Ofcom Want to Permanently Allow Low Audio Quality on DAB

Ofcom have proposed (in the main consultation document page 145) to remove the minimum bit rate levels that are currently in-place and favour a scheme whereby the audio quality should not be degraded by MPEG encoding by more than a certain amount. That sounds fair enough until you read exactly how much they propose the audio quality can be degraded to:

“the technical quality of received audio should not be degraded by the process of coding and transmission by more than two ‘diffgrade’ points on the scale defined within ITU-R recommendation BS. 1116”

Diff-grade values are the results of blind listening tests, and are an average measure of the audio quality for a given codec at a given bit rate (the ‘diff’ part of the term refers to the fact that it is the difference between the encoded versions and the CD-original versions that are measured).
The following table shows the measured diff-grades for various audio codecs at different bit rates, plus my own estimates of some analogue technologies such as FM and cassette tapes:

* my own estimates

As you can see from the table above, a diff-grade of –2.0 is setting the bar for audio quality to extremely low levels, and I do not see how such a decision could possibly be in the best interests of the general public. Moreover, such a proposal makes an absolute mockery of Ofcom's duty (from page 145 of main consultation document):

• to include appropriate conditions for securing that signals carrying radio multiplexes (including their ‘payload’ of sound channels) should attain high standards of technical quality (Broadcasting Act 1996).

because audio encoded with a diff-grade of -2.0 can not be described as being anywhere near having a "high standard of technical quality" -- nothing could be further from the truth.

After having spoken to some quite senior people in Ofcom who are in favour of this proposal, they said that the proposal was based on the fact that the general public wanted more stations. But let’s be clear about what this proposal would actually mean for, say, a DAB radio multiplex. This proposal is specifically designed to allow the commercial radio groups to lower the bit rates of their stereo stations to 112 kbps if they invest in new, slightly better, MPEG encoders, in order to add new stations to the DAB radio multiplexes. But in order to fit just one more 112 kbps stereo station onto a DAB multiplex then they need to reduce the bit rates of 7 x 128 kbps stations to 112 kbps. How on earth does that benefit the general public? Obviously, far more people will listen to the 7 stations whose bit rates would be cut than would listen to the one new station, and all the main analogue stations are already on DAB, so it is not as if they would be adding new stations that are likely to be particularly popular.
Ofcom frequently claim to be “fact-based”, and yet their proposal to allow such low audio quality on digital radio just in order to allow one more stereo station per DAB multiplex is in direct contrast to the findings from a MORI market research survey that Ofcom commissioned (from Appendix B of consultation documents, page 55):

The above facts speak for themselves, and Ofcom's proposal to allow the commercial radio groups to reduce the bit rates of 7 stereo stations in order to fit one more station on at the same low audio quality are the absolute opposite of what is in the best interests of the general public.

Ofcom's motivation for proposing this change becomes obvious when you read what they say in Appendix E, page 28:

"Developments in coders tested to date would allow about a 10% effective increase in capacity, which equates to a benefit of up to around £4-5 million/year based on access revenues alone."

In other words, they deem it to be acceptable to for the commercial radio groups to provide low audio quality in order to allow the multiplex operators (all of which are owned by the commercial radio groups themselves) to make an additional £4 - 5 million revenue per annum. I would suggest that any regulator that proposes actions such as these solely to benefit the big companies at the expense of the general public is morally corrupt. And given the findings of the MORI market research above, then Ofcom's argument that adding one new station per multiplex benefits the general public more than allowing improvements in encoder technology to be translated into improved audio quality is not just flimsy, it is completely and utterly wrong.

I also found out from speaking to people at Ofcom that it was employees from Ofcom, the BBC and the commercial radio groups that have carried out a blind listening test which has provided the diff-grade values that Ofcom are going off – although they have not published the results of this listening test. But when you consider how these blind listening tests are carried out, then it is incredible that Ofcom would allow the broadcasters to take part. To use an analogy, this is like letting Alex Ferguson referee a Manchester United football game and expect him to be impartial! The broadcasters – especially the commercial broadcasters -- have a vested interest in these results being favourable, because all along GWR and Emap especially have never hidden the fact that they wanted the minimum bit rates for DAB to be lowered:

• Nick Piggott of GWR made the following comment about DAB bit rate levels at the Radio Academy Conference 2002: “We may make the case to go down lower again if we can show that the audience can’t tell”, and there’s no “may” about it, he will have made the case for the bit rates to be lowered.
• Someone from Emap admitted to me on the phone that Emap wanted the minimum bit rates for stereo stations to be reduced to 96 kbps on DAB.

All it would take for the broadcasters to get favourable results from this “blind” listening test is just to give all the samples listened-to high marks, or higher marks than they deserve. And at the end of the day, it is these exact people that have deemed the audio quality on DAB to be acceptable at the low bit rates from day one, and that goes for the people that will have taken part from Ofcom, too.
If diff-grades are to be used for regulation instead of minimum bit rates, then experienced and respected people should undertake such blind listening tests so that those with vested interests can not influence the results. Such a person would be Gilbert Soulodre from the Communications Research Centre in Canada, because he is respected; he has experience of undertaking such listening tests on the MPEG Layer II codec as used on DAB and he has no vested interest in the results.

Another issue with Ofcom’s proposal is that they only plan to regulate the actual degradation to the audio caused by the MPEG encoding process (the transmission part they refer to is irrelevant to audio quality). This means that the audio going into the MPEG encoders can be of extremely poor quality to begin with – as it already is now in many cases of DAB stations – but because the bit rate used only apparently degrades the audio by 2 diff-grade points then that would be perfectly acceptable within these proposals.

Another issue of concern is Ofcom's belief that because market research conducted by the DRDB suggested that most people are happy with the sound quality on DAB then that means that there is no issue with the sound quality on DAB -- nothing could be further from the truth. There are the following problems with their views on this subject:

• Any positive market research results from DAB owners can easily be explained by the fact that DAB owners almost universally confuse the elimination of hiss with sound quality (for examples, see comments from DAB owners in the MORI market research in Appendix B). But the elimination of hiss applies to all bit rates on DAB, and is therefore 100% independent of audio quality. Therefore, all market research that doesn't differentiate between the lack of hiss and audio quality should be neglected in any decision-making with respect to the setting of minimum audio quality levels on DAB. And to date, I am not aware of a single piece of market research that has been conducted that differentiates between the lack of hiss and audio quality, therefore no decisions can be taken at present with regards to the adequacy of the audio quality on DAB.
• Somewhere in the region of 95% or more of all DAB products sold are small portable radios, and therefore the cross-section of devices used for radio listening via DAB bears absolutely no resemblance to the cross-section of devices used for radio listening via FM. An EBU document quoted that 60% of all radio listening takes place at home, 20% in the workplace, and 15% in vehicles. If only current market research is taken into consideration then the views of two large groups of radio listeners are almost totally excluded: Younger people very rarely buy small portable radios and are far more likely to listen via micro hi-fi systems and DAB micro systems have only very recently come on the market, so this whole group would basically be excluded from the market research; and anybody that listens to radio in vehicles (15% of all radio listening) would also be excluded due to the very low number of DAB car stereos sold to date.
• The sound quality on small portable radios is almost irrelevant when you are supposed to be considering the sound quality of DAB in general, because such small portable radios cannot produce good audio quality, and mask problems with the audio quality of the DAB signal.

Any decisions made about whether the current audio quality on DAB is adequate or not should be suspended until market research can be carried out that is representative of the cross-section and proportions of receiver-types typically used for listening to radio on, because any actions taken now cannot be reversed, because Ofcom will not instruct the commercial radio groups to increase their bit rates once they've been reduced.

 
 

Add a comment: Name (optional) Email (will not be published) (required) Subject (optional) Write the word radio in this box (HTML markup is allowed)

 
 

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