MQA is Vaporware

[daverich4]

6 hours ago, Paul R said:

 

No, you are rationalizing here. The point was there are plenty of "measurements" of MQA sound out there on the net, and from some pretty respectable sources. That was the question you know.

 

Mansr and Danny scoffed at it, but that is really arrogant tunnel blindness, and as you say, taking the review in utter isolation against what one might know. The review simply said turning on the MQA On switch improved the sound. I have seen nothing here that invalidates or even casts doubt on that. 

 

If it sounds better with the MQA switch on, then it is also quite reasonable to assume that the MQA switch engages MQA processing. In this case, probably just a MQA filter, which, by sufficient stretch of the imagination, could be construed as "turning on MQA."  

 

I personally would not agree with that, however, I would be enraged if they had pulled that on a iPhone. And shocker - Darko and others say their music sounds better on the iPhone with MQA processing. 

 

Obdurate is insisting that everyone else in the world must agree with you because you are RIGHT! Why can't those knuckleheads see that? I do not think I am often guilty of that. Perhaps occasionally, but not here on this subject. 

 

MQA is an. extreme example, you do find that behavior all over audiophile culture though. We used to regularly get missionaries from Hydrogen Audio over here to "save us" from our foolishness. How is this utterly cretinous MQA crusade any different? Or insisting that Digital sounds better than Vinyl, or vice versa? 

 

 

 

And what, except very specific knowledge and understanding of actual MQA performance, gathered pretty much exclusively here I suspect, would make you suspect that the MQA ON switch does not turn on MQA processing? 

 

 

Perhaps I am being influenced a bit because am researching some historical happenings surrounding WWI lately. I don't think that there has ever been a greater set of lies sold with "facts" than at that time.* An incredibly nasty war, fought for reasons that were nothing more than a fabrication of lies. Then many more lies and the even worse horrors of WWII. It may have been a hundred years ago, but people do not seem to have changed all that much today. 

 

In a very small way, this MQA crusade  is exactly like that. People are choosing a particular set of facts, in isolation, and pushing them as the absolute truth that can not be denied.That is exactly as inappropriate as telling someone they are not an audiophile because they have not spent enough money on their equipment, or because they don't like vinyl, or don't like digital, or are not subscribed to Tidal, or what have you. 

 

Its just like when we used to get missionaries from Hydrogen Audio every week or so, all determined to save us from our totally unscientific ridiculous beliefs that DACs could sound different, or any cable can be better than 18g zip cord. Except the people here are the missionaries today, torches, boiling oil, tar, feathers, and pitchforks arrayed against MQA or anyone who says a damn thing they don't like about it. 

 

*Well, except perhaps for some of  the damn yankees during reconstruction. Still telling the same lies today.  Or maybe during the McCarthy era. (*sigh*)  Maybe it just never ends. 

 

You are way more articulate than I am. And more willing to engage the lynch mob here...

[james45974]

I don't see the lynch mob here, just guys who aren't content being lead around blind by MQA fluffery and who are passionate about it.  Maybe Resistance would be a better name than lynch mob.

 

To my knowledge there has still been no verifiable comparison of every day recordings of the same verified master with and without MQA by MQA, don't you wonder why?  If its so great you think they would be all-hands-on-deck showing that instead of avoiding comparisons.

 

Finally, I am of the opinion that the only entity who should be doing any authentication is the original musical artist, not the producer, mixer, engineer, bean counter, cafeteria lady, or Universal librarian.  To that end I would not trust any "MQA" of deceased artists.

[KeenObserver]

28 minutes ago, daverich4 said:

 

You are way more articulate than I am. And more willing to engage the lynch mob here...

 

The mob has fashioned a rope made of actual truth borne of scientifically sound testing.

 

They hoist MQA on a petard of false claims.

[KeenObserver]

"You people that are launching all these scientifically proven facts at MQA are just awful people.  You're just like those awful, awful people that started World War One".

 

" There are "facts" out there that prove that MQA is just the most wonderful thing ever invented.  Honest.  Those facts are out there.  Swear to God".

[John Dyson]

 

(I hesitate posting this -- but REALLY feel that DRM and its negative effects for the consumer blows-away the quality issues -- even though the quality issues ARE not beneficial, and simply wasted against the vacuous advantage for the consumer.  The only positive is for licensors -- not really the consumers at all.)

 

I keep seeing arguments about MQA modifying the signal  (in the encode or decode process -- either one or both), but all of the relatively small differences are much less important than 'differences in mastering' and/or 'DRM and limting freedom of use'.   It is advantageous to the vehemently pro-MQA crowd to be distracted away from the more important differences.  Sure, DRM does damage the signal worse than a proper high quality distribution, but that isn't the big bugaboo...

1) Bugaboo #1 -- DRM and the loss of control -- both of use and modification, plus HW costs/exclusivity.

2) Bugaboo #2 -- (less important) MQA being sold as 'better', when it is actually 'inferior' when related to today's technology and capabilities.

 

Why am I most worried about Bugaboo#1?  Of course, the most obvious is a lack of effective 'ownership'/full control of a personal copy. Then, once you get some music, happens to be mismastered piece of garbage, then how do you do a local correction without using a lower quality copy to work from?  I predict (never hearing MQA material), that there is mismastered material, but with MQA, one is stuck with the mess -- having less control to correct the mess.  The local, consumer copy has much less value for future/improved  play.

 

Bugaboo#2 -- MQA being better...  It is NOT better in both of these ways:  1) it distorts the signal, removes data and then artificially tries to approximate a replacement.  Note that mp3 does that kind of thing also, but using a different technique.  2) MQA doesnt' actually address the real problem of quality, properly mastered material. 

 

Both Bugboos have similar attributes, but in reality 1) MQA doesn't fix ANYTHING, 2) MQA takes away data/freedom/etc., 3) MQA adds licensing headaches.  (Back 10-15yrs ago, a bit of data compression would have been nice -- now, the kind of data rates/amount of compression makes that MQA consumer benefit very unconvincing.)

 

MQA shouldn't have gotten very far, and the only reason why it has gotten as far as it has -- the situation kind of reminds me of 'The Emperor's new clothes'.

 

John

 

[firedog]

2 hours ago, Thuaveta said:

 

It's explained: "MQA mode" bypasses Android's audio pipeline, it's just that one has to take the time to read the text between the pictures: "This shows that the Android audio pipeline is limited to 16 bits so best to avoid it if you can. Fortunately the LG player does that and therefore, there is no need to buy USB Audio Pro (unless you like its interface better)."

Right, so that already shows that the device is setup so that the MQA path is a superior one, regardless of whether the file itself is an MQA file. Ergo, those improved results in MQA mode don't necessarily show anything about MQA itself. 

[Thuaveta]

38 minutes ago, firedog said:

Ergo, those improved results in MQA mode don't necessarily show anything about MQA itself. 

 

Exactly.

 

Same can often be said about subjective evaluations of MQA releases themselves, of course... Are people hearing improvements from MQA, or do they just prefer the (louder ?) remasters ? Dollars to donuts a serious study would demonstrate it's the latter.

[mansr]

45 minutes ago, firedog said:

Right, so that already shows that the device is setup so that the MQA path is a superior one, regardless of whether the file itself is an MQA file. Ergo, those improved results in MQA mode don't necessarily show anything about MQA itself. 

In fact, Amir's results necessarily show nothing about MQA itself for the simple reason that the test signal was not MQA encoded.

[tmtomh]

9 hours ago, Paul R said:

 

No, you are rationalizing here. The point was there are plenty of "measurements" of MQA sound out there on the net, and from some pretty respectable sources. That was the question you know.

 

No, you are rationalizing. The point of measurements of MQA is that they are valid only if they are... measurements of MQA. If an "MQA-enable" switch does more than simply enable MQA, but rather enables MQA and also change other aspects of the signal path/processing, then by definition any measurements of that switch are not in fact measurements of MQA.

 

Do the linked measurements purport to be measurements of MQA? Yes. Are you correct that MQA is not usually measured/listened to in isolation (i.e. other variables like mastering, equipment, or signal path are altered at the same time)? Yes. 

 

Where you're wrong is how you put together - or refuse to put together - these two facts. If the measurements that purport to be measurements of MQA are not in fact actually measurements of MQA, then there are not in fact measurements of MQA "out there."

 

You know all this, and it's a shockingly ironic display of tunnel blindness on your part for you to so stubbornly acknowledge it.

 

To put it another way, you have forthrightly and proactively acknowledged that measurements of MQA are not actually just measurements of MQA - but instead of acknowledging the significance of this fact, you instead want to tell others how stubborn, blind, and generally asinine they are.

 

You can do that of course; but in that case you should be surprised at the responses you get.

 

And P.S. @daverich4, I understand that it can be unpleasant to have to deal with an online mob - and I am on record repeatedly as very much opposed to some of the more polemical and nasty comments here. But I feel it is important to note that just because multiple people express disagreement, that does not make them a "lynch mob." Casting those who disagree as a mob or "groupthink" simply because they disagree and there is a group of them, is an intellectually lazy way to cast one's own position as rational and correct even if it is not supported by the facts or evidence.

[Paul R]

3 hours ago, james45974 said:

I don't see the lynch mob here, just guys who aren't content being lead around blind by MQA fluffery and who are passionate about it.  Maybe Resistance would be a better name than lynch mob.

 

To my knowledge there has still been no verifiable comparison of every day recordings of the same verified master with and without MQA by MQA, don't you wonder why?  If its so great you think they would be all-hands-on-deck showing that instead of avoiding comparisons.

 

Finally, I am of the opinion that the only entity who should be doing any authentication is the original musical artist, not the producer, mixer, engineer, bean counter, cafeteria lady, or Universal librarian.  To that end I would not trust any "MQA" of deceased artists.

 

 

I don’t see the lynch mob either, but that is because I have implacably hostile people Iike that Mikeyfresh guy and the Slappy fella on my ignore list. 😇

 

I just wrote a reply to firedog along the lines of what you suggest in your second paragraph, but did not publish it. I suspect the answer is simple. MQA does not want to pay for the research into why many people, some with very respectable listening skills, say MQA encoded files sound better. Why prove something people are not disputing? At least, are not disputing with research. 

 

Nobody here wants to research why MQA sounds better to some people because - horror of horrors - honest research might show that people do prefer the sound of MQA encoded files, and why. That would be a win for MQA. 

 

The arguments presented here as to as to why MQA just seems to sound better? They are all opinion, supposition, or rhetoric - not based upon solid research. All the research here, as very well done as it is, only suggests MQA has engaged in deceptive advertising, and that alternative existing technology probably sounds as good as MQA based upon not being lossy.  

 

Another reason the research is not being done is that positive results would also vindicate the much reviled “old guard,” who have been so unfairly vilified here. Without that old guard, most of us would not be enjoying our hobby today. That seems to be forgotten far too often. Ageism is disgusting.

 

I don’t trust MQA on much of anything at all regarding provenance, but I do think there is room for other opinions than just the artist. Rather, I think that engineering and mastering has artistic elements to it. I also think that the bean counters paying for the product should have some say in it. They are assuming the risk. 

[mansr]

3 minutes ago, tmtomh said:

Do the linked measurements purport to be measurements of MQA? Yes.

They don't, though. Amir shows a screenshot of a box labelled "MQA" being ticked followed by the same non-MQA 1 kHz test tone as before.

[Paul R]

5 minutes ago, tmtomh said:

 

No, you are rationalizing. The point of measurements of MQA is that they are valid only if they are... measurements of MQA. If an "MQA-enable" switch does more than simply enable MQA, but rather enables MQA and also change other aspects of the signal path/processing, then by definition any measurements of that switch are not in fact measurements of MQA.

 

Do the linked measurements purport to be measurements of MQA? Yes. Are you correct that MQA is not usually measured/listened to in isolation (i.e. other variables like mastering, equipment, or signal path are altered at the same time)? Yes. 

 

Where you're wrong is how you put together - or refuse to put together - these two facts. If the measurements that purport to be measurements of MQA are not in fact actually measurements of MQA, then there are not in fact measurements of MQA "out there."

 

You know all this, and it's a shockingly ironic display of tunnel blindness on your part for you to so stubbornly acknowledge it.

 

To put it another way, you have forthrightly and proactively acknowledged that measurements of MQA are not actually just measurements of MQA - but instead of acknowledging the significance of this fact, you instead want to tell others how stubborn, blind, and generally asinine they are.

 

You can do that of course; but in that case you should be surprised at the responses you get.

 

And P.S. @daverich4, I understand that it can be unpleasant to have to deal with an online mob - and I am on record repeatedly as very much opposed to some of the more polemical and nasty comments here. But I feel it is important to note that just because multiple people express disagreement, that does not make them a "lynch mob." Casting those who disagree as a mob or "groupthink" simply because they disagree and there is a group of them, is an intellectually lazy way to cast one's own position as rational and correct even if it is not supported by the facts or evidence.

 

We disagree. You assume much and put many separate pieces of information together to build an opinion about how  “Turning on the MQA switch” actually improved the sound. Heck, I do too. 

 

But what basis  are you using to assume that the great majority of audiophiles in the world are going to going to emulate your logic?  Research? Where?

 

In fact you are making an unjustified assumption based upon incomplete and unverified data, accepted because you do not like MQA. I can understand that, in fact, I might even agree with you for the same reasons and based upon the same assumptions.

 

Does not change the fact there is no research that supports that assumption. Tunnel vision,  or just willing self delusion? Does not really matter does it? 

 

Until solid, difficult to dispute research is done into why people sometimes prefer the sound of MQA files - and how often -  the arguments given to dispute it are just rhetoric and opinion. 

 

That’s okay by the way, because MQA just isn’t that important to a lot of people. If turning on the MQA switch improves the sound of their music, they quite rightly do not care what that switch actually does, and will continue to refer to it, however inaccurately, as “turning on MQA”.  Thinking otherwise is - to me - a case of true believer tunnel blindness. 

 

 

[mansr]

9 minutes ago, Paul R said:

Nobody here wants to research why MQA sounds better to some people because - horror of horrors - honest research might show that people do prefer the sound of MQA encoded files, and why. That would be a win for MQA.

What do you make of that McGill study? You know, the one that failed to show any consistent preference for MQA material despite being somewhat rigged in favour of producing a positive result.

[crenca]

3 minutes ago, mansr said:

What do you make of that McGill study? You know, the one that failed to show any consistent preference for MQA material despite being somewhat rigged in favour of producing a positive result.

 

Remind me again, what this refers to...I have in the back of my mind (lack of) volume leveling but that can't be it...

[Paul R]

13 minutes ago, mansr said:

What do you make of that McGill study? You know, the one that failed to show any consistent preference for MQA material despite being somewhat rigged in favour of producing a positive result.

 

Actually, I rather like it. I think that the the results were perfectly valid within the experimental framework they were conducted in. Well done, in fact.

 

I do think more more work needs to be done, as the emphasis was on one single aspect, namely the clarity of the sound. I strongly suspect there is more to any supposed preference for MQA than just the clarity. Also, the setup ran only through Brooklyn DACs. Using other DACs  to duplicate and verify the results would be good. 

 

Want to clarify what you mean by rigged there by the way? I do not think the experiment was rigged in favor of one result or the other. They made great efforts to eliminate bias

[firedog]

20 minutes ago, Paul R said:

 Why prove something people are not disputing?

Nope.

Some people like MQA. No one really cares.

 

The issue (beyond DRM and monopoly type issues) is that a)technical analysis shows MQA has serious issues of added distortion and aliasing; technical claims by MQA about deblurring, improved temporal response, unique to ADC/DAC etc, have been shown not to be true; b) and claims for how vastly and obviously superior MQA is (not just “I like it better”) are not heard at all by many people. Many people listen to some MQA cut that is supposedly so great and just don’t hear what’s described. 

That makes all those claims about how good it sounds suspect - especially since virtually all those testimonials are sighted.

So a lot of people are “disputing” - and MQA, if they were honest, would be trying to show that the critics are wrong. The fact that they don’t indicates they are aware they are selling snake oil. 

[Paul R]

2 minutes ago, firedog said:

Nope.

Some people like MQA. No one really cares.

 

The issue (beyond DRM and monopoly type issues) is that a)technical analysis shows MQA has serious issues of added distortion and aliasing; technical claims by MQA about deblurring, improved temporal response, unique to ADC/DAC etc, have been shown not to be true; b) and claims for how vastly and obviously superior MQA is (not just “I like it better”) are not heard at all by many people. Many people listen to some MQA cut that is supposedly so great and just don’t hear what’s described. 

That makes all those claims about how good it sounds suspect - especially since virtually all those testimonials are sighted.

So a lot of people are “disputing” - and MQA, if they were honest, would be trying to show that the critics are wrong. The fact that they don’t indicates they are aware they are selling snake oil. 

 

We look at the same data and come to different conclusions. 🤪

 

I guess that is half the fun though.  Where I see good technical research on the data and format, I do not see a lot of experimental proof of the assumptions from that research.

 

For example, one might assume the McGill testing recently discussed would have shown distinct and compelling faults in the clarity of the sound because of MQA introduced distortion. It did not, which suggests the assumption may be wrong. 

[Paul R]

3 hours ago, John Dyson said:

 

(I hesitate posting this -- but REALLY feel that DRM and its negative effects for the consumer blows-away the quality issues -- even though the quality issues ARE not beneficial, and simply wasted against the vacuous advantage for the consumer.  The only positive is for licensors -- not really the consumers at all.)

 

I keep seeing arguments about MQA modifying the signal  (in the encode or decode process -- either one or both), but all of the relatively small differences are much less important than 'differences in mastering' and/or 'DRM and limting freedom of use'.   It is advantageous to the vehemently pro-MQA crowd to be distracted away from the more important differences.  Sure, DRM does damage the signal worse than a proper high quality distribution, but that isn't the big bugaboo...

1) Bugaboo #1 -- DRM and the loss of control -- both of use and modification, plus HW costs/exclusivity.

2) Bugaboo #2 -- (less important) MQA being sold as 'better', when it is actually 'inferior' when related to today's technology and capabilities.

 

Why am I most worried about Bugaboo#1?  Of course, the most obvious is a lack of effective 'ownership'/full control of a personal copy. Then, once you get some music, happens to be mismastered piece of garbage, then how do you do a local correction without using a lower quality copy to work from?  I predict (never hearing MQA material), that there is mismastered material, but with MQA, one is stuck with the mess -- having less control to correct the mess.  The local, consumer copy has much less value for future/improved  play.

 

Bugaboo#2 -- MQA being better...  It is NOT better in both of these ways:  1) it distorts the signal, removes data and then artificially tries to approximate a replacement.  Note that mp3 does that kind of thing also, but using a different technique.  2) MQA doesnt' actually address the real problem of quality, properly mastered material. 

 

Both Bugboos have similar attributes, but in reality 1) MQA doesn't fix ANYTHING, 2) MQA takes away data/freedom/etc., 3) MQA adds licensing headaches.  (Back 10-15yrs ago, a bit of data compression would have been nice -- now, the kind of data rates/amount of compression makes that MQA consumer benefit very unconvincing.)

 

MQA shouldn't have gotten very far, and the only reason why it has gotten as far as it has -- the situation kind of reminds me of 'The Emperor's new clothes'.

 

John

 

 

Hi John. 

 

DRM is a concern, but mostly it is a blue herring. I can not think of any DRM system that was not swiftly bypassed the moment it became a nuisanice. Might be one, but I seriously doubt MQA would be that example. 

[mansr]

23 minutes ago, crenca said:

Remind me again, what this refers to...I have in the back of my mind (lack of) volume leveling but that can't be it...

The participants were asked to rate the "clarity" of the material they listened to. There are (at least) two problems here:

1. Clarity is not a well-defined, measurable property. As such, interpretations will vary, adding unnecessary noise to the data.
2. The test failed to establish that a difference was at all discernible before attempting to quantify it and pick a preference.

[firedog]

1 minute ago, Paul R said:

 

We look at the same data and come to different conclusions. 🤪

 

I guess that is half the fun though.  Where I see good technical research on the data and format, I do not see a lot of experimental proof of the assumptions from that research.

 

For example, one might assume the McGill testing recently discussed would have shown distinct and compelling faults in the clarity of the sound because of MQA introduced distortion. It did not, which suggests the assumption may be wrong. 

 

True. But there’s a bigger point you seem to be missing. MQA is marketed and touted by the audio press as “better sounding than CD”; and equal to or better sounding than hi res. All those flowery descriptions of how much better MQA sounds - even than hires versions of the same cuts. 
A lot of the hostility to MQA comes as a result of that. If they had simply introduced it as a compression scheme that’s slightly lossy, but “we think it’s perceptually lossless” when compared to hires, it would be much less of an issue.

And MQA DOES have something to prove when they make claims like that. Let’s see them prove something. 

[mansr]

2 minutes ago, firedog said:

True. But there’s a bigger point you seem to be missing. MQA is marketed and touted by the audio press as “better sounding than CD”; and equal to or better sounding than hi res. All those flowery descriptions of how much better MQA sounds - even than hires versions of the same cuts. 
A lot of the hostility to MQA comes as a result of that. If they had simply introduced it as a compression scheme that’s slightly lossy, but “we think it’s perceptually lossless” when compared to hires, it would be much less of an issue.

And MQA DOES have something to prove when they make claims like that. Let’s see them prove something. 

Does desk pounding count as proof? What if they use a shoe?

[crenca]

Just now, mansr said:

The participants were asked to rate the "clarity" of the material they listened to. There are (at least) two problems here:

1. Clarity is not a well-defined, measurable property. As such, interpretations will vary, adding unnecessary noise to the data.
2. The test failed to establish that a difference was at all discernible before attempting to quantify it.

 

Thanks.  Audiophile descriptive language such as "clarity", "resolution", and the like (to say nothing of descriptors like "musicality") are so variable as to be almost useless.  SBAF and others have made a conscious effort to find a common agreement through common experience with some (if limited) success, but yes in a study like this...

[Paul R]

1 minute ago, firedog said:

 

True. But there’s a bigger point you seem to be missing. MQA is marketed and touted by the audio press as “better sounding than CD”; and equal to or better sounding than hi res. All those flowery descriptions of how much better MQA sounds - even than hires versions of the same cuts. 
A lot of the hostility to MQA comes as a result of that. If they had simply introduced it as a compression scheme that’s slightly lossy, but “we think it’s perceptually lossless” when compared to hires, it would be much less of an issue.

And MQA DOES have something to prove when they make claims like that. Let’s see them prove something. 

 

Examine that from the other way around Danny. 

 

It is is proven that MQA is not a technically superior format than say, flac based high res 24/96k. 

 

So why do some people prefer the sound of a MQA file?

 

There are conspiracy theories and other even more unlikely theories floating around, but the simple and most logical reason they prefer the sound of a MQA file is that something is causing MQA files to sound better to them.  MQA is not going to do the research to explain why - they do not have to prove it sounds better, because people are agreeing with them.  And they might harm themselves with the results, besides being difficult and expensive to do.

 

Nobody here wants to do the research, because it is hard, expensive, and could possibly help MQA. The reason might turn out to be more of a legal or ethical problem than a technical one anyway. 

 

The arguments about the audio press press are in large part totally unjustified, IMNSHO. YMMV, etc.

 

[Paul R]

17 minutes ago, mansr said:

The participants were asked to rate the "clarity" of the material they listened to. There are (at least) two problems here:

1. Clarity is not a well-defined, measurable property. As such, interpretations will vary, adding unnecessary noise to the data.
2. The test failed to establish that a difference was at all discernible before attempting to quantify it and pick a preference.

 

The testing procedure defined clarity pretty precisely. They provided training for the test subjects too. 

 

Clarity is well defined in audio and the test subjects were not asked to provide objective measurements of their perception, just subjective better or worse. This was experimentally valid since the testing environment was artificial and completely controlled.

[KeenObserver]

27 minutes ago, Paul R said:

 

Hi John. 

 

DRM is a concern, but mostly it is a blue herring. I can not think of any DRM system that was not swiftly bypassed the moment it became a nuisanice. Might be one, but I seriously doubt MQA would be that example. 

 

Yes!  Pay no attention to what's behind the curtain!