Developing a simple petition for improved bitrates and increased spectrum allocation for DAB+ broadcasting in Australia

[MLXXX]

10 hours ago, alanh said:

If you read the link, the ages of the subjects is quoted. As I said you don't understand your google searches.

I'm sorry, alanh; I missed that part of the paper, I must confess. I see that on page 4 of the pdf, this explanation appears, "The test was performed on a group of  45 people, aged between 18 and 25 years old. None of them had hearing disorders.".

You stated earlier in this thread, "All the subjects were young, probably fellow students who are hardly average listeners. There age is well under the average Australian age.".

 A quick internet search tells me that in 2011 the average Australian was 37. 

We could take it that most if not all of the participants could be expected to have their hearing intact in the range 15kHz to approaching 20kHz, an upper range of hearing acuity that the average 37 year old might well not have, and something that people in their 70s would almost certainly not have.

I wonder whether that made the restriction in the upper limit of broadcast FM sound to 15kHz more noticeable than it might have been for an older group. That does seem distinctly possible.

And it could be that certain artefacts of HE-ACC were more noticeable to these young adults than to older adults, such as SBR.

I am not aware of, and have not attempted to search for, usual or recommended practice in choosing ages of participants in tests for absolute transparency of psychoacoustic codecs at high bitrates, or for degree of subjective impairment in MUSHRA style testing of codecs at medium and low bitrates. 

I am of the opinion though that a broadcasting system should be designed to give good quality sound for young adults and for older people.

 

11 hours ago, alanh said:

If the problems are as bad as you claim why can't you quote Australian examples, with broadcaster, date and time of transmission and a description of the sound and what is wrong with it.

You cannot hear any of these problems you keep quoting on syndicated Australian radio so how can you point out to an Australian broadcaster what is wrong with their digital sound.

 There is no likelyhood that any broadcasters will change their bit rate if you cannot demonstrate the problem with their received output on a set of studio quality speakers.

Have you complained to 4KQ yet?

Alanh, I think you have misconstrued some of my comments, I have not suggested that any of the broadcasts are faulty. As best I can tell from casual listening, they are what is normal and to be expected where a moderate to low bitrate is employed.

 In recent listening  I have found that unusually poor sound for the nominal bitrate is generally explicable in terms of the allocation of an unusually high PAD. Conversely, unusually good subjective sound quality despite a low nominal bitrate can be due to a minimal PAD or the use of HE-AAC v2 rather than v1. For my ears I prefer a sacrifice in stereo image at lower audio bitrates through the use of Parametric Stereo: a less nuanced stereo effect but combined with a better basic sound quality.   

Your accusation "You cannot hear any of these problems you keep quoting on syndicated Australian radio", is incorrect.

I have made no personal complaint to any of the DAB+ broadcasters in relation to bitrates. I think the petition envisaged by this thread, if and when it materialises, will be more effiective than a phone call, email or letter from me to a broadcaster.

Having said that, some years ago I did complain in writing to a broadcaster for their use on-air of the description "CD quality" for their DAB+ broadcast. I received an apology advising me that the matter had been brought to the attention of the particular announcer who had, in ignorance, made the incorrect claim about the sound quality of the digital broadcast.

[alanh]

 

You keep complaining that the audio bit rate is not high enough even in your last post.  In which programs, sounds and when are they is parametric stereo a problem? If parametric stereo is used the effect on the overall bit rate is tiny. I would rather have the stereo. Pity our broadcasters don't use stereo micing when there is more than one presenter.

Considering that Commercial Radio Australia hasn't used the "CD quality" in a very long time please drop this argument that there is insufficient data for the programs transmitted.

You still need to complain to 4KQ as they have a problem which is not evident on their AM program only their digital one.

So unless you can provide examples we can all assess please drop the poor sound quality tag. Remember that a lot of radios with speakers which are sold are more likely to have poor sound due to the speakers and the cabinet which contains them.

Alanh

[MLXXX]

On 4/7/2017 at 8:12 PM, Audiofile1503564314 said:

I think it's a good start... Empirical research is important... I think we'll need anecdotal reports as well as empirical research.

Hi Audiofile, just in case you or anyone else is wondering, I have by no means abandoned the idea of uploading some real life examples of DAB+ broadcasts. 

I have equipment for recording up to four analogue stereo sources simultaneously be they different model DAB+ radios using their line-out or headphone sockets , or FM or AM broadcasts. And, bypassing any analogue interface, I can capture a DAB+ digital transport stream using a USB dongle (RTL2832U).

I've done some preliminary testing comparing the subtle differences between different DAB+ radios but there is planning involved in not only deciding on which broadcasts to record with which radios, but how I might put the material together for a technically meaningful but easy to follow audio visual presentation on YouTube, possibly supplemented by direct links to uncompressed audio files. And then we would need a separate thread on this forum to give forum members the opportunity to comment. This would possibly be set up as a poll thread.

There is also the question of how many YouTube videos to prepare.  For example, it might make sense to to have one technical video comparing different DAB+ radios, another technical video comparing different FM or AM radios, and a third video inviting subjective comparisons for radio programs simulcast on DAB+ and FM or AM.  At this stage, my thinking is not to include any on-line streaming sources in the presentations.

[Pesto Lovin' Man]

2 hours ago, MLXXX said:

I've done some preliminary testing comparing the subtle differences between different DAB+ radios but there is planning involved in not only deciding on which broadcasts to record with which radios, but how I might put the material together for a technically meaningful but easy to follow

Keep it simple. If youre not trying to show differences between radios then go with transport stream decoded by a proper fp decoder (better than the integer decoders used by most chipsets) and the best FM receier you have recorded through the best ADC you can lay your hands on. If the station also streams for a 3rd sample then capture that without reencoding and if you can arrange to have all three be the exact same material recorded a tthe same time even better.

Leave your "audio visual presentation on YouTube" until after youve collected and analyzed results.

You wont convince the irrationally enthusiastic but you dont have to hand him arguments.

[MLXXX]

40 minutes ago, Pesto Lovin' Man said:

Keep it simple. If youre not trying to show differences between radios then go with transport stream decoded by a proper fp decoder (better than the integer decoders used by most chipsets) and the best FM receier you have recorded through the best ADC you can lay your hands on.

So by a proper decoder you'd mean something like this https://sourceforge.net/projects/aacstreamanalysis/ which uses faad2? Or is there something else that might somehow be better?

That also raises the question whether the FM should be captured by purely digital means, to keep a level playing field. I've found there are slight differences in the subjective sound quality between using a USB dongle for FM and using analogue capture of the output of the FM tuner in an AVR. And even with using a dongle and depending on the software used there are different settings for how to demodulate the FM signal, leading to very slight audible differences. I suspect I'll end up using a late model AVR FM tuner and an ADC. 

I have no problem with the ADC side of things. I use a Behringer Firepower FCA1616 Interface.  It's certainly good enough for analogue capture of FM radio and medium bitrate DAB+ radio. 

[Pesto Lovin' Man]

Sorry fp = floating point. Most digital radio chips & phone or tablet apps use fixed point i.e. integer math to save CPU or DSP cost & power. That means minor rounding errors etc & so the decompressed audio is a little less accurate. Most desktops programs use a floating point decoder library but on Windows Id pick the Fraunhofer FhG AAC encoder/decoder just to save arguments later. If you want a fixed point decoder to compare with then Fraunhofer FDK AAC is supposed to be good.

Id stick qith a good analog FM receiver for FM although software decoding should theoretically be better. Or Id do both if youve got the hardware to grab the simulcasts in parallel. I wouldnt bother doing an analog capture of DAB since thats adding the unknown quantity of the radio/chipset & you want to minimise those unknowns.

[MLXXX]

2 minutes ago, Pesto Lovin' Man said:

I wouldnt bother doing an analog capture of DAB since thats adding the unknown quantity of the radio/chipset & you want to minimise those unknowns.

Ordinarily I might agree but an enthusiastic member of this forum has suggested that new DAB+ radios produce noticeably superior sound to old DAB+ radios because of changes in the chips used (despite the fact that the relevant specifications for decoding HE-AAC remain unchanged).  This was partly behind my idea of a separate technical video just to compare the analogue outputs of different DAB+ radios. I would see that as a side issue. The main presentation would probably just use the DAB+ transport stream decoded with a floating point decoder. And I'd probably also make at least one example of the undecoded transport stream available for download (for geeks!).