brilliantly for radio. DMB was knocked up in a hurry to support mobile TV – it works, but it’s not elegant. But you can mix them all together in the same DAB multiplex just fine.

Radio companies of the world need to stick together

Together there is enough knowledge and understanding of the technology within broadcasters not to have to rely solely on technology suppliers for advice. The problem is that we appear to be really lousy at talking to each other about it. The WorldDMB Technical Committee helps a bit, but often decisions are being taken at higher levels than that, and there simply aren’t enough commercial radio broadcasters participating in WorldDMB.

That needs fixing before we fall into another IPR trap that will cost us all money.

9 replies on “Unbiased advice on DMB, DAB and DAB+?”

“Germany flicks off-switch on DAB”

“Part of the problem is that analogue FM never went away and most people didn’t seem to care for the clear digital-quality sound, and were left nonplussed by such benefits as easy tuning and message displays with song names and titles. DAB is struggling almost everywhere in Europe.”

“Report: Future Of U.K. Digital Radio May Be Bleak”

“LONDON — January 30, 2008: A report from Enders Analysis found that digital audio broadcasting, or DAB, is in trouble due to the high cost of transmission and slow revenue growth, U.K. newspaper the Guardian reports.”

“Macquarie Radio execs: delays have ‘killed’ the future of digital radio”

“Macquarie Radio Network says years of delays had ‘killed’ the future of digital radio, which was being overtaken by broadband services, third-generation mobiles and digital devices such as iPods, reports Australian IT.”

“Digital Radio in Canada”

“The Commission is very concerned about the stalled DRB transition. Roughly 15 of the 76 authorized stations (including the digital-only operation in Toronto) are not on the air. Some stations that once operated have since ceased operations. Few recievers have been sold, and there is no interest in expanding DRB service beyond the six cities where it exists.”

“Why don’t you use DRM on Shortwave?”

“Some large companies are pushing DRM by convincing program producers and broadcasters to start airing in DRM ahead of time, but unless there is mass availability and penetration of receivers on the listeners side, this will remain – unfortunately – a technological experiment, and broadcasters using it to reach their listeners now, are clearly throwing their money out of the window.”

“Straining to hear digital radio – Europe has a head start in terrestrial digital radio, but is anybody listening?”

“But today, digital radio is struggling to find its legs. While it’s still in the cradle in the United States, it has begun to crawl, a bit, in Europe and elsewhere… Europe has had a standard for digital radio for some time. The European Union adopted the standard, called Eureka 147, 10 years ago. But high prices and a lack of consumer interest have kept the market tiny.”

“Annual DAB sales 50% below forecast”

“The following graphs are copied from the DRDB’s (Digital Radio Development Bureau — UK DAB’s marketing and PR arm) sales forecast documents from 2004 and 2007, and they show that the forecast sales for 2008 are a massive 50% below what the DRDB had previously forecast they would be for 2008, and the cumulative sales will be 18% below previously forecast by the end of this year and 30% below what they had previously forecast by the end of next year.”

As PocketRadio stated, digital radio broadcasting (DRB) has either stalled, or is failing, worldwide.

Not sure you wanted to comment on this article particularly, or just have a bit of a rant. Either way, a collection of links to negative stories about Digital Radio (in all flavours, judging by the inclusion of DRM in there).

The Register article you quote is already discussed elsewhere in this blog, and was looking inaccurate the day it was breathlessly posted. The Enders report is similarly discussed here, and you probably ought to have looked beyond the headline to the detail.

As for the final link, I would expect that you have found a kindred spirit there.

Awww, bless.

And actually, you’re wrong, because he is no kindred spirit of mine – I constructively criticise, whereas he just criticises. For example, I suggested using mobile TV systems instead of DAB as far back as, oh, around 2003. And what is the DAB+ system? It’s basically just DMB mobile TV system without the video. Just think if you’d have taken my advice a couple of years earlier – it would have been a hell of a lot easier to carry out the switch to DAB+…

I’m afraid I disagree that DAB+ is basically DMB mobile TV without the video. Indeed, one of the key points I’ve made in this blog posting is that DMB requires the use of MPEG2 Transport Stream, and incidentally, enforces some rather awkward quantisation issues on the HE AAC v2 encoding which means that it is unable to spread the bit load across adjacent access units. Or to put it another way, HE AAC v2 in DMB sounds worse than DAB+ at equivalent coding rates. There’s also a rather nasty initial acquisition period with DMB, which isn’t the case in DAB+ where the access units are aligned with CIFs. And DAB+ appears to have better performance in marginal signal strengths.

But DMB only uses the MPEG-2 Transport Stream because DMB includes video, so where I say that DAB+ is just DMB without the video, I’m including the MPEG-2 TS in with the video part of the system.

I think we fundamentally view this from entirely different perspectives: I view things from a high level, whereas from what you’ve just written, you were focusing on the more nitty gritty dadta specification elements. The way I view the two systems is that they both consist of the following identical components:

AAC+ audio codec, OFDM modulation scheme, convolutional coding

The RS coding used on DMB and DAB+ is only very slightly different, because the code rate is almost identical (DAB+ uses 110/120 = 0.917, and DMB uses 188/204 = 0.922), and it is the code rate that determines the error correction performance – DAB+’s slightly lower code rate might explain the better performance you refer to, or it might be related to DAB+ using an additional error detection code (can’t remember what it’s called now, fire code maybe?) for the AAC+ audio headers.

The technologies above are by far the main technologies that determine performance, because at the end of the day, although DMB is less efficient than DAB+ due to DMB’s use of the MPEG-2 TS, which is less efficient for low bit rate radio stations, it’s only 20% less efficient at really low bit rates, which isn’t the end of the world, really. For example, DRM+ and DVB-H2 will both be, I dunno, 200 or 300 percent more spectrally efficient than DAB+ is, which is down to the ageing transmission scheme.

The reason why I object to you using words like “beautiful” and “elegant” to describe DAB+ is that when viewed from a high level, DAB+ is neither of those things. I’m not disputing that the people who worked on the design of DAB+ did good work taking into consideration the constraints they were working within, which was that it had to be backwardly compatible with DAB’s transmission scheme, but it’s still adding a state-of-the-art audio codec on top of an ageing transmission scheme – convolutional coding + RS coding was state-of-the-art in the early 1990s, so even though that’s a big improvement on DAB’s error correction coding, it’s a long way away from being state of the art.

And on the subject of state of the art, I noticed that John Ousby tried to patronise me on his blog by suggesting that using DVB-H2 was “naive”, but I couldn’t be arsed telling him why he was wrong. DVB-H2 is very likely to include both MIMO and transmit diversity (MISO), and on the page on my website where I mentioned DVB-H2, I didn’t say that you had to use it with MIMO.

DVB-H2 is going to be able to use 1.7 MHz channels, which means that it could be used in Band III, so if you only used transmit diversity rather than MIMO, it should be possible to incorporate DVB-H2 on a DAB/DAB+ receiver simply by adding software – which is no different to the addition of DAB+.

This does of course go against what you were saying in your blog, because you were making the point that using MPEG-2 incurs higher licensing costs, and so would using DVB-H2. But I don’t concur with your view that higher licensing costs are the end of the world, to be honest. I’ve no doubt Colin Crawford makes it sound like it’s the end of the world, but the choice of which technologies to use shouldn’t be dependent on how much profit receiver manufacturers can make – the decision should be on what’s best for the country, and in the long term DVB-H2 is a far, far better system than DAB+.

The way I look at it is that FM switch-off is still a long way away, so we should be asking the question: what system do we want to be using when FM is switched off? Do we really want to be using DAB+? I fail to see why we should have to put up with the relatively poor performance that DAB+ provides in comparison to state of the art systems when there’s so many years before FM is switched off, and a shiny new system could be silently introduced in time for FM switch-off. DVB-H2 is the best *current* system when viewd over the long term, but all I did was to suggest that it should be considered, but the DRWG seem to be so wrapped up in the short term of saving DAB that they seem unable to look beyond the end of their noses (and short-term decisions make for terrible long-term decisions…). If you want to avoid the additional licensing costs that using DVB-H2 would incur, then WorldDMB should design a successor to DAB+ so that we can switch to it when FM is switched off.

At the end of the day, DAB receiver modules are implemented using software-defined radio, so they’re just a microprocessor running some software (ignoring the RF part, which I’m not proposing should change), and over-the-air (OTA) software upgrades could be implemented if WorldDMB wanted to – over-the-Internet software upgrades have already been implemented on Wi-Fi radios. If OTA software upgrades were implemented, that would eliminate the problem we’re facing at the moment with switching from DAB to DAB+ which requires people to replace their receivers, so it also eliminates the need to build a market and so on.

Take DVB-H2 as an example. The technologies that DVB-H2 can use which DAB+ cannot are as follows:
(DVB-H2 is very likely to be the same as DVB-T2 but with additional technologies that are required for mobile reception)

* 16/64-QAM
* Near-optimal LDPC + BCH error correction coding
* MIMO and transmit diversity (MISO)
* OFDM with various guard intervals and numbers of subcarriers

Ignoring MIMO, because MIMO requires a receiver to have 2 or more aerials, all of the other technologies could be implemented in a DAB+ version 2 and then delivererd to DAB+ receivers via OTA software upgrade, and none of these technologies have any licensing costs associated with them. Implementing these technologies would increase the spectral efficiency of DAB+ (comparing for equal required C/N ratio levels) by about 100 to 200%.

The main arguments against using OTA software upgrades were that receivers would need to be over-specified if they want to be future-proof, but DSP processors are far more powerful than is required for digital radio reception / decoding (I’d bet Frontier-Silicon’s latest chips could handle the above technologies if they used a faster clock speed, and Radioscape just uses off-the-shelf DSP chips), and memory is cheap and it’s subject to Moore’s Law anyway, so cost per bit halves every 18 months or so. I’ve written a longer article about OTA software upgrades here:

It’ll be years before FM could be switched off, so I think it would be a piece of piss to implement OTA software upgrades and then implement a DAB+v2 and push it to receivers.

I’m sorry, but all you’re doing is continuing to theorise about potentially better transmission technologies, and thus missing the bigger picture. There are many factors required for a successful digitalisation of radio. Obsessing about perfect technology solutions does not make digital radio happen, and indeed, continues to delay it.

If you re-read your last-but-one post, it sounded to me like you had concerns with what I’d written regarding the difficulty of implementing new technologies and building new devices and a market for such new technologies. If you wanted me to address what you refer to as “the bigger picture”, then I feel you should have made that clearer in your post. I specifically responded to this sentence:

“The difficult thing is getting it implemented. Your comment, whilst long on (theoretical) technology, doesn’t address the issue of making devices and building a market.”

because I feel that by implementing over-the-air software upgrades (which would be simple to implement) you can postpone implementing new technologies for a while (so long as new receivers can support the technologies once they’ve been implemented), and it eliminates the need to build a new market for new devices.

But on the subject of “the bigger picture”, I’ve already said on numerous occasions on my website and also specifically in the article on The Register what I think needs to happen to make digital radio “better” so that it grows faster. For example, if the DRWG’s plan is to ignore the Internet and implement (effectively) a DAB-only plan, which is precisely what it looks like they’re trying to do from the DRWG stakeholder meeting presentation slides, then digital radio won’t grow anywhere near as quickly as it could do if the Internet were fully exploited – we’re moving to an on-demand world, so pushing everyone on to DAB, which can’t do on-demand, is about the stupidest idea it’s possible to come up with at the present time. Still, I suppose the DRWG’s planning will provide plenty of material to keep my website going for a few years to come, and judging by what the BBC people have been doing over the past year it’ll keep the BBC Trust busy with my complaints as well.

I think that the audio quality should be much higher than it is at the moment. I honestly think that if the audio quality were higher then growth would have been higher, because more people would give word of mouth recommendations and better online reviews and so on, which is what a “new” technology needs. I think your industry has used the market research into DAB’s audio quality simply to justify the original decisions regarding bit rates, as well as the decision to allow 112 kbps, but I think you’re collectively wearing blinkers about this issue: you seem more interested in being seen to be right than putting it right so that DAB can grow faster than it is doing – as an industry, you’re cutting off your nose to spite your face, IMO.

There are ways to improve the audio quality now, such as if the BBC put a couple of stations on Digital One or on Digital Two (if or when it launches), and local multiplexes with empty space could be reconfigured to allow some stations to provide higher quality – surely it’s not impossible for commercial radio groups to come to temporary agreements about how to use empty space on a multiplex?

Obviously, moving to DAB+ would be a better solution for providing higher audio quality, but your industry has managed to completely screw up the introduction of DAB+ up to now. The official go-ahead to design DAB+ was given at the WorldDAB general assembly in October 2005, so your industry had bags of time to plan for the introduction of DAB+, but look at how few DAB+-capable receivers there are on sale today. One thing you told Andrew Orlowski that he wrote in the article on The Register was that if people knew about DAB+ they’d stop buying DAB radios. But if the vast majority of DAB radios on sale were DAB+-upgradeable, it doesn’t make sense that people would stop buying DAB radios. And as for Peter Davies saying that if DAB+ were introduced early people would abandon digital radio forever, that’s simply laughable. DAB needs to be improved now – introducing DAB+ gradually will take *far* too long to fully reap the benefits it has to offer.

DAB+ is also the key to improving other areas where DAB’s proposition is weak:

* too many niche music genres aren’t covered – the fact that Planet Rock has been saved shows the importance of niche genres

* reception quality needs to improve, but improving reception quality on DAB could be very expensive

* DAB’s transmission costs are too high for you broadcasters

But as I’ve said, your industry has already decided to take the slow road to DAB+, so I suggest that you should concentrate on platforms where digital radio could grow quickly, i.e. the Internet and mobile broadband / 3G. DAB will grow due to the TV ads for the new Channel 4 stations, but Peter Davies is tripping if he thinks it will be “DAB’s Freeview moment”, and there’s simply nothing to gain by ignoring the Internet’s potential anyway.

Also, returning to what I said about using new technologies along the lines of those that DVB-H2 uses: it was implicit in what I was saying that by using more efficient technologies this makes more capacity available in the limited spectrum that’s available to DAB, and more capacity being available would allow the digital radio proposition to be better than it is when using less efficient technologies – I’ve made this point many, many times on my website on the subject of DAB+, and it is true in general. I don’t think you can be complacent over the long-term, because radio is under threat from other areas, so I don’t think you should turn your nose up at more capacity.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.