MQA is Vaporware

[adamdea]

20 minutes ago, adamdea said:

[oh dear another duplicate]

[adamdea]

2 hours ago, botrytis said:

It means absence of preconceived notions or goals.

Ah I see Knickerhawk has responded. Apologies for the attack of pedantry.

[adamdea]

13 hours ago, John_Atkinson said:

 

Er, https://www.stereophile.com/content/mqa-tested-part-2-fold

https://www.stereophile.com/content/more-mqa

https://www.stereophile.com/content/mqa-some-claims-examined

 

Perhaps you missed these articles?

 

John Atkinson

Editor, Stereophile

Why is the claim to remove time blur , or otherwise to target the time domain response not subject to any proper questioning?

Why no analysis of whether time blur exists in actual music? 

Why the tired old reliance on visual inspection of impulse responses?

 

I know you know that time domain and frequency (with phase) domain are mathematically interchangeable, so why allow the idea that you can be wrong in the frequency domain but right in the time domain to go unchallenged? You might thereby be able to target some aspect of the time domain response at the cost of some other aspect of the time domain response. That claim would not sound so snappy for marketing , but investigative journalism might involve saying things that don't sound so good in marketing terms.

 

There some people out there who are prepared to ask the questions, and they get clicks. 

[adamdea]

1 minute ago, psjug said:

Jim Austin did say that there will be a part of the series that discusses this.  I really doubt that Bob Stuart/MQA will give him much information to work with though.

Who knows? It would be interesting enough if he were asked  the right questions and if any failure to answer them were spelt out in the article.

[adamdea]

55 minutes ago, mansr said:

Especially when it is meaningless and misleading.

And especially especially so when apologies have to be made in Stereophile measurements section for the MQA filters' inability to deal with high level tones at high frequency. Of course, we are told, real music doesn't have HF at high amplitude (true); whereas an impulse response......

[adamdea]

Hang on a second. Why do we need a new format?

[adamdea]

7 hours ago, Brinkman Ship said:

Can you point out specifically what errors and misunderstandings Mr. Austin was guilty of in his articles..I believe there was Part One and Two?

"An impulse is a very short signal—the shortest possible signal, in fact—so it's tempting to think of a test of an audio system's impulse response as a test of its response to very short signals. An impulse-response test is that, but because an impulse contains all the frequencies—for band-limited systems, all the in-band frequencies—it's a useful and commonly used measure of a system's overall fidelity.”

Hmm really? If so how does that work? Talk me through an example of how we measure the overall fidelity of two systems in this way.

[adamdea]

4 minutes ago, John_Atkinson said:

 For Stereophile's speaker reviews I use the calculated impulse response (derived from either an MLS or a chirp signal) to measure the fidelity of the speaker under test. See, for example, https://www.stereophile.com/content/kef-reference-5-loudspeaker-measurements where I derive, from the measured impulse response, the frequency response, the radiation pattern in vertical and horizontal planes, the in-room response, the step response, and the cumulative spectra-decay plot.

 

For measurements of digital products I use a diagnostic signal that I created 20 years ago, comprising a single sample at 0dBFS to derive the impulse response of the reconstruction filter. This maps the filter coefficients, revealing if it is minimum- or linear-phase and whether it has a fast or slow rolloff.

 

I am currently using the analog equivalent of this signal, generated with a monostable multivibrator circuit I built,  to characterize all the A/D converters I have available, to examine the dispersion of their anti-aliasing filters. I actually used this signal many years ago, to examine the behavior of Wadia's spline filter; see

https://www.stereophile.com/content/wadia-850-cd-player-page-2

 

John Atkinson

Editor, Stereophile

Sure an impulse response does enable you to derive the filter coordinates and frequency response of a filter (or a system). I don’t deny that it may be usefully diagnostic for some purposes.

 

 But that isn’t really what is being claimed about impulse response analysis as a tool to get at time domain behaviour of the DACs filter. And you can’t measure the overall fidelity of a band limited system just by looking at the length of its impulse response can you? After all a zero order hold circuit doesn’t have a long impulse response does it? But that staircase isn’t accurate in the time domain is it? And Benchmark’s linear phase filter does not, as Jim Austin says, show that they are only interested in the frequency domain. 

 

This canard about the impulse response being the measure of the time domain behaviour of a digital system (as opposed to being a measure, in the time domain, of the system) has been popular since long before MQA. 

[adamdea]

9 hours ago, mansr said:

Bingo.

I'm slow, it took me about 5 years to work this out from a standing start of zero knowledge of electrical engineering or information theory. The only help I got was from people on online forums.

But wait there's more [I know you know this but...].

The real giveaway of how utterly useless and misleading it is to look at an impulse response and  attempt to deduce from it the time domain behaviour of an audio filter can be seen by the following

consider

• an orthodox sharp linear phase filter cutting in at 3 khz -it may very well have entirely audible ringing because its in the audible range, there is lots of signal material in the transition band and (something else to do with the time constant of the basilar membrane at that frequency which Fokus has tried to explain to me but I haven't grasped yet.)
• an orthodox sharp linear phase filter cutting in at 20 khz. Hmm inaudible frequency and not much program material.  
• ditto at 90 Khz - come on...
• An orthodox sharp linear phase filter at 500 Khz. No wait, not only can't you hear at that frequency, but there's nothing to hear. What's the issue?.

The problem is that the impulse responses all look pretty similar in general shape (especially if there is no label on the axes of the graph). Certainly they might all be made to "ring" for the same length of time. How does our visual inspection help us without considering the frequency range of hearing and the spectrum of the real world signal? Now see how this is used in the MQA documentation. Look Look even the 90 Khz LP filter is wider than our MQA filter. (and that within a few pages of the triangle of information showing that there is no musical information over 50Khz or so).

 

Now let's consider the question of whether a minimum phase filter might mangle the leading edge of a real world transient. The impulse response says it can't, because real world transients are impulses aren't they? This "proves" that the only time domain effect of the minimum phase filter comes after the transient. But is that true?

 

[adamdea]

11 hours ago, Archimago said:

 

Hi Fokus,

Yes, ultimately I'm certainly not saying that this is demonstration of something we do not know or intuit from before. Since Man's work and revelations of all the filters used in MQA - which includes the ones they use for hi-res - we can intuitively state that what MQA is doing results in group delays across the board with each of those filters demonstrated (though perhaps to varying degrees).

 

What I have not seen in any of the magazines/diagrams of course is anyone put it as explicitly which was the whole point of that post. To show with an actual MQA DAC the temporal nature of that filter in the context of everything else like the THD+N the filter induces, the frequency anomalies (imaging), the intersample overload, and now the phase anomaly/group delay...

 

The last piece of MQA's mythology that we should consider discussing and asking of MQA is that it "improves" time-domain accuracy. Beyond just showing the impulse response and talk of ringing and such. Or be impressed by claims of 5us this and that based on "neuroscience" or whatever.

 

Already, we know that the technique reduces bit-depth, employs lossy techniques, introduces imaging, has been shown to cause aliasing distortions even when downsampling to the 44/48kHz container, adds harmonic and non-harmonic distortion. Now we explicitly add to the mix temporal distortion inherent in the filter - which is not only significant for MQA CD's but also when all those 44.1/48kHz albums are upsampled for 88/96kHz "unfold" (stuff like Ed Sheeran, Beyonce, Bruno Mars, etc. already on Tidal)...

 

I'm sure most of us already are either aware or suspicious... Beyond "time domain" claims, there is nothing left of MQA that suggests that it is "better" in any way. It literally IMO is just a shell made of promises and claims to incorporate the cryptographic signature, and control "authenticated" playback whether on one's computer or thru DAC firmware.

 

Arch I would like to thank you for taking the trouble to demonstrate the time domain effects of filters on real world musical transients. [btw I seem to be unable to post my thanks on your thread. Perhaps this is a fault with my google account as I would like to hope I have not been banned] . It is a disgrace that the mainstream audiophile press has not done this. 

 

There simply cannot be and never has been any excuse for allowing purveyors of non LP filters to claim that they are better "in the time domain" fullstop and without qualification. This is not an issue of academic interest because the moment that one recognises that any non linear phase filter (and any filter without something like full attenuation at nyquist) must change the shape of the output in the time domain, then one knows that any proper time domain claim must properly be "under these circumstances", "in this respect" and "compared with this" .

 

The audiophile press had  only to require the proponents of MQA, the minimum phase filter. the nos dac  to explain under what circumstances their product will be better,  than what,  and in what respect  -to find the claim largely evaporating. And then they would have to concede that it was worse in the following respects and in the following circumstances. Already we wold be a long way down the line to being able to evaluate marketing claims and to assess whether the hype could match the real technical point.

 

But no one did ask that question so they never did have to answer. So someone had to actually do the work. Thank you for asking the right questions and showing independent thought and intellectual honesty.

[adamdea]

On 2/4/2018 at 8:43 AM, Fokus said:

 

 He found nothing new, he only documented it more explicitly.

 

'minimum phase', 'induced phase shift', and '40-50us delay of the high' all amount to the same.

 

The overload/distortion behaviour was also known.

 

And it must be remarked here that the filter Archimago tested is the oversampling filter for non-MQA CD-rate material, and presumable for MQA CD-rate as well. But this filter may, or may not, play a role in the decoding of high-res MQA. We simply do not know at this stage.

Strictly you are right of course. But it would be a mistake to underestimate the importance of actually showing all of this on a real signal. The idea that MQA is better "in the time domain" is sticky and has gained a lot of purchase because fourier transform pairs are not widely understood and I genuinely believe that most folks assume that what they are told in the press is at least mainly accurate. It therefore needs to be spelt out.

 

Once it is understood that MQA is not accurate in the time domain, then you just have to leave it up to audiophiles to decide whether their ears and systems are good enough to hear the faults.  After all Robert Watts makes a new dac with 1m taps instead of 250k and everyone can hear the difference. 

[adamdea]

15 hours ago, Brinkman Ship said:

I have seen a few posts on other forums that in Stereophile's new issue, they "As We See It" opening column may be the first

hedge against MQA they have published in almost 4 years, written by Jon Iverson.

 

Anyone with the issue, feel free to corroborate.

I've read Iverson's article. The conclusion is that becasue we can't separate the compression side of MQA from the deblurring there no ability to tell whether it is wirth it. Hence Iverson considers that MQA is not in the long-term interests of audiophiles. He hopes it's not too late.

[adamdea]

I'm not 100% sure I follow Iverson's reasoning, but it's interesting to note the independent thought. 

[adamdea]

9 minutes ago, Brinkman Ship said:

Interesting perspective. Compared to the rest of the MQA PR Staff at Stereophile, it is practically a knife in the back of Stuart.

This reminds me of the cold-war discipline of kremlinology, where pundits would try to acertain shifts in the balance of power from snippets like who sat where during the Red Army march past.

Does  this reflect a change in editorial thought? A change in JI's thought. if he always thought it, why has he only now chosen (or been chosen) to say it. Is this the equivalent fo Kruschev's "On the cult of personality and its consequences"?

My guess is that it reflects a recognition (which took too long to sink in) that the punters and many of manufacturers really aren't happy, and that Stereophile was out of line but being so single-voiced. So maybe the herald of more diversity , rather than a shift in the single voice. But let's see. 

[adamdea]

Ok whoa @Kal Rubinson on p. 169. He prefers DXD and DSD multichannel from 2L to the MQA without room correction- "more even, continuous and convincing", and much preferred them with DRC.

Holy smoke- Stereophile has now gone subjective on MQA's ass!

 

Ignore my previous post, it seems we have real shift. 

 

"I don't see the need for [MQA]."

[adamdea]

1 minute ago, Brinkman Ship said:

Call off the dogs!!!!! Has the emperor of Rome, errr. the editor, embraced Jeezus?

I don't know, but this seems big to me.

[adamdea]

47 minutes ago, Kal Rubinson said:

With all due deference to my editor, these are independent contributions and, as far as I know, unsolicited but welcomed by him.

Thanks 

Perhaps it's just synchronicity. Either way it's interesting. One former member of this forum insisted to me on the Stereophile forum only a day or so ago that no one thought MQA sonically inferior to ordinary hi res.

 

[adamdea]

1 hour ago, crenca said:

 

Does he lend credence to the marketing speak of "de-blurring", or is he skeptical of it based on what we actually know about digital sampling, filters, etc. etc.

He doesn't express a view on that part of the argument.

[adamdea]

11 hours ago, Indydan said:

According to Dale Thorn (who was banned from CA),  we have all made fools of ourselves opposing MQA. 

 

http://www.whatsbestforum.com/showthread.php?24970-MQA-(Master-Quality-Authenticated)-Better-Sound-or-DRM&s=82297a3b6b09502c63d349487a980c3a&p=499504#post499504

 

Dale's opinion is of zero importance. I just thought it was funny (as well as Dale's inflated sense of self importance). 

I had an exchange with him the other day. Quite a lot of it was edited out. Even by the standards of people banned from this forum he seemed to have an outstanding gift for obnoxious fatuity

[adamdea]

2 minutes ago, Indydan said:

 

It gets better!

 

Dale just compared his ban from CA, to Mandela's imprisonment and Jesus being crucified! 

 

"Jesus was executed, Dr. King was assassinated, Mandela was imprisoned, and Dale was banned. I feel very honored."

 

Post number 40.

 

http://www.whatsbestforum.com/showthread.php?24970-MQA-(Master-Quality-Authenticated)-Better-Sound-or-DRM/page4

Right after he said to you:

“You're obviously a victim of your misplaced emotional angst. Get a life.”

You couldn’t argue with a person capable of such obvious humiliating self-contradiction. 

[adamdea]

14 minutes ago, beetlemania said:

I'd like to think that folks on this forum will give some love to Schiit, Linn, Ayre, et al. and likewise avoid the manufacturers that signed up for MQA.

Perhaps but the manufacturer which consistently speaks the least amount of horseshit is Benchmark. Schiit and Ayre are ok on this subject. Why is it that one doesn't really hear much about lavry these days? Dan Lavry was (and probably is) a good guy. Daniel Weiss too, although his prices are out of my league. 

[adamdea]

1 hour ago, Fokus said:

 

Better not.

 

Typical of IAR it is boring, confusing, and full of errors.

I can't even work out if it's ironic.

This article will directly affect and influence millions of you around the world - virtually all of you who have a digital signal anywhere in your home entertainment system (or professional A/V equipment). To be honestly candid, this is probably the most influential article ever published on digital (especially when all its installments are released and assembled together).

Difficult to tell. The bombastic tone and tin ear are soon accompanied on their long rambling quest by some very approximate technical explanations. Do they make it to the end? I have no idea.

[adamdea]

Has anyone heard any more about the suppose deezer tie up? I can't see anything on deezer's website

 

[adamdea]

21 minutes ago, Jim Austin said:

 

Hey, thanks for the invite. 

 

I encourage you to read the article--my little interview. I mean actually read it--because it says most of what I have to say, and it doesn't say what you seem to think it says. This is a problem you guys have, I notice: You question everyone's motives (and competence) and assume everything is freighted with conspiracies and hidden meanings. 

 

I'm not shocked to discover that you seem to think sampling theory ended with Shannon, because you smugly assume you know everything. It didn't. Shannon's work was remarkable and brilliant and extremely important--decisive for digital audio--but it was almost an afterthought for Shannon, a thing that needed doing so that he could accomplish some other thing (I forget exactly what). Even when Shannon's paper was published, others were already moving the field forward. Post-Shannon sampling theory is a real thing. 

 

Does it circumvent Shannon? No. Shannon was correct, so his theory can't be "fixed." Did I claim it did? Did I imply it? Since you folks seem to be hard of reading, or of honest thinking, I'll answer my own question: No, I did not. Did Bob Stuart make that claim? Not that I'm aware of, at least not in our interviews. You still cannot perfectly reconstruct a non-band-limited signal.

..

. Besides, it seems that anyone on this forum who dares go against the prevailing view ends up getting banned. 

 

 

I'm out. 

 

Lots of hot air. No attempt to explain how you can be "post" something which is "correct".

Also, of you concede that you strictly aren't circumventing Shannon, why use an expression calculated to imply that you are?

Or to put it another way, why (as usual) use grandiose sciencey -sounding  marketing spiel?

[adamdea]