MQA is Vaporware

[adamdea]

1 minute ago, Jim Austin said:

 

You're welcome.

 

If you mean the math behind post-Shannon sampling, the answers are Yes, no, and no. I've read two papers as best I can, but it has been a long time since I did serious math--they are beyond me. If you mean, have Craven and Stuart shown me their work, I answer 'no' to all three. I've interviewed Stuart extensively--three intervals in total I think, and many emails, and I've had the luxury of the occasional MQA-encoded test track. But the only technical material I've seen is what's widely available in patent applications and scientific journals. But it holds together, and as I've said, Stuart and Craven are not charlatans. They're serious people with serious careers behind them--and I'm not talking about Meridian. In contrast to many hear, Stuart stands up in front of packed rooms and stands behind his claims. That doesn't mean he isn't lying, but to me the preponderance of evidence suggests otherwise. 

So, then. Cutting to the point- where is the evidence of

a) an audio signal

b) an audible audio signal

which required post-shannon sampling to encode and reproduce? If you can answer that you have a moved one part of the debate along. If you can't  then you have either not asked the right questions or you don't care about the answers.

[adamdea]

34 minutes ago, Fokus said:

And, by the way, MQA is not 'post-Shannon' at all. It is standard sampling with the additional question of "what if we allow a certain amount of aliasing, trying to keep it below the programme's innate noise level", based on the unproven premise that a compact impulse response is audibly desirable.

 

Since a literally perfect sinc filter has never been practically realisable engineers have always had to face the question of what compromise was acceptable.Is that post Shannon?

Equally some engineers have decided not to bother trying to get as close an approximation as possible- are they all using post -Shannon sampling?

 

What exactly does the post-Shannon thing mean? Why would someone who understood what he was taking about, and was trying to communicate not confuse, use such an expression in this context?

 

[adamdea]

18 hours ago, The Computer Audiophile said:

@Jim Austin

 

I also encourage the CA community to offer names of experts. 

 

 

Jim lesurf? I don't know whether he would be considered parti pris because he has already written on it.

Alan V. Oppenheim, if he is still going.

In the industry Daniel Weiss? Jim Lavry? John Siau

I guess it depends whether you want an independent expert (who has not expressed any view) or just someone to balance your view and test what BS has said.

[adamdea]

8 hours ago, Fokus said:

 

Indeed.

 

I agree, that is the grown up response; but there is something rather fun in watching professional horseshit-peddlars come unstuck in a debate with amateurs.

Ok, I should grow up. Also sadly memetic natural selection does not privilege truth

[adamdea]

1 hour ago, John_Atkinson said:

 

And take a look at https://www.stereophile.com/content/mqa-aliasing-b-splines-centers-gravity

There's an interesting listening test embedded in the text.

 

John Atkinson

Editor, Stereophile

 

Can you or anyone help with this sentence

“Relaxing that constraint restores the symmetry between the time and frequency domains that was missing from Shannon's theory.”
I’m baffled. It makes zero sense to me.

[adamdea]

I’m pretty much baffled by the whole interview. I can’t see anything but conjecture in it about what may or may not be important about time domain behaviour, and it appears to concede that there are trade offs involved. All of which makes a mockery of the marketing spiel and its parroting in the press.

This is just a form of lossy encoding with a vague theory about what may or may not  be acceptable to lose.

And the post Shannon thing still seems like blather. 

[adamdea]

1 hour ago, Audiophile Neuroscience said:

 

aha , I thought it was just me.

 

I am not an audio/computer professional but at least can get my head around the "folding" of bits into a more compact package requiring less bandwidth. I think I get the master "authenticated" bit, as approved by the artist, albeit a bit misleading IMHO. However the whole "time domain" thing totally eludes me. I just can't find anything in plain English that makes sense. That may well be my failing or perhaps it's a proprietary secret?

The odd thing about Jim Austin's sentence is that the frequency domain and the time domain are mathematically interchangeable. The Shannon's proof of the sampling theorem depends on this. That relationship is immutable. So how can they not be symmetrical and how can anything restore what can;t be lost?

 

And while we are at it this of course is the problem with saying that one can be wrong in the frequency domain but right in the time domain. 

So obviously (and now eventually BS pretty much concedes this) MQA can only be targetting one sort of time domain characteristc at the expense of another. And how is this set of targets justified? :

 " For a number of reasons based on the auditory science of object detection, it seems very plausible that the first moment is of prime importance to the ear and that higher moments are less important and (importantly) can be shown not to contribute errors such as jitter."

Wow - "it seems very plausible that", well that has me convinced. MQA targets one set of time domain targets at the expense of another because "it seems very plausible" to BS that they are what matters.

But frankly the bit where he exp;lains what he means by time smear is it even more pathetically weak

Stuart: Any deviations that aliasing brings to the "impulse response" (when analog is being uniformly sampled) are quite different from the impact of the filters controlling (and contributing to) end-to-end system response. The latter is there whether or not filtering is adequate to control or eliminate aliasing. Time smear relates to the fact that the "filter" spreads every sample out in time, irrespective of frequency—particularly in the "real world," where we take into account quantization (and sometimes aliasing) effects in A/D, workstations, and DACs. "*

So now we know what it is- its the possible maths errors in calculating the impact of each tap of the filter- - but wait- if this is signifcant this will show up in eavery single test signal you ever put through the dac. And how many orders of magnitude lower than the MQA alising are they? If 50 db below the signal is ok for the aliaising then whats the problem with 24 bit quantisation?

I would be happy to be corrected by the really smart people here, but as far as I can tell BS might as well be saying- look I've been talking crap from the word go, but you my friend had better keep eating it because you'll look silly if you stop.

 

* And he finsihes: This smear, we believe, can be material for the human listener who is extracting multiple cross correlations, as well as envelope and nonlinear measures of the audio. "

Ah "we believe"

[adamdea]

1 hour ago, Audiophile Neuroscience said:

 

Thanks Adam. I can't say I am any more enlightened but I can say that is not your fault

It was as a result of exactly this sort of puzzlement (about pretty much this point) that I ended up embarking on a sort of personal night school about information theory, and electrical engineering, culminating in the purchase of Morrison on Fourier Analysis. At the end of it, apart from mildly surprising some scientist friends at dinner parties, all I achieved was a firm conviction that one should be profoundly suspicious whenever one hears the phrase "in the time domain" in relation to audio. That and the loss of <I shudder to think how much> time I could have been making money, or doing something useful like watching porn.  

Maths does have a certain beauty though.

 

[edit] the reason for this rambling story was basically to say that the reason I bore on about this stuff is that I'm hoping to spare others the wild goose chase I went on. It's like going on a pilgrimage to lourdes to find out why stork tastes better than butter.

[adamdea]

12 minutes ago, Audiophile Neuroscience said:

 

hahaha. Adam I have developed a whole new respect for you ! I can't wipe the grin off my face -porn would have been the better option!

Tout comprendre c'est tout pardonner, I hope.

Mind you, it's perhaps dangerous to use the word "wipe" in the same sentence as "porn".

 

 

[adamdea]

9 hours ago, mansr said:

The Shannon-Nyquist sampling theorem provides a sufficient condition for fixed-interval sampling to fully capture a signal and enable subsequent reconstruction. Later research has defined other conditions allowing certain signals to be accurately captured without fulfilling the Shannon-Nyquist criterion. A search for terms like sparse signal, sparse sampling, compressed sensing, and finite rate of innovation will turn up hundreds of papers spanning decades.

 

None of this is new. The reason it hasn't been applied to audio is that it is unnecessary. An audio signal has such a low bandwidth to begin with that the Shannon-Nyquist requirement is easily met. The data rates involved also pose no problems for processing, transmission, or storage systems. Even if some form of sparse sampling of audio could cut the data rate in half, say, there are good reasons not to do this outside very specific applications. Traditional sampling produces a signal that is easy to process in a multitude of ways (think of all the operations a DAW can do). That isn't necessarily true of sparse sampling. Why should we complicate everything only for the sake of a data rate reduction we don't need? I can't think of a single reason.

 

With MQA it gets worse. Even if it is based on some clever sampling method that really does reduce the data rate, it is still a proprietary system. Nobody can do anything without Bob Stuart's blessing (and the handing over of a tidy sum of money). On top of that, the distribution format has various DRM features that can never be in the interest of consumers.

 

It is clear to me that MQA is a collection of (alleged) features, each designed to appeal to a particular step in the music production and distribution chain. Sound quality for the listener. Data savings for the distributors. DRM and control for the content owners. Money for Bob from all of the above.

 

More damning still is the fact that every time a claim about MQA is poked, it falls apart like a house of cards. Bandwidth reduction? Standard FLAC does better at comparable quality. Sound quality? Accurate comparisons are made extremely difficult (why?), and with the material available, listeners are anything but united in favour of MQA. DRM? We don't know what the labels have been promised, so assessing this is impossible at this time. We do know that every DRM system to date has been broken. The odds are not in MQA's favour. Authentication (a subset of DRM)? Shoddy at best. Mastering engineers report that their work has been MQA'd without their knowledge, let alone approval. Based on my analysis of MQA firmware, tricking it into turning on the blue light is probably not difficult.

 

In the end, what have we got? By the looks of it, MQA either fails to deliver or is outright harmful in every aspect it purports to improve.

This is such a good post that an up arrow is not enough to do it justice. 

 

Leaving aside the immediate issues, the twos JAs might wish to ruminate on the following.

 

There are some of us who really want to learn and to understand the technical issues involved in digital audio. I have spent in aggregate months of my time trying to find out what sampling means, what filters are for, etc etc. And I have found in the end that there was more to learn from individuals on forums with real expertise than from the mainstream audio press. 

 

In some cases professional audio writers just don’t have the technical understanding to help anyone else out, but more depressingly, it seems that many just don’t seem to care about getting to the truth.

 

Not for the first time we have an example in Mansr’s  post of something which is simply better written, more carefully thought out and more intellectually honest than anything that Stereophile, let alone TAS, has seems to want to contribute.

 

I come to audio forums because, apart from the banter, I get to meet people who have something to contribute. It’s a privilege to kick ideas round with people who really care. In among the blizzard of mad, dull people or weird stuff  there is some real gold. 

 

I have noticed that some Hifi journalists seem to keep coming unstuck on forums by behaving in the way that forum posters do in their imagination, throwing aggressive an ill-considered insults around and generally posting as though this were a Snapchat squabble with words melting away once typed. But that is not how people achieve credibility and respect on forums.

 

Trust is like virginity, once lost it cannot easily be recovered. 

 

I think that any fair minded and patient reader who wades through this stuff will be able to discern who is sincere, who is thinking really hard about stuff, who wants to get to the bottom of a point.

 

Now this “post-Shannon” expression? Does it illuminate any point of substance? If not, what is it doing in Jim Austin’s article? This is not a timed essay, a tv debate or a playground argument. If there is a proper, reasoned answer I would like to hear it. 

 

[adamdea]

16 hours ago, mansr said:

 

I do find it curious that BS suddenly starts bringing up quantisation. He never mentioned that before. It's almost as if he's given up on the original angle and is trying a new excuse in the hopes that we'll fall for this one.

That’s how it appears. And identifying quantisation as an issue is a bit rich when you are trying to justify turning 16/44 into 13/44 and 24/96 into 17?/96 

 

[adamdea]

On 5/19/2018 at 10:00 PM, Jim Austin said:

 

Because music in the analog domain is not inherently band-limited. So, one has to apply an antialiasing filter pre-conversion, thus altering the signal that will be "perfectly" reconstructed (ignoring some complications related to amplitude quantization). 

 

It is in the assumption of bandwidth limitation that there is an implicit lack of symmetry. One counts errors in the frequency domain while ignoring errors introduced by antialiasing. In the generalized post-Shannon approach, those errors are counted--integrated into the theory. 

I've been out of the loop for a little while and I'm afraid I didn't see this answer. I'm afraid it takes us right back in a circle to where we were- the problem is that there doesn't seem to be a coherent time domain target which could be balanced against a sensible specification of the required frequency range; and beyond the vague impulse response compactness, no time domain target which an orthodox sinc-type filter can't meet; and then there's the problem, acknowledged (finally!) by Bob Stuart that the mqa type filter trades off one sort of time domain inaccuracy for another.I really can;t see that the expression "asymmetry" helps here. Equally his explanation for what the "time smear" has more or less disappeared down its own plug hole.

 

Ages ago on another thread I pointed out an example given in dspguide of a sampling used in a cardiogram where not having a reconstruction filter might be better for one sort of purpose. But of course in that example there the sample rate is way less twice the spectrum of the signal, and they were only trying to get out very specific information. The problem is-what has any of this got to do with audio ie what exactly is it that using a decent sinc style filter for 44khz sampling upwards (or for that matter an ordinary minimum phase filter) will blur or spoil?

[adamdea]

3 hours ago, mansr said:

Probably something using phrases like boxed in, veiled, digital, artificial, lifeless, robbed of its soul, and so on.

My money would be on “ etched, superficially impressive but ultimately fatiguing and lacking in nuance”.

[adamdea]

11 hours ago, Audiophile Neuroscience said:

Imagine if people are asked if they prefer chocolate to vanilla ice cream and the result comes back 50:50. Does that mean they were guessing or have no preference?

The answer to this is pretty clear if each person has consistent preferences. But if they choose which is in fact chocolate one time and vanilla another time, then we draw our own conclusions.

It's worth bearing in mind that in archimago's test the results showed

"9 listeners selecting all MQA and 12 listeners selecting all hi-res PCM" That's out of 83 listeners. This strongly suggests that  the results are more likely to be random than just spilt between mqa lovers and haters. If it were the latter than we would expect many more people to chose consistently. As it is  the results seem to match a random profile (ie about 1/8 for 3 MQAs or hi rezs.). It is also noteworthy that there was no tendency to prefer mqa amongst "experts" or the young. If there really were some interesting hi frequency/time domain accuracy effect at work you might expect this to show up amongst those who think they have better hearing, or even amongst those who actually do,.

 

It could of course be possible that preferences vary from track to track . If so I guess you might want to repeat the whole thing. My money would be that there would be considerable variation for individuals if it were repeated. In any event this seems a bit academic as either way MQA would not really be most people's first choice for most things.

 

 

[adamdea]

11 hours ago, Audiophile Neuroscience said:

 

I agree the difference stats do not support obvious differences. It would have been interesting to further test the cohort that nominated clear or moderate difference in a straight up ABX comparison to see how they fared but understandably beyond the scope of what could be achieved . That said I am on record as challenging the scientific validity of such test methodologies (as yet) but am open to the possibility. Kudos on what you did achieve, there should be more taking the time and effort to establish objective methods.

Whether or not one objects to ABX (and I'm not convinced there is any reason to), it might not be necessary. Since the number of people choosing 3 of either looks consistent with chance, why not just get them to repeat? If they don't consistently  come up with the same results second time then I wouldn't bother testing further.  Obviously the sample isn't very large, and it isn't entirely foolproof, but it might be a useful and fairly easy to perform  threshold test. If more than (say) 6 of them got the same result twice, it might be worth checking them. 

[adamdea]

12 hours ago, Archimago said:

?

 

I'd love to see the text of the McGill University study:

http://www.aes.org/e-lib/browse.cfm?elib=19396

 

While the methodology is of course different, sounds like the result is similar with listeners unable to overall significantly discriminate MQA vs. unprocessed file.

 

They were asked to assess the relative clarity of the two formats. They did not do so consistently ie there was no consistent preference.

Lots of analyses are done for particular types of music and categories of listener. There was ultimately no clear evidence of any individual having a consistent preference. There was a suggestion of a significant result for casual listeners preferring MQA for jazz on headphones,  but PCM speakers, but it's not clear whether a bonferri correction was done for it, and it seems a bit half hearted.

 

[adamdea]

4 hours ago, ARQuint said:

I am not disputing (or misunderstanding) the conclusions Archimago drew from his Internet exercise. Rather, I looked at his data and noted that there were plenty of instances (78, to be specific) when a listener heard a "moderate difference" or "clear difference" between the high-resolution PCM file and the MQA-treated one. I mention them as examples of a perceived positive effect of MQA on SQ outside of the audiophile press.

 

 

To quote the main himself

 

 

An exact 50:50 coin toss even within the group of listeners who thought they heard significant differences to a moderate or obvious degree. Again, there is no preference towards MQA Core or just standard hi-res PCM playback. "

 

Seriously have you thought this through, Mr Quint?

[adamdea]

1 hour ago, ARQuint said:

 

I have. The fact that half preferred PCM and half preferred MQA is interesting but not the point at all. People have different tastes / value systems when it comes to reproduced sound. The point is that these people heard a difference and then decided which version they liked better—Archimago required that they do so.

 

Imagine this hypothetical scenario. You need to replace a failing power amplifier and have a budget of $5000. Your dealer lets you take home, for the weekend, two solid state products, "A" and "B", that have a similar design and power ratings. "A" costs $4000 and "B" costs $5000. You listen to both for the weekend with familiar recordings, taking copious notes, and conclude that you prefer "B" for its smoother top end, better spatiality and superior bass clarity. You're then provided with the information that in a blinded trial with 100 experienced audiophiles, all heard a difference between the two amps—but 50 preferred "A" and 50 preferred "B". Would that information effect your decision regarding which product to buy? Would you choose to save $1000 because of the results of the trial?

 

I also want to emphasize that my own modest preference for MQA-encoded files comes not from demos at shows but from my experience of reviewing the Aurender A10 when I had a couple of months to make my own comparisons of streamed MQA Tidal files with files of the same resolution downloaded from HDtracks.

 

Andrew Quint

No Mr Quint you have not thought this through

Unfortunately your hypothetical example is of no application here. The 50/50 split did not show a consistent preference. It is not the case that 50% of listeners consistently favoured MQA. Only 9 out of 83 voted MQA for all three tracks. This is exactly what you would expect if they were guessing. The 50/50 is an aggregate of what people chose when they thought they were hearing a clear difference.

 

Guessing heads or tails correctly 50% of the time does not make you a part-time genius.

 

 

[adamdea]

13 hours ago, Audiophile Neuroscience said:

 

But we can't read more into it than what is known. My point based on 50:50 preferences alone was simply we can't know if their choice of chocolate or Vanilla was a clear preference or random ( nor if they would flip next time asked). It could conceivably be either, therefore not wise to make conclusions. I believe Archimago agreed but added additional clarifications so all good. Such distributed tests have their limitations as Archimago mentioned but I agree with him that the whole evidence thus far is not supporting MQA as a better SQ medium. To the extent that there is doubt the burden of proof is on those promoting MQA.

I agree with all of this in principle, but we do know rather more than 50:50 aggregate preferences. We know that only 9/83 look like candidates for having a clear and consistent preference for MQA.

Particularly when you factor in the McGill result then it does all look pretty random.

But I accept that you couldn't rule out some people having most subtle preferences.

 

I agree that the distributed file tests have their limitations but at least people get to decide for themselves based on preference rather than identification, and can do so at home on a familiar system.

 

Anyway, bearing in mind Archimago's previous results, one shouldn't be too surprised.

[adamdea]

15 minutes ago, Em2016 said:

 

Regarding your bold parts (and I would honestly prefer MQA would just go away) but if you compare the 1st unfold (I'm only talking the 1st unfold and nothing beyond that) to the same master that can be purchased, they are essentially the same. It's easy to verify if you're able to capture and analyze the streams and purchases on your computer.

 

You can test with the album MAGNIFICAT on Tidal (353kHz) and purchase the 24/96 version and DXD master. The 1st unfold on Tidal stream and the 24/96kHz purchase will look identical. 

 

Again I'd prefer MQA go away. Again, I'm only talking about the 1st unfold. 

 

@Archimago said similar....

 

"Objectively with the songs I examined, the software decoder works well to reconstruct what looks like the equivalent 24/96 download."

 

and

 

"Bottom line: TIDAL/MQA streaming does sound like the equivalent 24/96 downloads based on what I have heard and the test results"

 

https://archimago.blogspot.hk/2017/01/comparison-tidal-mqa-music-high.html

 

Sure there is a sound point underlying the MQA triangle- there is very little information on any recording which couldn’t be captured with noise shaped 16/96. 

Unfortunately BS is not interested in distributing noise shaped 16/96, which could be packed in 24/96 with the bottom 8 digits frozen and packed in a similar size flac container to MQA. 

It would have the same information content as pretty much any 24/96 file (and  pretty much any 24/192 either).

[adamdea]

19 hours ago, esldude said:

If you play back the 1st unfold does it not engage the rather lax filtering of MQA?

 

 

Yes. True. You could presumably play it back with a proper filter but you'd be stuck with alising from the 24/192 downsampling to 24/96. It's not ideal, but probably still unlikely matter too much.

[adamdea]

On 7/14/2018 at 1:57 AM, Brinkman Ship said:

Certainly....HDCD also did not share any of the nefarious traits of MQA..in the end it is a footnote.

Actually it did share some - Jim Lesuf has written about this - first the dynamic expansion was supposed to give greater than 16 bit performance but in fact meant that even though it was supposed to be compatible it meant that on non HDCD machines you were running the risk of reduced dynamic range or even dynamic peak compression. [I'm not sure MQA does the latter]

http://www.audiomisc.co.uk/HFN/HDCD/Enigma.html

Second there was a promise of adaptive filters but there is considerable doubt as to whether the feature was ever used. I checked with Cambridge audio and linn engineers who resigned hdcd machines and they seemed to think it never was. In this regard it might be better than MQA which seems to have actually delivered on the dodgy filters.

The interesting point is that there was virtually no proper examination of the implementation, just a load of cheerleading.  Very familiar.

[adamdea]

18 hours ago, Brinkman Ship said:

I was actually not referring to the technical aspects..more so other implications.

Ah yes- right you are. I dare say the original founders were hoping to get a royalty from chip manufacturers and the music industry. But it didn't seem so sinister as a land grab and didn't have the same drm issues. I spose it could have threatened the availability of proper redbook if it had become the universal standard.

[adamdea]

On 9/23/2018 at 12:34 PM, John_Atkinson said:

 

 

 

You then get the advantage, as I described in my September 2018 article on A/D conversion - https://www.stereophile.com/content/zen-art-ad-conversion and was also touched on by Paul Miller in the July 2017 issue of Hi-Fi News - of more accurately preserving the time-domain aspect of the original signal. But again you don't get something for nothing: that time-domain optimization of the digital transmission chain allows for there to be aliased energy in the reconstructed signal.

?

 

The problem is that this is a misdescription and one which could and should have been corrected from the outset. At most (and only arguably) it optimises one sort of time domain behaviour at the expense of others. Only in the case of synthetic pulses could it possibly represent a pareto improvement. In other cases its reconstruction will be inaccurate in the time domain. This has to be the case.

This is not just a careless use of language. When stated accurately the slogan  evaporates. Making the accurate statement would be good journalism but poor advertorial. You have a choice.

.

[adamdea]

5 hours ago, John_Atkinson said:

Thank you. Had you posted this analysis in response to my question two days ago , a lot of bandwidth would have been saved.

 

John Atkinson

Technical Editor, Stereophile

 

Leaving aside the froth and noise rising on top of this exchange, there is a real issue in here, John., which merits sober reflection.

 

Mansr and others have made a very valuable contribution by some painstaking research into MQA. There is careful and rigorous thought involved. That care and rigour has not gone unnoticed, and has drawn unfavourable comparison with that found in print media..  Had you been paying attention to what was going on you over here you may have taken his points on board.

 

Mans is a very clear thinker, I'm sure that if you can identify a flaw in his reasoning I'm sure he will consider and reconsider as appropriate.