Quality of BBC Internet radio streams to overtake DAB

The BBC has finally said that it's going to start using new audio formats for its live and on-demand Internet radio streams in July and April respectively, which should mean that the audio quality of the Internet radio streams will overtake the quality of the BBC's stations on DAB. The reason I say this is that Real Player supports both the AAC and AAC+ audio codecs (official name of AAC+ is HE-AAC, which stands for High-Efficiency AAC), and the BBC's Real streams are currently using 64 kbps, so the minimum you would expect them to use is 64 kbps AAC+ for the streams, and as the following quote from a document published by the WorldDMB Forum (the organisation in charge of DAB around the world) shows, 64 kbps AAC+ provides higher audio quality than 128 kbps MP2, which is what 98% of stereo stations on DAB in the UK use:
 

because 64 kbps AAC+ obviously provides significantly higher quality than 40 kbps AAC+.

However, I think they should only use 64 kbps AAC+ for delivery to mobile phones, and they should use 128 kbps AAC for the streams to stationary listeners, for the reasons given in the section on Ultra-low bit rate levels below -- basically, the cost of Internet bandwidth to the BBC has been falling in-line with Moore's Law, i.e. it's been halving every 18 months, so I think it's about time that the increased the bit rates of their radio streams to decent levels.

In the same blog entry, the BBC also said we "will be getting" the multicast radio streams. The BBC has been trialing multicast for several years, and the multicast streams were due to launch last year, so it now seems that they will be going live in time for Virgin Media's cable network supporting multicast later this year -- at the same time as when Virgin is going to launch their 50 Mbps broadband package. Multicast streams will always provide higher audio quality than DAB+, let alone DAB. Multicast is extremely bandwidth-efficient both for the broadcasters and the ISPs, so the multicast streams will always use higher bit rates than on DAB+, and the multicast streams are using modern audio codecs, so the audio quality will be higher. For instance the BBC has been using 128 kbps AAC for its trial multicast streams, and more recently some of the commercial radio groups have been using 192 kbps WMA for theirs, and the bit rates of the multicast streams are only likely to go up over time.

HOWEVER, and that is a deliberately big however, before we go patting the BBC on the back for what they're about to do, I'm sure that most people reading this will at least have listened to the BBC's Internet radio streams at some point over the last few years prior to the BBC increasing the bit rate levels of the Real streams to 64 kbps last September. And I'm sure those who did hear them would agree that they sounded utterly attrocious -- some of the Listen Again streams still sound utterly attrocious to this day. Well, it turns out that they've had the option of using AAC+ on the Real Player radio streams for over four years now, and they've also been butchering the audio at the same time, so there's been absolutely no reason for them to have sounded anywhere near as bad as they have done for a long time now, so all they're doing now are things that they should have done a long, long time ago, so in my opinion they don't deserve a pat on the back, they deserve a good clip round the earhole -- or talking or earholes, perhaps it would be better to punish them by locking them in a Guantanamo-style room and pipe the BBC Internet radio streams into the cells at 32 kbps so that they .

Also, as I'll explain in the last section of this article, the reasons why I think this ridiculous level of inaction has happened has been both due to the BBC executives who make the decisions not understanding the technologies that they're making decisions about, because they don't have a technical background, and also due to the BBC's massive bias towards DAB, and I think at least part of why they haven't improved the quality before now is simply that providing them at poor audio quality gives DAB a helping hand, because the audio quality of the streams has been so awful that they will have put a hell of a lot of people off wanting to listen via the Internet, which obviously increases the likelihood that people will buy DAB instead.

They've had the option of using AAC+ since 2004

Real Networks added support for the AAC+ audio codec to its Real Player 10 software in January 2004, so the BBC has had the option of using the most efficient audio codec available today, which performs head and shoulders above all other codecs at very low bit rate levels, and the BBC just happens to have been using extremely low bit rate levels for its Internet radio streams since they first launched -- they were using a bit rate of just 32 kbps for Radios 1, 2 and 3 until last year, and they're currently using 64 kbps. The AAC+ codec has therefore been the ideal audio codec to use ever since Real added support for it, and yet it looks like we're only now going to get it -- or they'll use AAC if they decide to increase the bit rate levels as well.

I rang up Via Licensing (which is the company that looks after the collection of licence fees for the use of various MPEG-4 audio and video codecs) a few weeks ago to check whether the BBC would have to pay the licensing costs associated with the use of AAC+ (official name HE-AAC), and it turns out that the BBC wouldn't have to pay a penny to use AAC+, because what matters is where the "decoding event" takes place, and the decoding for the BBC Radio Player is performed by the Real Player software in the "back-end", so Real would be responsible for paying the licensing costs, not the BBC. So there's been absolutely nothing stopping the BBC from using AAC+ whatsoever for over four years. And the person I spoke to at Via Licensing told me that the BBC had actually contacted him a few months earlier to find out how much it would cost them to use AAC+ -- which obviously suggests that they're now going to be using it. But why on earth has it taken them almost four years to realise that the most efficient audio codec on the planet was sitting there under their noses ready and waiting to be used when the audio quality of the radio streams has been so utterly diabolical? I find it completely incomprehensible that the people in charge of digital radio at the BBC could go four years providing such dire audio quality without even taking the most basic of steps to investigate how to improve things. At the very least it shows an utter contempt for the way the BBC has viewed the Internet radio streams, because there's no way on this earth they would have gone four years using the wrong codec on DAB -- if DAB had the choice of using a different codec, which it doesn't without changing to using DAB+, of course. So the very least that can be said is that they were hardly encouraging people to listen to the online streams.

AAC+ vs Real G2

The audio codec that the BBC has been using for its Internet streams all these years is called 'Real G2', and it's interesting to compare the audio quality performance of Real G2 with AAC+ using results from a listening test that engineers from the BBC's own R&D department took part in in 2003 (which had been co-ordinated by the EBU (European Broadcasting Union)), because it shows that people at the BBC have known all along not only that they've been using a bit rate/codec combination that guarantees that the audio quality will be absolutely terrible, but also that AAC+ performs far, far better than Real G2 does -- I'm not blaming the engineers in the BBC R&D department for this; I'm laying the blame for this squarely at the feet of the BBC executives who make the decisions, because all they had to do was ring up the R&D department to ask them if anything could be done to improve the quality, but clearly that was too much trouble.

The figure on the left below shows the results for the Real G2 audio codec at bit rate levels between 16 kbps and 64 kbps ('S' means stereo, 'M' means mono), and the figure on the right shows the test results for all codecs that were tested at a bit rate of 48 kbps (Real G2 is referred to as 'RL'). The BBC was using the Real G2 codec at a bit rate of 32 kbps for Radios 1, 2 and 3 until the first half of last year, and that was rated as providing audio quality that was just above being "bad" -- and this was on the most lax kind of listening test it's possible to use, because it is used for testing bit rates/codecs that provide "intermediate quality" (which is a polite way of saying "crap quality"), and in my opinion this test overestimates how good the quality is! The AAC+ codec at 48 kbps, on the other hand, is on the cusp between being rated as providing "good" and "excellent" quality (for some reason, AAC+ was only tested at 48 kbps).

1 - The listening tests were using the 'MUSHRA' testing methodology, which is meant to be used for 'intermediate-quality' audio codecs (listening tests for "high quality" use the BS.1116 listening test methodology, which is the "gold standard" for such tests), because it was originally designed to test Internet radio streams in the late 1990s when the bit rate levels were extremely low -- 20 kbps and the like, so it is actually very easy to achieve a high score on the MUSHRA test, and the ratings given overestimate how good the quality actually is.

Ultra-low bit rate levels

Internet bandwidth costs have been plummetting for content providers

Before I look at the bit rate levels the BBC has been using for its Internet radio streams, it's worth looking at how the cost of Internet bandwidth has varied over time for the BBC. One thing you have to consider with the BBC is that because they run the third-busiest website in the UK, with 16 million unique users per week, to save paying another company to provide them with bandwidth, the BBC basically runs its own ISP, and normal ISPs -- the kind we pay every month -- "peer" with the BBC so that the ISPs save on costs by exchanging data directly with the BBC at Internet Exchange Points (IXP), such as LINX in London. So to look at how some of the costs of running a high-traffic website have changed over time, the following table uses information from here that shows the relative costs of running a web server in 1995 compared with 2005:

1 - Quoting from the araticle: "In addition to the price reductions, we also have to look at the compute performance of a web server class machine in 1995 vs. today. Given five or six performance doublings since 1995 courtesy of improvements in clock speed, bus speed, architecture changes from 32 to 64 bit, additional cache memory and faster RAM, a conservative estimate would be that today's single CPU 1-U "pizza box" web server is roughly fifty times faster than last decade's model. Couple that with the 25x price difference for this pizza box, and your 2005 dollar buys you more than one-thousand times as much compute power as it did in 1995."

The "cage space" can be ignored because the overall cost will be tiny compared to everything else; and the disk storage and effective overall computing power have gone up by a factor of about 400 and 1,000 respectively, so the "limiting factor" is the bandwidth costs, and they've fallen almost exactly in line with Moore's Law -- i.e. they've halved about every 18 months! Bandwidth costs following Moore's Law isn't all that surprising really if you look at the following figure (slide 49 from this), which shows how Internet router speeds have changed since the mid 1980s, because a chain is only as strong as its weakest link, and optic fibres can handle effectively anything you throw at them (the theoretical maximum bandwidth that a single optic fibre can handle is 25 Tbps, or 25,000 Gbps), so the weakest link for Internet bandwidths is the speed of Internet routers, which have increased in-line with Moore's Law since the mid 1980s. So assuming the cost of a router stays constant (you'd really expect the cost to go down rather than up, if anything), the cost of Internet bandwidth will fall in-line with Moore's Law.

The BBC was using 32 kbps for Radios 1, 2 and 3 until last year

Despite the cost of bandwidth halving every 18 months, the BBC actually reduced the bit rate levels they were using for their Internet radio streams for Radios 1, 2 and 3 from 64 kbps to 32 kbps in 2003! And not only that, but they didn't return to being 64 kbps until September 2007.

I don't know exactly when the BBC originally started providing its Internet radio streams at 64 kbps, but I'd guess it was around the start of this decade when broadband was just in the early stages of taking off (there was little point in providing 64 kbps radio streams when everyone was using 56k dial-up modems...). So let's just say that they launched the 64 kbps streams on 1st January 2000, that would mean that there have been 5.44 iterations of Moore's Law between then and the beginning of March 2008, so the cost of bandwidth would have gone down by a factor of 2^5.44 = 43.5. How on earth can they justify still using 64 kbps for the radio streams when the cost of bandwidth has gone down by that much?

The fact that bandwidth has become as dirt cheap as it is also explains why the BBC has been able to lavish bit rate levels of 550 kbps on the BBC iPlayer TV streams -- 17-times higher than the 32 kbps used for Radios 1, 2 and 3 -- and the iPlayer has had up to 660,000 users per day in its first two months since launch, and the number of users, and hence the bandwidth required for the service, will increase exponentially over the next few years.

And the fact that bandwidth is now dirt cheap, the bit rates that they're currently using are obviously far too low (these bit rates were what they were using about three weeks ago, but I'd imagine they'll be the same today):

Because the cost of bandwidth falls with Moore's Law, the BBC should increase the bit rate levels to 128 kbps when they start using new audio formats. And when using a bit rate level of 128 kbps, it is better to use AAC rather than AAC+ -- AAC+ is meant to be used at bit rate levels below about 85 kbps, and AAC provides better quality for bit rates above that level. It is completely unjustifiable that we've seen so many iterations of Moore's Law and yet we still haven't seen the bit rates of the radio streams rise from 64 kbps. Furthermore, the BBC is using 400 kbps H.264 for the video and 116 kbps AAC for the audio for the iPlayer TV streams that iPhone users can receive, so I fail to see why iPhone users get better quality than someone who might want to listen to the Internet radio streams on a hi-fi system or who have forked out specifically on a Wi-Fi Internet radio. And the last time I looked there were 5,770 Internet radio streams using bit rate levels of 128 kbps or higher on the Shoutcast Internet radio portal, and GCap has been providing 128 kbps WMA streams for its stations for over a year now and they're at a far higher quality than the same stations are at on DAB. So if these other broadcasters can provide 128 kbps, the BBC?

The BBC has been butchering the audio for the Internet radio streams

Last but by no means least, the BBC has been 'transcoding' the audio for its Internet radio streams. Transcoding means that compressed audio is decoded and then it's re-compressed to a different format. In the case of the BBC's Internet radio streams, they've been receiving the BBC's stations off-air via digital satellite at Maidenhead, where the Internet radio servers are located, and the audio from the satellite receivers is then re-compressed to the Real G2 format. Transcoding is very poor audio engineering practice, and the lower the bit rate you're re-compressing to the worse the effect of transcoding becomes, so it's been particularly important not to do this for the BBC's Internet radio streams when they've been using such low bit rates as 32 kbps and 64 kbps. And similar to the case where people from the BBC already knew that AAC+ vastly outperformed the Real G2 codec, people from the BBC R&D department published a white paper on the issue of transcoding -- although it's common knowledge that transcoding degrades the quality anyway. Again, I don't hold the BBC R&D engineers responsible for this happening, because they're not the ones controlling the purse strings, because the reason it has been done the way it has has been down to saving money -- see below for how pathetically little they've saved each year.

The person in charge of the technical side of digital radio admitted that they've been doing it this way ever since the radio stream servers moved to being located at Maidenhead in 2002, so they've had six years to sort this out. Apparently "plans are afoot" to sort this problem out as well and it will be done in the next few months, but considering that they've spent £130m on developing the BBC iPlayer, why couldn't they have spent the tiny amount of money needed to do this years ago?

The UK population spent 25 million hours per week listening to the BBC's Internet radio streams in January this year, and asking around I've been told that it would have cost the BBC about £5,000 per annum to run a leased-line to get the pre-compressed Real G2 data to the servers in Maidenhead to avoid having to re-compress them. Or in other words, the BBC degrading the audio quality in the way that they have had saved the BBC about £5,000 per annum. So as the UK population spends 25 million hours per week listening to the BBC's Internet radio streams, this works out as being that the BBC saves £1 -- that's 100 glorious pennies -- for every 260,000 hours of listening displeasure. Presumably they needed the £5,000 per annum to put towards the £130m that they spent on developing the iPlayer? Or perhaps they needed it for the £9.6m they spend each year transmitting DAB, which is soon to go up to £14.6m per annum even though the 83% increase in costs for transmitting the BBC's national DAB multiplex (gonig from £6m to £11m per annum -- the other £3.6m is spent on transmitting local BBC stations on the commercial multiplexes) will only increase the population coverage from 86% to 90%! And the cost of extending their national DAB multiplex to provide the same 99% coverage level that FM provides would have cost £40m per annum, and they only realised last year that DAB had become "prohibitively expensive" to roll out to 99% of the population (see page 48 of this pdf for the figures), so they're now going to increase the coverage to 90% of the population, which they'd already committed to, and they're then going to stop there. The remaining 10% of the population -- who have no idea that they're never going to receive DAB and yet they're still watching the BBC's TV advertising campaigns telling them to go and buy DAB -- will have to wait for the BBC to investigate other technologies instead, such as DRM, satellite radio and bizarrely Wi-Fi was mentioned, which links nicely to the next section.

The problem with the BBC is that non-technical people are making technical decisions

The example given in the previous sentence is a perfect example of what is wrong with the way the BBC does things regarding technology. I've no doubt that Mark Friend, the current BBC Controller in charge of digital radio, is highly competent in his own field, whatever that may be, but his field of expertise is clearly not wireless digital communication technologies, because his suggestion that Wi-Fi is a potential technology that could be used to cover the remaining 10% of the population is about as far wide of the mark as you can possibly get. Wi-Fi is a short-range wireless technology that was not designed to be received when mobile, and it doesn't have 'handover' capabilities like you have on mobile phones, and which don't need to be supported on DAB due to the use of SFNs (single-frequency networks). So Wi-Fi is the opposite of what you'd need to cover the last 10% of the population, because you need something that can cover very wide areas as cheaply as possible -- with Wi-Fi you'd need another hotspot every 5km or so, and then you couldn't receive it in the car without retuning every 5km.

The people I blame for this situation arising aren't the likes of Mark Friend, but I blame the people right at the top of the BBC, because they arrogantly think that it's perfectly okay to allow non-technical people to make technical decisions, but I'm afraid that they cannot hope to fully comprehend the implications of the decisions they're making.

You would have thought they'd have learnt their lesson with DAB. The BBC chose to stick with DAB due to the current Director of Radio, Jenny Abramsky, and the ex-Controller in charge of digital radio, Simon Nelson, choosing to go ahead with it. Neither have a technical background -- Simon Nelson has a background in marketing, and Jenny Abramsky studied English at university according to her biog web page -- and it is purely down to the decision they made to adopt DAB that we're in the ridiculous mess that we're now in -- if the BBC didn't support DAB, then DAB would have been a non-starter, because commercial radio couldn't sell it on their own. But that means that whoever takes the decisions at the BBC needs to know exactly what they're doing, and I'm afraid that Jenny Abramsky and Simon Nelson couldn't have hoped to fully understand the implications of their decision.

The BBC R&D engineers were obviously saying long before DAB was properly launched in 2002 that there were better audio codecs available, because they wrote the following in a brochure for a BBC R&D open day in 1999:
 

"New audio coding systems (such as AAC) can halve the bit-rate"

 "Don't squeeze the bit-rate"

and engineers from the BBC R&D department had taken part in a multi-channel listening test in 1996, then a stereo test in 1998, both of which concluded that AAC was twice as efficient as MP2 -- i.e. it needs half the bit rate to provide the same level of audio quality.

Obviously, if Jenny Abramsky or Simon Nelson had asked the people who wrote those comments, they obviously wouldn't have said "yeah, Jenny and Simon, it's a fantastic idea to use MP2 when the bit rates are as low as 128 kbps, in fact you could reduce the bit rates a bit more because MP2 is such a great codec" -- for those that don't know, MP2 is the codec used on DAB, and it was designed to be used at bit rate levels of 192 kbps, and any audio codec operated at bit rates well below what it was designed to be used at will inevitably provide poor audio quality, hence the poor audio quality on DAB. Using 128 kbps MP2 on DAB is like encoding your own audio to 64 - 80 kbps MP3.

Basically, engineers should make engineering decisions, and for as long as the executives continue to arrogantly think that non-technical people can make technical decisions then we will continue to see major mistakes being made, such as those that I've described above on the Internet radio streams and as were made when the BBC incompetently chose to adopt DAB and they landed us in the ridiculous mess we're now in.

Simon Nelson's reign of audio terror in charge of digital radio at the BBC

Put very simply, Simon Nelson was ridiculously biased towards DAB, and his bias towards DAB seemed to govern every decision he made when he was in charge of digital radio at the BBC, and it's because of the ridiculously biased way he acted that it wouldn't surprise me in the slightest if he actually deliberately decided to keep the audio quality of the Internet radio streams low in order to deter people from listening to them so that more people would buy DAB instead.

One candid admission he made that speaks volumes of his attitude was on Radio 4 Feedback when he said: "of course the BBC would prefer it if everybody listened to digital radio via DAB". Having said that, it's hardly a surprise that the BBC is heavily biased towards DAB at the expense of all other digital platforms that carry radio, because the BBC has broadcast 20 high-impact TV advertising campaigns for DAB (which I worked out would have cost £163m if they'd have been shown on commercial TV; so the BBC has been pseudo-subsidising each one of the 6.5m DAB receivers to the tune of £25 each!), whereas they've broadcast zero TV adverts let alone any ad campaigns for their Internet radio streams.

The crucial thing to consider with Simon Nelson is that he was the "artchitect" of "BBC DAB": he was the Head of Strategy when the plans were being formed to launch DAB, and he went on to be the Head of New Services Development, so he was in charge of the overall strategy as well as which stations to launch on DAB and why to launch them -- so it was Simon Nelson that decided to lower the audio quality to the level it's now at.

Apart from his extreme bias towards DAB, the main problem with Simon Nelson was that he thought nothing of being dishonest when dealing with the general public as a whole. For example, in an email that someone forwarded to me, he admitted that he had chosen not to tell the public in the public consutlation for the five new digital-only stations in 2000 that the result of the statinos being added to the BBC's already almost full national multiplex would be that the audio quality of the existing statinos would be drastically degraded. That is about as crucial a piece of information as you could possibly think of considering that the stations whose audio quality would be affected were Radios 1-5, and if the public had been told that the audio quality of the stations that they listen to on analogue radio would be massively deraded then the feedback the BBC would have received would obviously have been far worse than it was, and that would have made it far less likely that all five new stations would have been launched. As it happened, the feedback was hardly positive anyway, because 1Xtra and the Asian Network both received less than 50% acceptance figures, and only BBC7 received a big thumbs up from the public. But all five digital stations were launched anyway. He basically dishonestly withheld information that was crucial to the public being able to make their own minds up. It's as simple as that.

Another example of him being dishonest was whenever he appeared on Radio 4's Feedback programme to answer complaints about DAB's audio quality. One example of this consisted of when he was asked why Radio 4 on DAB was sometimes broadcast in mono in the evenings, to which he replied "yes, that does happen from time to time, erm, it's because the programme has been made in mono" -- if you listen closely, you can actually hear the cogs turning round in his head and they're whispering "how can I get out of this when there's millions of Radio 4 listeners listening?" But rather than doing the right thing and fessing up, Simon took his default route of being dishonest to the millions of Radio 4 listeners instead. The truthful answer to that question is as follows: Radio 4 on DAB is always reduced to mono whenever Radio 5 Sports Extra is on-air after 5pm (and Radio 3's bit rate is reduced to 160 kbps before 5pm when Radio 5 Sports Extra goes on-air). This is due to Simon Nelson choosing to cram so many stations onto the BBC's national DAB multiplex that whenever Radio 5 Sports Extra starts transmitting other radio stations have to reduce their bit rates in order to accommodate it. Therefore, programmes that are made in stereo for Radio 4 but which are broadcast when Radio 4 is in mono no DAB are still broadcast in stereo on Radio 4 on FM and on the digital TV platforms. Someone in the DAB industry suggested to me that "he might not have known" that programmes made in stereo were broadcast in mono on DAB. But he'd just spent the last 3 or 4 years working on the plan to launch five new stations, and he was the Controller in charge of digital radio at the time, so he will have made the decision to reduce Radio 4 to mono on DAB in the evenings, and he will have weighed up the possibilities. I think the chances of him not knowing that programmes made in stereo are broadcast in mono on DAB is so vanishingly small that it's not even worth considering. Basically, he told a porkie pie.

Also, right at the end of one of his other appearances on Radio 4 Feedback, after numerous listeners letters had been read out which were virtually all complaining about the audio quality on DAB, he had the gall to say the following, in a manner befitting Tony Blair when he was trying his hardest to make persuade people to believe him: "“And the message we need to get across is that for the vast majority of people, the sound quality they're getting through their digital radio set is vastly superior to that which they've ever had through their analogue radio.” The reality is of course that the vast majority of people are getting vastly inferior audio quality from DAB.

And as discussed above, Simon Nelson was the person that took the decision to transcode the audio for the BBC's Internet radio streams by receiving the stations off-air via digital satellite, and he took the decision to reduce the bit rates of the Internet radio streams as well.

I also suggested that the BBC should increase the bit rates for the radio stations on digital satellite, because the BBC has got 231 Mbps of capacity on digital satellite, of which only 0.7% is consumed by the radio stations, so it would be easy to increase the bit rate levels so that they provide high quality -- they are at a higher quality than DAB, but they still could easily have increased the bit rates further. The technical journalist Barry Fox also wrote in one of his columns that he thought doing this was a good idea, and there was also a letter published suggesting that this should be done in the magazine for people who've retired from the BBC. The BBC replied to that letter -- which will no doubt have been written by Simon Nelson -- by saying that "the BBC hasn't received any complaints about the audio quality of the radio stations on satellite", and that was that. Similarly, the BBC refused to increase the bit rates for the radio stations on Freeview, albeit that there is less free capacity on Freeview. But the BBC Trust published a consultation last year considering options for adding the BBC HD channel to Freeview, which included plans to add it to their existing multiplexes, so the BBC could find about 10 Mbps of capacity to fit the BBC HD channel onto their Freeview multiplexes, and yet they couldn't even find 200 - 300 kbps to increase the bit rates of the radio stations.

I also said that the BBC should mention in their numerous high-imapct TV adverts for DAB that digital radio was also available on platforms other than DAB, because some people would prefer to listen via these other platforms, but of course Simon Nelson wouldn't have that, and all of the TV adverts only mentioned DAB.

Simon Nelson also wrote in the BBC's submission to an Ofcom consultation on digital radio that he wanted the UK to be "an all-DAB country", which in the context meant that he didn't want DRM or DRM+ to be used at all, despite the fact that they're far cheaper to transmit than DAB. Therefore, Simon Nelson was in favour of spending the £40m per annum needed to transmit DAB to 99% of the population, as discussed in the section above. To say that this is poor value-for-money when covering 86% of the population only costs £6m per annum goes without saying, and again it shows his massive bias towards DAB.

I also think it speaks volumes that just a few months after Simon Nelson had moved on to another job in the BBC, his successor increased the bit rates of the Internet radio streams, because his successor, Mark Friend, obviously thought that the quality wasn't anywhere near good enough -- Mark Friend has doubled the bit rates used for Radios 1, 2 and 3 since he took over (although as discussed above, due to the dirt-cheap cost of bandwidth, they should be higher than the 64 kbps they are using today). Also, over the last 3 - 4 months, the BBC has finally started broadcating TV ad campaigns promoting the fact that their radio stations are available on the digital TV platforms, and the adverts don't even mention DAB, which is something that would never have happened when Simon Nelson was the Controller.

The BBC promoted Simon Nelson last year, which either speaks volumes about the BBC as an organisation in that they're willing to reward dishonesty, or that it shows that the BBC management above Simon Nelson are so lax that they're unaware of just how dishonest he is.

Comments

Bookmark with:
Digg	del.icio.us	Reddit	Facebook
Google	Stumbleupon	Slashdot