MQA is Vaporware

[crenca]

3 hours ago, ARQuint said:

.

As a new CA member (Welcome!), shtf and may not yet be fluent in the anti-MQA catechism but the above is an argument that's not frequently deployed by those opposed to the technology. Early on in this debate...Audiophiles need to seriously consider the questions raised by Archimago in the level-headed summary published here but they also need to personally undertake the sort of comparative listening that has been the basis of high end audio's "subjective" approach for half a century.

. One's ears should indeed remain "trustworthy" and individuals can make their own judgements about MQA without undue influence of the loudest voices on both sides.

 

 

Actually, that is exactly what they need not do.  It is a misappropriation to imply that     @Archimago's work would support this sort of approach to MQA as well.  "High end audio's subjective approach" is itself what gets Audiophiles into the sort of mess and debate you complain of, because it claims for itself something it can not do - judge MQA on it's sound quality, technical, and digital eco-system merits.  

 

Question:  If you have a hammer, is everything a nail?  

 

Question:  If you have antibiotics, do you treat every disease with it (e.g. parkinson's, high blood pressure, viral infection, and bacterial infection)?

 

MQA was designed, from the very beginning, to game and beat the "subjective approach".   You were bamboozled (Bob S and company really had your number), and you continue to recommend the wrong tool for the job - the job of finding clarity around MQA.  There is no "debate".  If you like leaky, min phase, IM upsampling you might like the sound of MQA and that is perfectly fine "subjectively".  The thing is, you don't need a proprietary, DRM, "end to end" solution for that - you can do that with standard PCM and DSP/software tools.

 

So why are almost all of the writers for the standard trade publications myopically recommending the wrong tool for the job, the wrong medicine for the diagnosis?  A somewhat complex question but in my opinion it is a mix of lack of technical competence (so the utter reliance on one approach), an unwillingness to admit past mistakes, and a general laziness to try to understand the complexities of a product like MQA.  Am I right?  probably, but even if I am wrong the fact remains that the very subjectived "high end" you describe is incapable of understanding MQA and what it means for Audiophiles and the larger music lover community...

[beetlemania]

4 hours ago, ARQuint said:

Now that Roon allows MQA decoding and more DAC chipsets will have MQA capability,  motivated audiophiles will be able to do the sort of familiar-environment comparisons that we're used to doing when we bring home a new amplifier, cartridge, or loudspeaker.

 

I recently did this very thing. To my ears, TIDAL MQA sounds fine in and of itself. But in direct comparison to 24 bit versions (of what I assume is the same master), the PCM original sounds a bit more liquid and full. I resent than I might be having to pay MQA licensing via Roon or TIDAL subscription fees for a product that is no better, apparently worse, than existing digital. And I'm hugely disappointed in JA for fawning over this farce (sadly, not surprised that RH jumped the shark).

[Nikhil]

On 5/9/2018 at 12:03 PM, shtf said:

Don't know if this MQA review by International Audio Review's Peter Moncrief has been posted here yet.

Twenty years ago Peter Moncrief was known by some to be head and shoulders over all other reviewers because of his supposedly exceptionally well-trained ears.

Peter was also known to tell it like it is. That he did and this upset any number of product manufacturers. Rumor has it that reviewers (and mfg'ers) would sit around the audiophile campfire telling scary Peter Moncrief stories.

In my limited experience with products I've owned or auditioned and Moncrief reviewed, he seemed spot on in those cases.

Anyway, thus far I've only read the first few pages of his paper but wow! I also suspect this should be quite an eye opener for many and for a number of reasons.
 

http://www.iar-80.com/page170.html

 

I have been reading the Peter Moncrief's article "Digital Done Wrong" from IAR.  Wow!

Has anybody else read the piece.  I am 5- 6 pages into it and my head is spinning.

 

That article flattens everything I've read about digital.    

It is relatively heavy reading but would be interested in knowing what @mansr or @Miska have to say about it

 

Regards.

 

 

  

 

 

[Sal1950]

On 5/7/2018 at 8:49 PM, Brinkman Ship said:

woah...there are still not even a bucketful of MQA albums..these are hollow "victories"...

 

Tidal is still the single major source for MQA...and just how many premium tier subs to they have?

 

100K? 150K? Yawn.

And word is that Tidal is circling the bowl. May the big flush come soon, taking JayZ and his little crew of anti-social friends with it. 

[Brinkman Ship]

1 hour ago, Sal1950 said:

And word is that Tidal is circling the bowl. May the big flush come soon, taking JayZ and his little crew of anti-social friends with it. 

"Tidal Mystery surrounds the number of subscribers the ostensibly artist-owned streaming service, which launched when Jay-Z acquired Aspiro, Norwegian parent company of a streaming service formerly known as Wimp, for $56 million in 2015. In September of that year, he tweeted that Tidal had hit the 1 million-member milestone, though internal payments to record labels cited in Norwegian publication Dagens Næringsliv said it was closer to 350,000;  around six months later, Jay claimed it had reached the 3 million subscribers, which the Norwegian paper said was closer to a million; no further numbers have been circulated. The company has played up its exclusive content — which includes videos, films and podcasts as well as music — and while it suffered some bumps with high-profile exclusives like Kanye West’s “The Life of Pablo,” Rihanna’s “Anti” and Jay’s own “4:44,” the rollouts for Beyonce’s “Lemonade” in 2016 and Deadmau5’s “Where’s the Drop” last week went smoothly. Whatever Tidal’s future may be,  Jay is likely to come out ahead: Last year he was able to sell a third of the company to Sprint based on a $600 million valuation."

 

http://variety.com/2018/music/news/as-spotify-goes-public-how-do-its-competitors-measure-apple-music-amazon-1202741460/

 

 

[Ajax]

More from Mark Waldrep who is currently attending the Munich Audio show where he obviously bumped into Chris. IMO his last paragraph below is where we are now at with MQA.

 

Dr. AIX

..........  I had the opportunity to speak with a few familiar faces. Andreas Koch of Playback Designs was just around the corner demonstrating his new components — a server, DAC, and streamer. Juan Perez of Digibit (they make a terrific server called ARIA) came by in the afternoon. Chris Connaker of Computer Audiophile and his brother wondered by too.

 

I hadn't chatted with him in a while but congratulated him on publishing the Archimago article on MQA. He assured me that he sent the piece to Robert Stuart prior to publication but was informed that the company didn't wish to respond. Apparently, the previous Q&A that Chris had published was sufficient in their mind to "refute" the observations made in the new piece. It didn't. I'm sure that MQA is hoping that the critical analysis of their lossy codec will pass without adversely affecting their business model. I'm not so sure. It's important for our community to make sure everyone reads the article.

It's true that some of the strongest advocates in the audiophile press still cling to their "I just like the sound" subjective positions but I've noticed even some of them have recently pulled back. Changing a position from MQA is the "single biggest development in the history of audio" to "it gives musicians control of their music" or "it's at least better than MP3" is substantial.

 

However, I've heard from a large number of companies and individuals — that wish to remain anonymous — that agree MQA is a bad thing for the industry but they fear the "experts" in the audiophile press will ignore their products. Makers of DACs, servers, and other high end products depend on positive press reviews to bring attention to their products, expand their distribution network, and ultimately sell more units. Having an MQA advocate slam their latest DAC as lacking because they didn't pay the hefty MQA license fee and incorporate the codec would be bad for business — even though ignoring MQA is the right thing to do.

It's time for those that initially hyped MQA to take another look, read the technical and scientific analysis and back off their unreasoned support. MQA is NOT needed, doesn't deliver fidelity improvements, and actually takes the music distribution business backwards. 
 

[opus101]

On 5/11/2018 at 12:56 AM, Nikhil said:

Has anybody else read the piece. 

 

Read parts of it. What little I sampled was rubbish.

[FredericV]

More BS about time smearing, with some simulations where he does not explain how the examples were created:
 

 

Did Bob Stuart create the files for him to serve on a plate? We don't know. Therefore his claims cannot be peer reviewed.

 

Then he pushes the MQA renderer, as being a requirement to compensate for time smear. The solution to a non-problem. The renderer is an upsampler with one cycle of postringing. It actually makes the time domain worse, as such filters are not linear, every frequency has a different phase shift.

The far superior Archimago Intermediate Phase filter does not suffer from these flaws:

http://archimago.blogspot.de/2018/01/musings-more-fun-with-digital-filters.html

What Hans probably cannot grasp - or does not want to admit, is the fact that time smear from traditional speakers is far worse than the time smear from a linear phase digital filter. Most speakers are not phase correct. They also are not fast enough to reproduce timing events correctly, such as events originating from a point source, such as ticks. I have seen the distortions of most dome tweeters.

To name 2 exceptions, bending wave drivers (such as the Manger driver) are fast enough for high frequencies, and electrostatic drivers also can do this, but at a lower max spl.

Let's first fix these flaws in speakers, before fixing time smear at a much lower level in digital audio.

 

The underlying tactic in this video, is that now that more products have an MQA software decoder, there may be a fear (by MQA) for not buying an MQA dac. He tries to convince us we need such DAC. The underlying fear is the same as what was in the MQA financial report some time ago, that software decoding may compete with MQA hardware DAC's. This was stated in the PDF as posted earlier here.

 

[Andyman]

On 10/05/2018 at 11:50 AM, ARQuint said:

Now that Roon allows MQA decoding and more DAC chipsets will have MQA capability,  motivated audiophiles will be able to do the sort of familiar-environment comparisons that we're used to doing when we bring home a new amplifier, cartridge, or loudspeaker. One's ears should indeed remain "trustworthy" and individuals can make their own judgements about MQA without undue influence of the loudest voices on both sides.

 

 

If I were the unscrupulous head of marketing (eyeing a substantial performance related bonus) for a business such as MQA, I would be be sorely tempted to collude with my recording company partners, to provide Tidal et al with superficially similar yet compromised non-MQA versions of music to sit alongside the MQA files. I would then be happy to let audiophile ears judge.

 

Fortunately this could never happen in the real world...

 

 

 

 

[psjug]

3 hours ago, FredericV said:

More BS about time smearing, with some simulations where he does not explain how the examples were created:
 

What a joke.  His piano "simulation" has an obvious large amplitude difference from the original.

[psjug]

4 hours ago, opus101 said:

 

Read parts of it. What little I sampled was rubbish.

What does he say that is rubbish?  I find him to be so long-winded that I just can't get through the article, but what he is saying about impulse response I think is true, and the same as many here have been saying.

[opus101]

1 hour ago, psjug said:

What does he say that is rubbish?

 

...we happen to know that, for technical reasons (to be explained in another installment of this article), low sample rate digital in general and PCM in particular do have problems that lead to their sounding too dull and closed in at high frequencies.

 

Consider a transient attack's sharp and narrow single peak, from a Bob Dylan guitar pick, just as we heard it in MQA's own MQA vs. PCM A-B. Suppose the signal waveform of this sharp, narrow peak happens to be initially sampled unluckily (just as most high frequency peaks are likewise unluckily sampled). The sample dots will be way down on the flanks of this transient attack peak, and the high peak itself will have occurred somewhere near the center of the sampling interval, so its high peak amplitude will not be represented directly by any sample dots at all.

 

Just two examples of complete nonsense.

[mansr]

6 minutes ago, opus101 said:

...we happen to know that, for technical reasons (to be explained in another installment of this article), low sample rate digital in general and PCM in particular do have problems that lead to their sounding too dull and closed in at high frequencies.

 

Consider a transient attack's sharp and narrow single peak, from a Bob Dylan guitar pick, just as we heard it in MQA's own MQA vs. PCM A-B. Suppose the signal waveform of this sharp, narrow peak happens to be initially sampled unluckily (just as most high frequency peaks are likewise unluckily sampled). The sample dots will be way down on the flanks of this transient attack peak, and the high peak itself will have occurred somewhere near the center of the sampling interval, so its high peak amplitude will not be represented directly by any sample dots at all.

 

Just two examples of complete nonsense.

Someone has fundamentally misunderstood sampling of band-limited signals.

[psjug]

25 minutes ago, opus101 said:

 

Consider a transient attack's sharp and narrow single peak, from a Bob Dylan guitar pick, just as we heard it in MQA's own MQA vs. PCM A-B. Suppose the signal waveform of this sharp, narrow peak happens to be initially sampled unluckily (just as most high frequency peaks are likewise unluckily sampled). The sample dots will be way down on the flanks of this transient attack peak, and the high peak itself will have occurred somewhere near the center of the sampling interval, so its high peak amplitude will not be represented directly by any sample dots at all.

 

Just two examples of complete nonsense.

On this 2nd example, I think he is just saying a proper reconstruction filter is required to accurately render the higher frequencies within the band of interest.  So I would not say it is rubbish but he certainly does have a strange way of putting things.

[opus101]

3 minutes ago, psjug said:

So I would not say it is rubbish but he certainly does have a strange way of putting things.

 

One reason why it really IS rubbish is because his example takes an incorrectly bandlimited signal. With a correctly bandlimited signal there's no such thing as 'sampled unluckily'.

[psjug]

3 minutes ago, opus101 said:

 

One reason why it really IS rubbish is because his example takes an incorrectly bandlimited signal. With a correctly bandlimited signal there's no such thing as 'sampled unluckily'.

OK I'll take your word for it; I really don't want to go back and look.  Like I said I had a hard time with his writing style.  Also, the introduction is so grandiose that you would hope he's joking.

[Fokus]

21 minutes ago, psjug said:

On this 2nd example, I think he is just saying a proper reconstruction filter is required

 

Could be. I stopped reading Moncrief years ago.

 

Back then he was rallying against proper, conventional filters (not that he understood that they were proper) and demanded something he called 'high power averaging' as the only true way.

 

Really ... that man is clueless and full of himself.

 

 

[opus101]

14 minutes ago, psjug said:

Also, the introduction is so grandiose that you would hope he's joking.

 

That was a huge red flag for sure. What's the chance that this guy's 'most influential article on digital' is right and 'thousands of digital engineers around the world' are completely deluded?

[mansr]

1 minute ago, opus101 said:

That was a huge red flag for sure. What's the chance that this guy's 'most influential article on digital' is right and 'thousands of digital engineers around the world' are completely deluded?

"Slim" would be overestimating it by a few orders of magnitude.

[Indydan]

9 hours ago, FredericV said:

More BS about time smearing, with some simulations where he does not explain how the examples were created:
 

 

Did Bob Stuart create the files for him to serve on a plate? We don't know. Therefore his claims cannot be peer reviewed.

 

Then he pushes the MQA renderer, as being a requirement to compensate for time smear. The solution to a non-problem. The renderer is an upsampler with one cycle of postringing. It actually makes the time domain worse, as such filters are not linear, every frequency has a different phase shift.

The far superior Archimago Intermediate Phase filter does not suffer from these flaws:

http://archimago.blogspot.de/2018/01/musings-more-fun-with-digital-filters.html

What Hans probably cannot grasp - or does not want to admit, is the fact that time smear from traditional speakers is far worse than the time smear from a linear phase digital filter. Most speakers are not phase correct. They also are not fast enough to reproduce timing events correctly, such as events originating from a point source, such as ticks. I have seen the distortions of most dome tweeters.

To name 2 exceptions, bending wave drivers (such as the Manger driver) are fast enough for high frequencies, and electrostatic drivers also can do this, but at a lower max spl.

Let's first fix these flaws in speakers, before fixing time smear at a much lower level in digital audio.

 

The underlying tactic in this video, is that now that more products have an MQA software decoder, there may be a fear (by MQA) for not buying an MQA dac. He tries to convince us we need such DAC. The underlying fear is the same as what was in the MQA financial report some time ago, that software decoding may compete with MQA hardware DAC's. This was stated in the PDF as posted earlier here.

 

 

What a drama queen blowhard that Hans. 

[crenca]

5 hours ago, opus101 said:

 

One reason why it really IS rubbish is because his example takes an incorrectly bandlimited signal. With a correctly bandlimited signal there's no such thing as 'sampled unluckily'.

 

This concept, waveform "sampling", is really hard to grasp for so many.  I think it is unfortunate that the word "sampling" is used because for so many folks it implies that you are putting the humpty dumpty waveform back together again with incomplete knowledge - in a since ending up with an incomplete picture of the original waveform.  They don't understand that the waveform is calculated...

[miguelito]

Just read the latest interview with Bob Stuart on Stereophile.

 

First and foremost, the implication that “Post-Shannon” upsampling discovered some form of mathematically circumventing a theorem is deceitful and simply complete nonesense. In this era of “Post-Truth” we are living in, it is fitting that the name BS has given to this BS is “Post-Shannon”.

 

Secondly, for once I would appreciate honesty and simplicity in an explanation. There’s no way around math, honey, no matter how many pints you’ve had. If you played around with some sampling and can’t hear the difference, great, no dispute for that. But please don’t paint this as Shannon’s theorem being superseded. What you’re saying is the higher frequencies can be lossy and it still sounds good to you and those that have listened to it. Just say that honey. Noone would be upset by this.

 

And can we drop the bull about “just as the musician intended” or “like in the studio” because none of this is every done. Ok, sweetie?

 

I have listened to a LOT of MQA on TIDAL. I have a fully decoding DAC (dCS Rossini + Rossini Master Clock). Some of it is great sounding. Almost never better than high res versions. Decoding MQA sounds usually better than not decoding it or the redbook (but files are 2x the size so it better be so). It is fine, a lossy encapsulation of high resolution files. Fine.

 

I wish all the bullshit stopped and they just let the technology be judged by its own merits. Just please stop the “Post-Shannon” BS and call it “lossy encapsulation that sounds great.” And stop the “just as the artist intended” cuz NO artist is checking (well, some in some oddball labels might hav checked at some point, which does not live up to the implication in the claim). 

 

Sorry for the rant... Post-truth claims truly drive me up the wall...

[psjug]

3 minutes ago, miguelito said:

Just read the latest interview with Bob Stuart on Stereophile.

 

First and foremost, the implication that “Post-Shannon” upsampling discovered some form of mathematically circumventing a theorem is deceitful and simply complete nonesense. In this era of “Post-Truth” we are living in, it is fitting that the name BS has given to this BS is “Post-Shannon”.

 

Digital New Agers

[mansr]

1 hour ago, crenca said:

This concept, waveform "sampling", is really hard to grasp for so many.  I think it is unfortunate that the word "sampling" is used because for so many folks it implies that you are putting the humpty dumpty waveform back together again with incomplete knowledge - in a since ending up with an incomplete picture of the original waveform.  They don't understand that the waveform is calculated...

It's ok to not understand sampling and the underlying maths. However, insisting you know better if you don't have a relevant degree is ignorance at its finest.

[HalSF]

14 hours ago, Ajax said:

[Mark Waldrep}:

It's true that some of the strongest advocates in the audiophile press still cling to their "I just like the sound" subjective positions but I've noticed even some of them have recently pulled back....

It's time for those that initially hyped MQA to take another look, read the technical and scientific analysis and back off their unreasoned support. MQA is NOT needed, doesn't deliver fidelity improvements, and actually takes the music distribution business backwards. 
 

 

Art Dudley is covering the Munich show for Stereophile and takes a "controversial" [his word] stand in favor of the "I just like the sound" school of subjective MQA boosters Waldrep is referencing. Art is all in with a fairly ringing endorsement calculated to inspire a renewed Brinkman Shipstorm of ridicule in this thread:

 

Quote

 

Knowing that Wilson Audio brought to Munich a full three pairs of the company's new Alexia 2 loudspeaker ($65,000/pair; watch for John Atkinson's review in the July Stereophile), and hopeful of hearing at least one of those pairs, I steered toward the room sponsored by digital specialists dCS ltd. My reasoning: on that system, I would also be able to hear back-to-back comparisons of MQA and non-MQA versions of the same music files, derived from recordings made by Wilson Audio's Peter McGrath.

 

The less controversial of my two take-aways: driven by a Dan D'Agostino Momentum stereo amplifier and fed by a dCS Rossini DAC ($24,000) and Rossini clock ($7500)—the system also included the brand-new Rossini SACD/CD transport ($22,000)—the new Alexias sounded effortless, and harked back to the unapologetically pretty tonal balance of Wilson's Sophia 2 loudspeaker, which until now was my favorite of the brand's creations: I left the room thinking more about the music I had heard than the gear.

 

The conclusion that may sit less well with some: all of the four abbreviated selections I heard—two different piano recordings, a string quartet, and a recording by Renée Fleming, with Michael Tilson Thomas and the San Francisco Symphony Orchestra, of one of Strauss's Four Last Songs—sounded better in their MQA versions. The two piano recordings seemed to benefit the most—the sense of touch was more apparent, and more human—and although the Fleming-Tilson Thomas file also sounded more focused, musically and spatially, I also admit that the music making was so over-the-top gorgeous that the sonic distinctions, though real and in MQA's favor, mattered less to me. Still, in four out of four cases, MQA's aptly described de-blurring was an unambiguous improvement.

 

Speak of the devils (just a figure of speech!), here are Peter McGrath of Wilson Audio (left) and Bob Stuart of MQA (right), seen here taking a break from all things audio and chatting about another mutual passion, photography.

 

Read more at https://www.stereophile.com/content/arts-thursday-munich-part-two#qvtTYf0kHyvrQfCl.99