MQA is Vaporware

[kumakuma]

8 minutes ago, John_Atkinson said:

This is literally correct. When 24/48 MQA data is sent to the D/A processor, the DAC chip is presented with 24-bit data sampled at 192kHz. If every such passing mention had to include a full discussion of how MQA works or doesn't work, reviews would become unwieldy. 

 

John Atkinson

Technical Editor, Stereophile

 

 

Surely adding the words "fake" and "data" before and after "24/192" wouldn't have made that review unwieldy. 👺

[crenca]

23 minutes ago, John_Atkinson said:

This is literally correct. When 24/48 MQA data is sent to the D/A processor, the DAC chip is presented with 24-bit data sampled at 192kHz. If every such passing mention had to include a full discussion of how MQA works or doesn't work, reviews would become unwieldy. 

 

John Atkinson

Technical Editor, Stereophile

 

 

@mansr,

 

Is this correct?  Is not a 24/96 stream sent with metadata to instructing the DAC to upsample using specific MQA supplied filters? 

 

In any case, the difference between 

 

"24/48 MQA file unfolded to 24/192"

 

&

 

"24/48 MQA file unfolded to 24/96 and then upsampled to 24/192" (or something similar)

 

is the opposite of "unwieldy"

 

What it is, is an anti-consumer publication and its anti-consumer writers once again going with "the industry"...which thankfully is now mostly MQA/Bob Stuart and the "audio press" (and  perhaps the labels, though their near perfect silence is instructive) one corner and everyone else in the other, including the manufacturers who otherwise would be MQA's natural partners...

[The Computer Audiophile]

It's interesting that none of the old guard are fans of clearly upsampled high resolution releases for sale at download stores, but they are OK with clearly upsampled MQA. 

[mansr]

16 minutes ago, crenca said:

Is this correct?  Is not a 24/96 stream sent with metadata to instructing the DAC to upsample using specific MQA supplied filters? 

Right. The upsampling itself might be done on an XMOS (or whatever) processor and the result sent to the DAC chip. This is how the iFi DACs work. Some others, including the Dragonfly range, simply program the MQA filters into the DAC chip. This is of course only possible with chips supporting custom filters. Either way, the end result is the same.

[Rt66indierock]

7 minutes ago, The Computer Audiophile said:

It's interesting that none of the old guard are fans of clearly upsampled high resolution releases for sale at download stores, but they are OK with clearly upsampled MQA. 

 

Very true

[Rt66indierock]

I saw the Qobuz  numbers 200,000 overall and 25,000 in the United States. Did someone start a boycott or is this more evidence that high resolution audio is a very hard sell?

[botrytis]

3 minutes ago, Rt66indierock said:

I saw the Qobuz  numbers 200,000 overall and 25,000 in the United States. Did someone start a boycott or is this more evidence that high resolution audio is a very hard sell?

 

 Good Question. On other audio forums, it was how Qobuz inter-operates with ROON that is a complaint and the fact people didn't like the Qobuz software. It was a beta test - jeez.

[botrytis]

On 8/16/2019 at 5:10 PM, John_Atkinson said:

 

Our conversation took place in the fall of 2014, when I Charley and I had dinner during the RMAF.  Charley showed me a prototype Pono Player and we discussed the design goals and how they had been achieved, as well as the product's back story. As best as I can remember, we didn't talk about MQA other than the problem with battery life it had led to with the original Meridian proposal.

 

John Atkinson

Technical Editor, Stereophile

 

 

I guess this is what you are looking for as far as an explanation.

 

[firedog]

56 minutes ago, John_Atkinson said:

This is literally correct. When 24/48 MQA data is sent to the D/A processor, the DAC chip is presented with 24-bit data sampled at 192kHz. If every such passing mention had to include a full discussion of how MQA works or doesn't work, reviews would become unwieldy. 

 

John Atkinson

Technical Editor, Stereophile

 

Isn't this incorrect? My understanding is that the highest sample rate when converting a file to MQA is 96k, which at some point in the playback process - whether before  the DAC chip or not - is upsampled to 192. 

 

How "unwieldy" is it to write "upsampled to"  192? As in "the 24/48 MQA file unfolded, and then upsampled to 192":

That's hardly a "full discussion of how MQA works".

In any case the "unwieldy" excuse is totally lame and shows how far the Stereophile people have bought  into the Orwellian newspeak (aka marketing lingo of MQA). They don't even know they are doing it. 

As I previously wrote, before MQA they would have been sure to point out that a file was an upsample if they knew about it. Now that MQA is on the scene the seem to have a different standard. 
 

[crenca]

10 minutes ago, firedog said:

In any case the "unwieldy" excuse is totally lame and shows how far the Stereophile people have bought  into the Orwellian newspeak (aka marketing lingo of MQA). They don't even know they are doing it. 

 

 

On the one hand, true.  They don't know because they are technically incompetent.  They are like a group of (mostly) guys who started a car magazine who are not only wholly ignorant of thermodynamics (somewhat understandable) but of the most basic workings of an internal combustion engine.

 

On the other hand, not true.  They have been informed of their ignorance for about 3 years now.

 

 

10 minutes ago, firedog said:

As I previously wrote, before MQA they would have been sure to point out that a file was an upsample if they knew about it. Now that MQA is on the scene the seem to have a different standard. 
 

 

Right, which tells us that they are not interested in the truth about digital sampling and upsampling.  Rather, they are interested in serving their customers and $selling$ MQA - who are not their readers.

 

[botrytis]

1 minute ago, crenca said:

 

Right, which tells us that they are not interested in the truth about digital sampling and upsampling.  Rather, they are interested in serving their customers and $selling$ MQA - who are not their readers.

 

 

The readers are the product. The customers are the manufacturers.

[mansr]

34 minutes ago, crenca said:

On the other hand, not true.  They have been informed of their ignorance for about 3 years now.

That makes them wilfully ignorant. Unless they are outright malicious.

[Rt66indierock]

25 minutes ago, mansr said:

That makes them wilfully ignorant. Unless they are outright malicious.

 

Willfully ignorant goes right with tool of special interests or pawn in the  game of life.

[Sonicularity]

When I did listening tests between mp3 and lossless, I would convert the mp3 file to the same format as the lossless version.  Both files under test were technically the same bit depth and sample rate.  I would not call the version converted from the mp3 to be high resolution or lossless.  

[christopher3393]

On 8/16/2019 at 2:36 PM, The Computer Audiophile said:

Just finished Neil’s book. Tons to discuss September 9, 2019, the publication date. 

 

First 5 chapters available for preview here:  https://play.google.com/store/books/details?pcampaignid=books_read_action&id=5m2QDwAAQBAJ

 

Several mentions of MQA, BS in early chapters.

 

NB: I was able to purchase and download the ebook just now.

[Jud]

52 minutes ago, christopher3393 said:

 

First 5 chapters available for preview here:  https://play.google.com/store/books/details?pcampaignid=books_read_action&id=5m2QDwAAQBAJ

 

Several mentions of MQA, BS in early chapters.

 

Perhaps you have an extra comma?

 

😏

[crenca]

1 hour ago, christopher3393 said:

 

First 5 chapters available for preview here:  https://play.google.com/store/books/details?pcampaignid=books_read_action&id=5m2QDwAAQBAJ

 

Several mentions of MQA, BS in early chapters.

 

NB: I was able to purchase and download the ebook just now.

 

Are you sure you did not just "pre-order" it... In other words what you downloaded is just the sample chapters already available?

[Ishmael Slapowitz]

7 hours ago, John_Atkinson said:

This is literally correct. When 24/48 MQA data is sent to the D/A processor, the DAC chip is presented with 24-bit data sampled at 192kHz. If every such passing mention had to include a full discussion of how MQA works or doesn't work, reviews would become unwieldy. 

 

John Atkinson

Technical Editor, Stereophile

 

I and many others find your answers to common sense situations and questions to be insufficient at best, and disingenuous at worst.

[esldude]

3 hours ago, crenca said:

 

Are you sure you did not just "pre-order" it... In other words what you downloaded is just the sample chapters already available?

I can see the same thing as posted.  First 4 chapters.  I've not pre-ordered it. 

 

BTW, is it just me, or is not Phil Spector's famous Wall of Sound one of the most ridiculous over-hyped awful recording methods ever?  Only surpassed by the modern pervasive hyper-compression of all music.

 

Also can't let Neil off the hook for so many old wive's tales about digital being put in the first 38 pages I've read.  Enough already.  He has ruined his credibility for anything remotely technical.  He has gone about of his way to paint an inaccurate picture of how digital works.  He looks foolish and thinks his direct access to his music makes him an expert. 

[firedog]

4 hours ago, esldude said:

TW, is it just me, or is not Phil Spector's famous Wall of Sound one of the most ridiculous over-hyped awful recording methods ever?  Only surpassed by the modern pervasive hyper-compression of all music.

I'm also not a big fan, although he was a good songwriter, arranger, and judge of talent. I think his method might have been very good for the late 50's early 60's when people listened on crappy AM radios and crummy tabletop "stereos".

I don't think it translates well to any decent system that can reproduce a reasonable amount of detail and something close to a full range frequency response. 

[Jeff_N]

Not sure if anyone already posted this. No mention of MQA so maybe I should have posted it somewhere else, but I saw people here were talking about NY.

 

https://www.nytimes.com/2019/08/20/magazine/neil-young-streaming-music.html?action=click&module=Editors Picks&pgtype=Homepage

 

 

[Currawong]

19 hours ago, The Computer Audiophile said:

It's interesting that none of the old guard are fans of clearly upsampled high resolution releases for sale at download stores, but they are OK with clearly upsampled MQA. 

 

Quoted for truth. HDTracks got reamed when it was discovered that they were selling up-sampled music. Supposedly the record labels had sent it to them after up-sampling it. Now the labels have been busted, AGAIN, trying to market up-sampled music. 

 

Now it's on my mind, does anyone have a list of the tracks where it has been confirmed that their MQA sample rate is higher than their recorded rate?  Including the one where all the high-res material was an alias of the audible music.

 

20 hours ago, John_Atkinson said:

This is literally correct. When 24/48 MQA data is sent to the D/A processor, the DAC chip is presented with 24-bit data sampled at 192kHz. If every such passing mention had to include a full discussion of how MQA works or doesn't work, reviews would become unwieldy. 

 

John Atkinson

Technical Editor, Stereophile

 

 

No, it is NOT correct.  You yourself first wrote about the unfolding process. "Unfolding" implies that data was folded in the first place. We know that NONE of the audio above 48 kHz (corresponding to a 96 kHz PCM file) is included, so nothing above that is folded at all*. We KNOW that the "rendering" is up-sampling. How hard is it to state this? You and the other writers always go to the trouble of stating when up-sampling is being used in a DAC, so I don't see how this is different.

 

You could write a good technical article on how MQA actually works, using the wealth of data from people here, showing that it is essentially 17/96, including an explanation of how the "rendering" is really up-sampling using MQA's selected digital filters. The article could also correct all the previous errors, such as the ones stating that it was lossless. That article could be linked to every time you mention MQA being played back.

 

In the past, I gave you the benefit of the doubt that you, and many other audio writers, had simply been misled and were simply too embarrassed to admit that you'd been given false info. Now when you and Jim write the kind of replies you have, I don't understand how can you so flagrantly ignore the facts, and so transparently brush them aside. Even more so when (and good on you for doing so) you tested the output of the 10 kHz sine tone rigged with an MQA flag and saw for yourself a piece of what has been going on.

 

Remember all the times you've posted, after measuring a NOS DAC, how it is "fundamentally broken"? How about the tube amp (manufacturer's name forgotten) which had serious electrical issues? How about when measurements have revealed issues with manufacturer claims? Have you got the balls to post as much about the serious issues with MQA, just as you have with all the other products that had equally major issues?

 

*Probably because, as Dan Lavry pointed out many years ago when he was refusing to make DACs and ADCs that went above 96 kHz, that anything above that is just electrical noise, so MQA must have realised, going by tracks such as the one which I posted the spectrogram of which, that attempting to "fold" a bunch of noise was not going to work.

[christopher3393]

11 hours ago, esldude said:

I can see the same thing as posted.  First 4 chapters.  I've not pre-ordered it. 

 

 

First 5 chapters available in preview. After purchasing yesterday, the entire book was accessible -- I looked at two later chapters to verify. This morning when I went to check, I no longer had access to the book. Perhaps the seller was notified?

[mansr]

8 minutes ago, Currawong said:

Now it's on my mind, does anyone have a list of the tracks where it has been confirmed that their MQA sample rate is higher than their recorded rate?  Including the one where all the high-res material was an alias of the audible music.

How would you know what the recording format was? This information is rarely provided by the labels. All we can spot with any certainty is 44/48 kHz material upsampled and tagged as a higher original rate in MQA. That's possible since MQA does preserve at least some content up to 48 kHz. Above that it's all fake, even if the recording was in fact done at the advertised rate.

[botrytis]

21 hours ago, John_Atkinson said:

This is literally correct. When 24/48 MQA data is sent to the D/A processor, the DAC chip is presented with 24-bit data sampled at 192kHz. If every such passing mention had to include a full discussion of how MQA works or doesn't work, reviews would become unwieldy. 

 

John Atkinson

Technical Editor, Stereophile

 

 

I will give you the benefit of the doubt, as a scientist, one can be wrong. I have seen many papers that have been pulled because of errors and omissions in the explanation or data. Why can't you and Stereophile do the same? We know that what you have said above IS WRONG and this is based on data compiled here and by others. It would behoove Stereophile to get ahead of this and really state that MQA is in essence a file encoded UP TO 96/24 at full resolution of the original file (The "renderer" upsamples to whatever rate it pleases regardless of what the MQA tag says. The actual upsampled rate can be higher or lower than the displayed number.) and then anything past that is upsampled to the resolution in encoded information in that file.

 

This is a way out. MQA is a fraud that is being perpetrated on the world-wide consumers. You can be for the fraud or against the fraud, there is no middle ground.