A discussion on digital radio technology

Charles Slane · 01-01-2006 10:01PM #1

Let's hope they keep up the high bit rates.

For anyone who doesn't know much about dab radio, here's a link to an interesting and informative UK site - http://www.digitalradiotech.co.uk/index.htm

Propellerhead · 01-01-2006 10:23PM

That site is chock full of anti-DAB propaganda though?

Conor S · 01-01-2006 11:10PM

In fairness a lot of what is on that site is fact.
We would be better off with DRM plus instead of DAB but as there are no DRM plus receivers available in this country I cant really see it happening.Maybe in the future sometime
Does anyone else find that 2fm sounds crap at 192kbps compared to the likes of the BBC at 128kbps??
Conor

Charles Slane · 01-01-2006 11:12PM

It may be a bit anti-DAB skewed, but it does also have tons of info regarding bit-rates, channels, and what radios are available in the UK (and their prices in each shop).

As far as I can see the anti-DAB bias has more to do with the low bit-rates used in the UK rather than the DAB system itself (although most would agree that it IS old technology at this stage).

If the bit rates are kept high in Ireland I'd be very interested in buying a DAB radio.

DMC · 02-01-2006 07:39AM

Ahhh... the RD1.... meant to be a MW/LW inside it... never happened.

I'm glad you got the BBC back, I was sure that if there you had an external aerial, this was the case. 12B and 12C are so close to each other, that a big aerial might like RTÉ DAB too much. This doesnt seem to be a problem the further you are away.

As for Steve Green's site.... He knows his onions, no doubt about it. His posts on alt.radio.digital on usenet can have a bit of warmth, but he's very against broadcasters squeezing the living daylights out of the spectrum, while compromising on sound quality.

6:40am update... no sign of Three Rock yet, but I found a sweet spot for Clermont in my bedroom.

marksi · 02-01-2006 10:57AM

Unfortunately the guy from the digitalradiotech website appears to be rather obsessed with bit rates. While he does have valid points his "holier than thou" approach means he's incapable of believing that actually not everyone will notice the quality, and no, there are no more important issues in the world right now.

I've tried talking to him on newsgroups before and given up because his attitude sucks. Apparently low bitrates on DAB are the fault of every single person who works for the BBC and therefore all staff members deserve being slagged off because they're all as bad as each other.

Propellerhead · 02-01-2006 01:02PM

I'd agree - I don't rate him as a source of info on DAB as he's completely obsessed with running down DAB. While DAB isn't perfect and some of the points he makes about low bitrates are valid, his whole "I hate DAB" attitude both on the site and in newsgroups marks him out for me at least as a obsessed propagandist rather than a source of accurate information.

No doubt he won't mention the RTÉ tests as he's more interested in news about DAB failure.

lawhec · 02-01-2006 03:24PM

Propellerhead wrote:

I'd agree - I don't rate him as a source of info on DAB as he's completely obsessed with running down DAB. While DAB isn't perfect and some of the points he makes about low bitrates are valid, his whole "I hate DAB" attitude both on the site and in newsgroups marks him out for me at least as a obsessed propagandist rather than a source of accurate information.

No doubt he won't mention the RTÉ tests as he's more interested in news about DAB failure.

Have to agree. While he obviously knows his stuff and that there's little doubt that more modern technologies like DRM hold several advantages over DAB, his constant rabid put downs of anything said about DAB that isn't negative means he can't be held to have a NPOV on the subject. Moreso, he seems to think that DABv2 (AAC instead of MP2) would cure most of the problems of DAB, perhaps not realising that all you'd probably end up with is more automated jukebox stations than there already is in the UK on DAB. Also on newsgroups and forums he comes across as a complete **** especially if you dare say something that is contary to his thoughts.

The main problem in DAB doesn't mainly lie with the use of MP2 (though that's hardly a plus point) but in its specification rigidness and that it was designed for national and regional networks. For local radio outside of cities like Dublin and Belfast, DAB doesn't really work out well without it being a bandwidth hog, moreso than FM.

watty · 02-01-2006 06:34PM

At higher bitrates MP2 is probabily better than MP3. Its used for SVCD and recommended at 192K to 256k for as non AC3 for DVD when AC3 seems to make source material sound distorted.

The only problem with UK DAB was that they didn't reserve enough of the old Band III TV spectrum for it. UK DAB uses a tiny chunk of spectrum in Band III.

The solution isn't to change from MP2 or critize the BBC, but the people that had out licences and BW in UK.

DAB is fine in Germany and could be fine here too.

Though I don't kow what they will fill it with without upsetting incumbent Indepent Local Radio Ad selling Executives.

Isn't MP2 used for DVB-t and DVB-s radio and TV audio stereo/mono? Almost no DVB transmission uses AC3 etc, very little is in Surround sound.

digitalradiotec · 03-01-2006 07:15PM

Propellerhead wrote:

I'd agree - I don't rate him as a source of info on DAB as he's completely obsessed with running down DAB. While DAB isn't perfect and some of the points he makes about low bitrates are valid, his whole "I hate DAB" attitude both on the site and in newsgroups marks him out for me at least as a obsessed propagandist rather than a source of accurate information.

No doubt he won't mention the RTÉ tests as he's more interested in news about DAB failure.

You don't rate me as a source of info on DAB? Do you rate the WorldDAB website? Do you rate the DRDB's digitalradionow.com website? How about bbc.co.uk/digitalradio?

How many of those sources reported Finland switching off their DAB transmitter network? Did they accurately report DAB in Canada being sidelined? Did they report the Swedish government's decision to halt DAB accurately? Did they report that the Australian minister said that DAB is a 10-year-old (actually, it's 15 years old) system and that it would be "irresponsible" to use it?

digitalradiotec · 03-01-2006 07:24PM

marksi wrote:

Unfortunately the guy from the digitalradiotech website appears to be rather obsessed with bit rates. While he does have valid points his "holier than thou" approach means he's incapable of believing that actually not everyone will notice the quality, and no, there are no more important issues in the world right now.

Audio quality doesn't matter on a small portable radio -- I know that. And some people aren't bothered about audio quality anyway -- I know that too.

The main advantage that analogue listeners say they want from DAB in market research surveys is that digital radio should provide better audio quality, so why is it acceptable for the broadcasters to provide such very poor audio quality??

I've tried talking to him on newsgroups before and given up because his attitude sucks. Apparently low bitrates on DAB are the fault of every single person who works for the BBC and therefore all staff members deserve being slagged off because they're all as bad as each other.

I don't say that at all. It's only a small number of people at the BBC that you can blame for the current situation. They're mainly BBC Radio execs, but there's a couple of engineers as well who were part of the team that thought up the strategy that led to the new stations being added and the subsequent poor audio quality being provided.

digitalradiotec · 03-01-2006 07:56PM

Northern Correspondent wrote:

Have to agree. While he obviously knows his stuff and that there's little doubt that more modern technologies like DRM hold several advantages over DAB, his constant rabid put downs of anything said about DAB that isn't negative means he can't be held to have a NPOV on the subject.

Rabid put downs?

Moreso, he seems to think that DABv2 (AAC instead of MP2) would cure most of the problems of DAB, perhaps not realising that all you'd probably end up with is more automated jukebox stations than there already is in the UK on DAB.

I'm perfectly aware that DABv2 could be used simply to provide more stations at the same level of audio quality. But the reason I favour DABv2 (and all the other efficient newer systems) is because it hugely increases the likelihood that significantly better audio quality will be used, for the following reasons:

1. radio stations are charged for the bit rate they use, so if they only have to pay for, say, 48kbps (HE AAC) instead of 128kbps (MP2) it's far less likely that they'll choose to go even lower, e.g. 40kbps, and 48kbps HE AAC provides better audio quality than 128kbps MP2, therefore the audio quality would be improved.

2. spectrum is a limited resource everywhere and there's only a finite number of radio stations that would be economically feasible, so if a far more efficient system is used it is far more likely that the audio quality will not have to be degraded due to too many radio stations wanting to get on the limited number of DAB multiplexes. The UK is a case in point here, because even after the UK gets 4/5 new Band III channels there will still be about 120 analogue stations unable to get on DAB even though they're already using low bit rates.

3. public service broadcasters, such as the BBC, that have a full multiplex to themselves are far more likely to provide high audio quality rather than cramming more stations onto the multiplex. For example, if the BBC could change its DAB multiplex to DABv2 tomorrow, they wouldn't be allowed by the government to launch any new stations, so Radio 1, 2, 4, 6 Music and 1Xtra could all be changed from using 128kbps MP2 to 128kbps AAC, thus hugely increasing the audio quality.

Propellerhead · 04-01-2006 01:38AM

digitalradiotec wrote:

You don't rate me as a source of info on DAB? Do you rate the WorldDAB website? Do you rate the DRDB's digitalradionow.com website? How about bbc.co.uk/digitalradio?

How many of those sources reported Finland switching off their DAB transmitter network? Did they accurately report DAB in Canada being sidelined? Did they report the Swedish government's decision to halt DAB accurately? Did they report that the Australian minister said that DAB is a 10-year-old (actually, it's 15 years old) system and that it would be "irresponsible" to use it?

AM is nearly a hundred years old. FM is over sixty five. For how long do we have to chase technical excellence on ever changing systems before affordable receivers become available for that particular system? :rolleyes:

I'm not prepared ever to pay €500 + for a receiver that I can't get anything on.

I may be prepared to pay €50-€75 for a receiver that will expand my choice of radio in a country dominated by too few radio formats. There is only so much Top 40, Country Music and Death Notices on Local Radio in this country that I can take.

If DAB is ever fully rolled out here, and there is no guarantee that it will be, then I would expect that the bitrates here will be higher than the UK precisely because we are a much smaller market than the UK and that there will not be the same scramble for bandwith that caused the bitrate attrition that annoys you so much.

lawhec · 04-01-2006 11:26AM

digitalradiotec wrote:

1. radio stations are charged for the bit rate they use, so if they only have to pay for, say, 48kbps (HE AAC) instead of 128kbps (MP2) it's far less likely that they'll choose to go even lower, e.g. 40kbps, and 48kbps HE AAC provides better audio quality than 128kbps MP2, therefore the audio quality would be improved.

This makes a presumption. When DAB started in the UK, there wern't too many presumptions that we'd end up with a situation like there is right now

2. spectrum is a limited resource everywhere and there's only a finite number of radio stations that would be economically feasible, so if a far more efficient system is used it is far more likely that the audio quality will not have to be degraded due to too many radio stations wanting to get on the limited number of DAB multiplexes. The UK is a case in point here, because even after the UK gets 4/5 new Band III channels there will still be about 120 analogue stations unable to get on DAB even though they're already using low bit rates.

But this disregards the fact that outside main cities and large towns which may have several "local" stations, in other areas where there might be only one or two local stations, regardless of codec DAB is several times more spectrum hungry than FM. The spectrum rigidness of DAB speaks for itself - it is fine and dandy for national and regional networks but cannot work out for local radio without being a bigger bandwidth hog than FM. This is why IMHO DAB is only an "interim" system and will not survive in the long term, not even a movement to V2 would be enough.

3. public service broadcasters, such as the BBC, that have a full multiplex to themselves are far more likely to provide high audio quality rather than cramming more stations onto the multiplex. For example, if the BBC could change its DAB multiplex to DABv2 tomorrow, they wouldn't be allowed by the government to launch any new stations, so Radio 1, 2, 4, 6 Music and 1Xtra could all be changed from using 128kbps MP2 to 128kbps AAC, thus hugely increasing the audio quality.

Hello? The BBC already have crammed more stations on to their multiplex at the expense of providing high quality audio. And was it not the BBC who decided against using AAC for DAB in the late 1990's?

digitalradiotec · 04-01-2006 12:31PM

Propellerhead wrote:

AM is nearly a hundred years old. FM is over sixty five. For how long do we have to chase technical excellence on ever changing systems before affordable receivers become available for that particular system? :rolleyes:

I read a document written by someone relatively senior in the worldwide DAB industry that said that new DAB standards should be released in the next few months. I read that in about September / October, so they should be available soon. There's 2 DAB chip/module design companies that dominate the DAB receiver market: Frontier-Silicon and Radioscape, and both use software-defined radio techniques (i.e. the receivers are implemented in software), and both have already designed DMB modules (DMB is almost identical to DABv2), so receivers could be ready quickly. Radioscape's modules are *all* pin-compatible as well, so you could literally remove a DABv1 module and put a DABv2 module in its place in existing receiver designs.

And here's what Ofcom said recently on the subject of the UK adopting DABv2:

http://www.ofcom.org.uk/consult/condocs/vhf/statement/band3statement.pdf

"4.85 However, Ofcom is aware that some other countries, which
currently have no DAB services or where set sales have been minimal, are considering adopting the new standard.
4.86 These developments mean that receivers may be developed for the
international market which can receive both the existing MPEG 2 services and any new AAC MPEG 4 services (together with DRM, as discussed below). Ofcom would
welcome this development and encourages the radio industry to work with manufacturers to develop and market such radios as standard.
4.87 Such a development could mean that, a number of years from now, when the vast majority of radios in the market can receive both MPEG 2 and AAC MPEG 4
services, multiplexes in the UK may be able to make use of AAC, thus further
enhancing spectrum efficiency, allowing the provision of additional services to
increase listener choice."

It is logical from an economies of scale point of view for all DAB receivers to be version 2-enabled (because it's cheaper to manufacture 1 chipset/module rather than 2).

At the end of the day, why would you want Ireland to use DABv1 when it is well known that it uses an inefficient audio codec that requires very high bit rates to provide high audio quality when you could wait a few months to adopt DABv2 which will allow you to use the AAC audio codec which provides very good audio quality at bit rates as low as 96 - 128 kbps?

If DAB is ever fully rolled out here, and there is no guarantee that it will be, then I would expect that the bitrates here will be higher than the UK precisely because we are a much smaller market than the UK and that there will not be the same scramble for bandwith that caused the bitrate attrition that annoys you so much.

How many Band III channels are or will be available?

The current test is only using 192kbps MP2, which doesn't even provide FM-quality.

At the end of the day you've got a choice between using MP2 and AAC, and AAC requires half the bit rate MP2 requires, so if you want more choice and high audio quality then you will still need to use AAC unless you've got loads of free Band III channels.

digitalradiotec · 04-01-2006 12:59PM

digitalradiotec wrote:

1. radio stations are charged for the bit rate they use, so if they only have to pay for, say, 48kbps (HE AAC) instead of 128kbps (MP2) it's far less likely that they'll choose to go even lower, e.g. 40kbps, and 48kbps HE AAC provides better audio quality than 128kbps MP2, therefore the audio quality would be improved.

Northern Correspondent wrote:

This makes a presumption. When DAB started in the UK, there wern't too many presumptions that we'd end up with a situation like there is right now

Less than 50% of all analogue stations are on DAB now and yet they're already using low bit rates and even with the 4/5 new Band III channels there will be about 120 analogue stations unable or unwilling to transmit on DAB. The Radio Authority planned how many local and regional multiplexes there would be with the 7 existing Band III channels, so it should have always been obvious that we were never going to get high bit rates on DAB if they wanted to carry a high percentage of analogue stations on DAB.

digitalradiotec wrote:

2. spectrum is a limited resource everywhere and there's only a finite number of radio stations that would be economically feasible, so if a far more efficient system is used it is far more likely that the audio quality will not have to be degraded due to too many radio stations wanting to get on the limited number of DAB multiplexes. The UK is a case in point here, because even after the UK gets 4/5 new Band III channels there will still be about 120 analogue stations unable to get on DAB even though they're already using low bit rates.

Northern Correspondent wrote:

But this disregards the fact that outside main cities and large towns which may have several "local" stations, in other areas where there might be only one or two local stations, regardless of codec DAB is several times more spectrum hungry than FM. The spectrum rigidness of DAB speaks for itself - it is fine and dandy for national and regional networks but cannot work out for local radio without being a bigger bandwidth hog than FM. This is why IMHO DAB is only an "interim" system and will not survive in the long term, not even a movement to V2 would be enough.

I agree with you that DAB's wide bandwidth makes it unsuitable for many local stations. And I support DRM+ anyway, which I presume you're in favour of seeing as you don't like DAB's wide bandwidth. But I disagree that DAB is an interim system -- if Ireland chooses to mainly go with DAB then you're stuck with it for decades. You may use DRM+ alongside DAB, but other than changing to a subscription-funded and away from an ad-funded model I see no reason for radio broadcasters to want to invest in another digital radio system for a long time.

digitalradiotec wrote:

3. public service broadcasters, such as the BBC, that have a full multiplex to themselves are far more likely to provide high audio quality rather than cramming more stations onto the multiplex. For example, if the BBC could change its DAB multiplex to DABv2 tomorrow, they wouldn't be allowed by the government to launch any new stations, so Radio 1, 2, 4, 6 Music and 1Xtra could all be changed from using 128kbps MP2 to 128kbps AAC, thus hugely increasing the audio quality.

Northern Correspondent wrote:

Hello? The BBC already have crammed more stations on to their multiplex at the expense of providing high quality audio. And was it not the BBC who decided against using AAC for DAB in the late 1990's?

The BBC has to get permission from the government before it launches any new services, and they've already launched 5 new radio stations, so if they tried to launch any more the commercial radio groups' toys would be well and truly thrown out of the pram. So the BBC won't be launching any new radio stations for a long long time. And anyway, if they used DABv2 there is actually room to add a couple more stations while still providing excellent audio quality, so in the unlikely event that they are allowed to launch new stations the audio quality would still be high.

watty · 04-01-2006 04:08PM

The problem with both DTT and DAB is simple.

Analog Satellite used FM video (25MHz) or even hybrid analog/digital (25 to 30MHz) D2MAC.

Using the SAME transponder, you get a 10:1 saving roughly because MPEG2 (and now MPEG4) can do a huge amount of compression. Over 450 times.

For Terrestrial TV the saving is really only 3:1 at comparable quality to Satellite.

Audio compression even using the "best" and newest codecs simply doesn't come near the compression and subjective quality even of good MPEG1 video.

DRM is the newest and basically uses twice the bandwidth of pilot tone SSB, which could be used instead if AM to save 1/2 the power and spectrum at the same quality.

FM today could actually use less deviation and with good receiver design achieve at least 1970s FM quality probabily on 1/4 the bandwidth (NBFM or Phase modulation).

The overiding reason for Digital on Satellite was NOT quality. FM video with correct dish and the newer higher power satellites is mostly better than MPEG2 or MPEG4. The reason is a 10:1 saving in bandwidth = ten times more channels. Since Digital has FEC and a degree of multiplex separation by using different Symbol rates the channel separation can be less and even the same frequency used for H and V at the same time (not possible terrestrially). Thus near to 15: 1 can be achieved.

Lower bit rates for less important content bring saving up to maybe 20:1 or 30:1 for video MPEG on satellite compared with FM video.

But since analog Terrestrial is 1/4 to 1/5th of bandwidth and no H/V polarity overlaps can be allowed, savings on terrestrial digital video is only 1/10th of the savings on Satellite.

Thus it could be argued that DTT in UHF replacing analog TV is silly unless MPEG4 or reduced quality DVB-h.

IMO the case for DAB at all is dubious. If you are going to have DAB it has to be for different reason than change over to Digital on Satellite. It also really needs most of Band III, not just the tiny bit it got in the UK.

The DV camcorder, home DVD player and CD Audio outperform 8mm tape, VHS tape and cassette tape (or FM radio). But digital does not mean automatic quality or automatic bandwidth savings.

Serious mistakes have been made in UK because the nature and reason behind success of Satellite DVB, DVD players and CD Audio was not analysed. The false premise was that Digital = better quality and Digital = less bandwidth.

Given the difference between how we perceive sound and how we see, it is likely Digital audio (at similar subjective quality to "best" DVD film transfer video) will never use much less bandwidth than analog Audio transmission.

Who pays for DAB?

DRM has to compete with AM which suffers badly from noise, fading, co-channel, image channel, intermod and wasting 1/2 the spectrum. It has lots of advantages over AM for DX reception, especially on HF.

Even DRM+ is of dubious value compared with FM or Narrowband FM for LOCAL radio. Local VHF radio does not generally have a noise, fading, co-channel, intermod or image frequency interference problem like HF during day or MF / LF at night.

In any engineering solution:
1) What existing problem are we trying to solve?
2) What are the goals?
3) What new problems will the solution proposed have.

With Satellite and Cable the advantages of Digital outweigh the difficulties and problems.

With Terrestrial the goal of shutting off analog PAL Video is artifical. Only HD needs digital and the BBC can't see how more than 2 or 3 of those can fit in even if analog is turned off.

What advantage and disadvantages would DAB bring to Ireland?

What will it cost?

AM is going to go. It was once thought that it would be replaced by SSB with pilot tone carrier (Automatic tuning). But a DSP chip based receiver for DRM will deliver better quality and other advantages.

The Argument for ditching FM is less clear. Extra FM stations using NBFM could be accomodated in Band III more cheaply than DAB at higher quality!

lawhec · 04-01-2006 04:18PM

I agree with you that DAB's wide bandwidth makes it unsuitable for many local stations. And I support DRM+ anyway, which I presume you're in favour of seeing as you don't like DAB's wide bandwidth. But I disagree that DAB is an interim system -- if Ireland chooses to mainly go with DAB then you're stuck with it for decades. You may use DRM+ alongside DAB, but other than changing to a subscription-funded and away from an ad-funded model I see no reason for radio broadcasters to want to invest in another digital radio system for a long time.

As the Irish Republic only has two Band III DAB allocations cleared, DAB in its current form will never be able to match and add to the system of radio broadcasting on its own. Its quite obvious that if RTÉ regards these tests as "successful", they'll stick with their 12C allocation while the 12A allocation would be for a national commercial allocation like D1 in Britain. Therefore the local stations would simply not get on to DAB and would have to remain on FM for the forseeable future until either DRM/DRM+ became established or analogue television is switched off. Very similar arguements apply to Northern Ireland too - there is simply no more room to expand DAB here, with only one local station outside Belfast on DAB on the Score NI Multiplex. It ends up with the irony that in many parts of Northern Ireland there is more choice to be had on FM than on DAB.

According to another poster on this forum, RTÉ is planning to do DRM tests this year, and RTÉ don't lack in having spare high powered LW and MW allocations in which to test this on and so should prove interesting. These DAB tests are however interesting to certain people on this board as DAB receivers are readily available in shops - DRM receivers right now are near impossible to get outside of anorack circles. If only for selfish reasons I would hazard to say that most of us here with receivers would like these tests to "succed", but that we're well aware that better, if not yet mature platforms exist. I would like to hope that whatever the outcome of the DAB trial, RTÉ don't drag its heels over over the likes of DRM (IMHO DVB-H falls into the same traps as DAB). To me DRM (within its current and potential specs for DRM+) makes a lot more sense to use than DAB V2 as it isn't tied into fixed width multiplexing. The only way DAB could compete on that level is if DAB could work in a SCPC mode similar to DVB-S, but if you're doing that, you effectively end up with something similar to DRM + anyway.

digitalradiotec · 04-01-2006 06:56PM

watty wrote:

The problem with both DTT and DAB is simple.

<massive snip>

There's too many points to answer them individually cos I'd be here all day, so I'll just pick out a few:

The ratio of DTT to analogue terrestrial equals the number of TV channels you can fit into a DTT multiplex, which for the UK DTT muxes, which are either 24 or 18 Mbps, the ratio is 8:1 and 6:1 respectively. The saving is even more if you use large-scale SFNs, which the UK doesn't because it jumped in too quickly when there were only 2K chips available.

Although I actually agree with you that the case for using 400 MHz or so of prime spectrum (for mobile applications) for TV is wasteful in this day and age, there are legitimate arguments for why it should be used, such as the large user base of people that use analogue terrestrial and the fact that not everybody can install a satellite dish. So just comparing it to satellite doesn't tell the whole story.

I disagree with your point that audio compression doesn't come close to video compression ratios therefore it's a waste of time, because it never will be able to come close due to the nature of video being largely static, and only differences need to be encoded and similar reasons. I mean, do you think compressed audio has no use just because the compression ratios are lower? The massive success of MP3 players shows that it is very useful.

And I disagree with your view that digital radio is a waste of time because it isn't much more spectrally efficient than FM. For starters, you can't use SFNs on FM, whereas you can with DAB, DRM etc, so the spectral efficiency will be greater when you take the multi-frequency FM network planning into consideration, albeit that only applies for national / regional stations.

Taking DRM+ as an example, with a 50kHz channel it will be able to carry a radio station at near CD-quality, which is already much better than FM's spectral efficiency before you take the gain from SFNs into consideration.

Digital radio systems also use far lower transmitter power, which is beneficial to the broadcasters and it allows more radio stations to be fit in per unit area.

It's pretty obvious that I agree with your view that DAB is pretty much a waste of time in its current form, but I do support DABv2, because it'll have the HE AAC and AAC audio codecs, so it'll be much more efficient.

DRM+ is still leagues ahead of DAB or DABv2 though, because it should allow a 112kbps AAC station (near CD-quality) to be carried in a 50kHz channel, whereas to achieve the same audio quality on DAB you need to use a bit rate of 224kbps MP2 and that requires 350kHz of spectrum -- so DRM+ is 7 times as efficient as DAB.

And in fact, if you used digital satellite distribution for radio you'd be less efficient than DRM+, because DRM+ allows 16-QAM and 64-QAM whereas satellite is usually limited to QPSK.

I do agree with you that the UK made some major mistakes when they decided to go digital, but that's governments for you.

I don't see how DRM+ is of dubious value compared to FM for local radio. You can question why bother going digital at all, but if the desire is to go digital then DRM+ will be an excellent system.

Regarding whether or not to go digital, I think it does have clear benefits:

* better audio quality if the right bit rates are used (and you can still save spectrum if you use an efficient system)
* more choice
* no hiss

Basically, I'm in favour of digital radio and always have been, but it's just the UK DAB implementation that I dislike.

digitalradiotec · 04-01-2006 07:08PM

Northern Correspondent wrote:

If only for selfish reasons I would hazard to say that most of us here with receivers would like these tests to "succed", but that we're well aware that better, if not yet mature platforms exist. I would like to hope that whatever the outcome of the DAB trial, RTÉ don't drag its heels over over the likes of DRM (IMHO DVB-H falls into the same traps as DAB). To me DRM (within its current and potential specs for DRM+) makes a lot more sense to use than DAB V2 as it isn't tied into fixed width multiplexing. The only way DAB could compete on that level is if DAB could work in a SCPC mode similar to DVB-S, but if you're doing that, you effectively end up with something similar to DRM + anyway.

I agree that DRM+ is by far a better system than DABv2 let alone DABv1. But that begs the question, why do you want these DAB tests to succeed? You can say that RTE will keep bit rates high (although they're already using MP2 bit rates that provide sub-FM quality in these tests when the mux isn't full...), but what about the commercial multiplex? Don't you think that would just end up like all the DAB muxes we have in the UK?

One of the main advantages of DRM+ compared to DAB is that with a fixed bandwidth you don't get the "let's trade off audio quality to add another station so we can make extra profit" effect.

If you're basing your views on selfish reasons, I'd have thought that you'd want these tests to be an abject failure so that you don't have to put up with this lousy DAB system for decades. Or at the very, very least, surely you'd want them to be using AAC than MP2, i.e. DABv2 rather than v1?

watty · 04-01-2006 07:32PM

MP3 digital is successful because it is solid state or small hard disk. The comparison is not with analog but CD audio format.

Obviously since Hard Disk or solid State players MUST be digital, ANY compression is worth while.

Satellite doesn't have a great BW problem. Even using QPSK there is 10:1 to 30:1 saving. At Best Satellite Quality and robustness, COFDM/DTT is more like a 4:1 saving on Analog PAL. Satellite can't use QAM64 etc as the noise margin won't allow it.

I agree that DRM comes out best in Bandwidth.. Analog NBFM actually uses LESS BW than DAB V2 for same audio quality.

digitalradiotec · 04-01-2006 08:31PM

watty wrote:

MP3 digital is successful because it is solid state or small hard disk. The comparison is not with analog but CD audio format.

MP3 became successful through file-sharing networks, which were long before MP3 players arrived.

The issue I had with what you said though was that you seemed to question the validity of audio compression per se just because it cannot compress to anywhere near the compression ratios that video can achieve. But I think audio compression has proved to be very successful. If you only meant audio compression with regards broadcasting applications then that's a different matter.

Satellite doesn't have a great BW problem. Even using QPSK there is 10:1 to 30:1 saving. At Best Satellite Quality and robustness, COFDM/DTT is more like a 4:1 saving on Analog PAL. Satellite can't use QAM64 etc as the noise margin won't allow it.

Satellite for digital radio still has bandwidth problems because they're not going to use 12 GHz like that used by Sky et al. There's not a great deal of spare spectrum when you look at it for satellite digital radio, because from what I've read they plan to use spectrum around 1400 - 1500 MHz, and mobile TV applications will probably be competing for quite a bit of it.

If you look at XM and Sirius satellite digital radio systems in the States they're only using around 4 MHz of spectrum each!

I agree that DRM comes out best in Bandwidth.. Analog NBFM actually uses LESS BW than DAB V2 for same audio quality.

Any system uses less bandwidth than DABv2 or v1 for the same audio quality.

rlogue · 04-01-2006 09:10PM

I see DAB Sounds Worse than FM's site is still leading with the Swedish story.

watty · 04-01-2006 09:41PM

digitalradiotec wrote:

If you only meant audio compression with regards broadcasting applications then that's a different matter.

Yes. I'm a Communications Systems Engineer and Telecoms Design Engineer a while ago. I'm comparing Analog audio transmission with Digital Audio Transmission.

In 1986 I designed a voice mail system using custom chips to do Analog to Digital and POTS line functions. Because 64K codecs used to much disk space in 1986 we convinced ourselves that 1bit 4 x oversampled digital conversion was OK (16kbps CVSD modulation). Actually it sounded terrible. Some early motorola radios with encryption used it though as the A/D and D/A. I have a Racal designed frequency hopping radio set that simply used SSB and had much better audio quality and was more secure too!

lawhec · 04-01-2006 09:45PM

digitalradiotec wrote:

I agree that DRM+ is by far a better system than DABv2 let alone DABv1. But that begs the question, why do you want these DAB tests to succeed? You can say that RTE will keep bit rates high (although they're already using MP2 bit rates that provide sub-FM quality in these tests when the mux isn't full...), but what about the commercial multiplex? Don't you think that would just end up like all the DAB muxes we have in the UK?

One of the main advantages of DRM+ compared to DAB is that with a fixed bandwidth you don't get the "let's trade off audio quality to add another station so we can make extra profit" effect.

If you're basing your views on selfish reasons, I'd have thought that you'd want these tests to be an abject failure so that you don't have to put up with this lousy DAB system for decades. Or at the very, very least, surely you'd want them to be using AAC than MP2, i.e. DABv2 rather than v1?

You're trying to read my mind, and you're failing miserably.

The main reason why I would like to see RTE take forward these DAB tests (though it'll be no skin off my nose if it fails) is that it would finally establish RTE to take on a digital broadcasting setup of its own without having to rely on Sky or cable as its partners. We haven't a DTT network either though tests are again expected soon. The Irish Republic is currently the poor man of Western Europe regarding digital technologies but the past has shown that the Irish public is reasonably quick to take up on new technology while still maintaining a back-up redundancy, the most obvious being 405 line vs. 625 line TV transmissions back in the early 1960s hence the phrase "Irish Solution to an Irish problem". Local radio in the Republic, which accounts for a large amount of listenership, cannot be accomodated on VHF Band III DAB within current allocations. My prediction would be that these stations will bypass DAB altogether and appear on DRM / DRM+ when the right time comes. DAB, if RTÉ decides to roll it out nationally, will probably end up as 405 line TV in its time, making it available to those who want it but not give too much publicity.

I also don't really see any reason why at this stage using AAC codecs on DAB is worthwhile - existing DAB receivers would become obselete, which I believe in the UK stands at around 2 million. The UK government after WW2 decided to keep the 405 line system going even with only a few thousand sets going in London even with a better spec 625 line system being made available across contentinal Europe. As I see it in the UK, the current system for DAB is past the point of no return. If AAC was to be implemented, it should have been done five years ago at least. It's too late now. Even if AAC was tried now it still wouldn't solve the problems that DAB inherently has with its fixed bandwidth properties. As it stands, the 128kbps MP2 DAB stations range in quality from tolerable to shocking - I can listen to the BBC stations though a set of headphones but not the Score NI stations. If AAC/HE-AAC is going to be thrown in somewhere it might as well be done with something like DRM which gives the flexibility for national, regional, local and community radio. For DAB as it stands , Ofcom has been too toothless to try and ensure a minimum bitrate standard and letting commercial interests beat them down.

watty · 04-01-2006 09:47PM

digitalradiotec wrote:

... Satellite for digital radio still has bandwidth problems because they're not going to use 12 GHz like that used by Sky et al. There's not a great deal of spare spectrum when you look at it for satellite digital radio ...

Also compared with Analog transmission (or Video compression) the compression at decent quality isn't great.

For most people it is personal players that is the big use of MP3, not the minority people who are internet users and steal music. I setup an iPod Nano on Win 98 original (not SE) tonight for a neighbour and he doesn't even have Internet. Audiograbber for his CDs and ephPod for iPod control. And a rather 3rd party USB storage device driver.

lawhec · 04-01-2006 09:58PM

watty wrote:

Thus it could be argued that DTT in UHF replacing analog TV is silly unless MPEG4 or reduced quality DVB-h.

France uses MPEG4 for its Pay-TV Channels on DTT, with its FTA channels using MPEG2.

Unless someone else knows better, the rest of the world uses MPEG2 encoding for its DTT systems both FTA and Pay,allowing for various transmitting systems like DVB-T, 8-VSB etc.

watty · 04-01-2006 10:08PM

The variant of MPEG4 best suited for radio transmission is not quite the same as MPEG4 computer files. The chips are only out last month to the setbox makers, so any MPEG4 in French DTT must be an earlier type.

Since the chips are only released now, the opertunity is for new DTT systems to be all MPEG4. Obviously older DTT will all be MPEG2.

The chipsets will do MPEG2 and manage Disk transfers so new boxes will work on older networks and can be designed with or without hard drives etc.

The norm in Europe for terrestrial Digital is COFDM, which uses multiple slower symbol rate carriers (up to several thousand carriers). This allows multipath signals to enhance the S/N rather than degrade the signal. Strangely the USA is not keen on COFDM.

I could be wrong but I think that sadly Digital MMDS seems to just use cable TV type modulation, a single carrier with QAM, so it is more interference prone and can have huge error rate if there is Multipath (Ghost on Analog). I guess it lets then use a Cable TV set box with the ordinary MMDS downconvertor used for Analog.

Metro uses single Carrier QAM but at 10Ghz with array antenna, interference and multipath is not an issue as you need almost optical LOS anyway!

digitalradiotec · 04-01-2006 10:45PM

Northern Correspondent wrote:

You're trying to read my mind, and you're failing miserably.

And you have the cheek to say this about me?:

"Also on newsgroups and forums he comes across as a complete **** especially if you dare say something that is contary to his thoughts."

The main reason why I would like to see RTE take forward these DAB tests (though it'll be no skin off my nose if it fails) is that it would finally establish RTE to take on a digital broadcasting setup of its own without having to rely on Sky or cable as its partners. We haven't a DTT network either though tests are again expected soon. The Irish Republic is currently the poor man of Western Europe regarding digital technologies

You say the Irish Republic is the poor man of Western Europe, and yet you actually want a test of the worst possible digital radio system to succeed?? Why?

but the past has shown that the Irish public is reasonably quick to take up on new technology while still maintaining a back-up redundancy, the most obvious being 405 line vs. 625 line TV transmissions back in the early 1960s hence the phrase "Irish Solution to an Irish problem". Local radio in the Republic, which accounts for a large amount of listenership, cannot be accomodated on VHF Band III DAB within current allocations. My prediction would be that these stations will bypass DAB altogether and appear on DRM / DRM+ when the right time comes. DAB, if RTÉ decides to roll it out nationally, will probably end up as 405 line TV in its time, making it available to those who want it but not give too much publicity.

I also don't really see any reason why at this stage using AAC codecs on DAB is worthwhile - existing DAB receivers would become obselete, which I believe in the UK stands at around 2 million.

Nobody's going to make 2 million receivers that they've just bought obsolete. What Ofcom said is that other countries are probably going to adopt DABv2 and it would be better if all DAB receivers were v2-enabled so that the UK can change over to v2 when the vast majority of DAB receivers were v2-enabled.

Ireland has the choice of adopting a completely out-of-date system now, and which is in the process of being replaced, or waiting a very short time and adopting a far more future-proof system.

The UK government after WW2 decided to keep the 405 line system going even with only a few thousand sets going in London even with a better spec 625 line system being made available across contentinal Europe. As I see it in the UK, the current system for DAB is past the point of no return. If AAC was to be implemented, it should have been done five years ago at least. It's too late now.

It is too late for the UK. But where else is it too late? Denmark, and that is it, everybody else can use DABv2.

Even if AAC was tried now it still wouldn't solve the problems that DAB inherently has with its fixed bandwidth properties. As it stands, the 128kbps MP2 DAB stations range in quality from tolerable to shocking - I can listen to the BBC stations though a set of headphones but not the Score NI stations. If AAC/HE-AAC is going to be thrown in somewhere it might as well be done with something like DRM which gives the flexibility for national, regional, local and community radio. For DAB as it stands , Ofcom has been too toothless to try and ensure a minimum bitrate standard and letting commercial interests beat them down.

AAC/HE AAC is going to happen with DABv2, so why on earth would you want to use DAB with MP2?

lawhec · 05-01-2006 12:42AM

digitalradiotec wrote:

And you have the cheek to say this about me?:

"Also on newsgroups and forums he comes across as a complete **** especially if you dare say something that is contary to his thoughts."

You are the one who was trying to put words into my mouth. You are trying to make people to agree with what you think without allowing the right to disagree.

You say the Irish Republic is the poor man of Western Europe, and yet you actually want a test of the worst possible digital radio system to succeed?? Why?

I've already explained.

Nobody's going to make 2 million receivers that they've just bought obsolete. What Ofcom said is that other countries are probably going to adopt DABv2 and it would be better if all DAB receivers were v2-enabled so that the UK can change over to v2 when the vast majority of DAB receivers were v2-enabled.

Ireland has the choice of adopting a completely out-of-date system now, and which is in the process of being replaced, or waiting a very short time and adopting a far more future-proof system.

It is too late for the UK. But where else is it too late? Denmark, and that is it, everybody else can use DABv2.

Ireland as a general rule of thumb tends to adopt the same technolgies that the UK follows on with, in a similar fashion to Mexico and Canada to the USA. It adopted the use of 405 line broadcasting (for part of the country anyway, including Dublin) around 15 years after 625 line broadcasts started on the continent which ironically the Irish Republic also adopted. Also DABv2, I don't believe has been rolled out anywhere yet. RoI is too small of a country to really go out and be the pioneer of a new broadcast system. Some of the larger European countries might, Australia, but I don't see Ireland doing so. Also RTÉ didn't plan to resume DAB test broadcasts until "receivers became affordable", now while receivers in the UK are not in general cheap, they are affordable for many people. Why RTÉ decided to resume the tests that are now being transmitted, I don't know, but the previous reason is the most likely answer. Certainly the announcement came out of the blue from RTÉ considering many of us are still waiting for the DTT trial to commence from the same transmission sites.

AAC/HE AAC is going to happen with DABv2, so why on earth would you want to use DAB with MP2?

Personally I do not see the point in adopting a DAB v2 where the only major change is adopting AAC as the audio codec. It's a bit like adding 256MB of RAM onto a 486 PC, sure it'll do good but really what's the point? The extra memory will not make up for the shortfalls of the PC in the same way that AAC will not make up for the lack of transmission flexibility that is inherient with DAB. If other European countries do adopt the DABv2 standard then there will be pressure on the UK to seek a gradual switchover with dual codec receivers. And what can be got on sale in the UK and usually be got in the Irish Republic as well, heck you can even buy DAB receivers in places where there is no hope of a signal! And when that happens, if a DABv1 system does apply in Ireland on a permenant basis, the BCI or Comreg will be likely to make a similar mandate.

watty · 05-01-2006 09:43AM

There is no point to DAB v2 and I would be surprised if many adopt it. Exsiting receivers can't use it. If you are going to need new receivers then DVB-h and DRM (and Eutelsat's scheme to recycle inclined orbit TV satellites as mobile Radio ) all make more sense.

For portable radio "quality not important" the digital medium that is going to kill DAB in UK is DVB-h on mobile phones. Sure the TV will mostly be susbscription, but I bet a significant amount of free radio. How many DAB radio sets fit in your pocket?

DRM may bring a resurgence in MW, LW & SW listening with easy tuning and clear signals. But not a medium for Local Radio, more for National and International networks.

I'm not sure what RTE is trying to achieve with its DTT and DAB tests other than familarising themselves with the technology. It is mature technology with well documented advantages and problems.

The MW & LW remote monitoring by RTE outside Ireland (co-Channel with Algeria etc) is more on line of real tests. This monitoring is preparation for DRM, which unlike DTT and DAB apparently is definately going to happen.

BTW, Homebase in Limerick still has huge stock of DAB Radio sets, in perhaps four models. Maplin and PC World have DAB and DTT cards / USB adaptors for PCs in Limerick too. Yet for years Argos wouldn't sell UHF only TVs in its UK catalogue here (Even though without Cable/MMDS they are fine in Limerick).

A discussion on digital radio technology

Comments