MQA is Vaporware

[The Computer Audiophile]

13 minutes ago, crenca said:

 

Wrong again.  The truth of MQA was never "in the middle", nor is sugar always the best strategy.  It's a both/and not an either or.  Sugar sometimes works, but how did that work out for you at RMAF?  Fail! as the kids say.    MQA is counting on the status quo audiophile sugar - it's a play on the sugar.  Neither yourself, nor Jud, nor tmtomh are going to convince the Lee's, ARQuint's, and JA's of the world that they are wrong about MQA.  If they don't know already, they ain't never going to know.  What they know and what they say are always going to be different because their whole position/livelihood depends on them not being wrong about MQA.

 

ARQuint civility sugar play is just that, a play.  These guys are going to work that play for the next 4 years just as they have for the last 4 years.  Sometimes, the truth needs a little spice...

I’d say my RMAF sugar showed exactly who MQA Ltd is. Believers weren’t going to believe me or facts any way. Now we have the MQA team on video acting like children. I’d say that’s a win. 

 

You actually argue my point well. Neither jud not tmtom etc... will change their minds and as you say if they don’t know already they ain’t ever going to know. Well, there you have it. They ain’t ever going to know if you beat them with a spicy hammer. 

 

Minds aren’t going to change with spice. Reasonable people are the only people who will change their minds and they usually accept sugar much better than spice. 

[rando]

27 minutes ago, crenca said:

Sometimes, the truth needs a little spice...

 

Sometimes the meat of an issue benefits from a little spice to highlight the full range of qualities it contains. Very little, applied at the correct time.

 

Sugar rarely features in anything, bar means to loosen tongues, until everything else has been satisfactorily digested. 

 

 

 

[crenca]

Since when is Audiophiledom built on reason?

 

We will have to agree to disagree on this one.  Reason is necessary, but not sufficient.  MQA, and all the other fraudulent voodoo in audio relies on distortion of the truth - it's not a competing signal/parallel signal.  Lee, Quint, JA - they are (whether they know this or not - and they do) the sources of the distortion.  They distort in accordance with their interests.  The civility complaint is but another play and I am no buying it.

 

Another good thing about MQA is that it is showing not only who MQA is, but who and what Stereophile, TAS, all the rest of the Old Guard trade publications are as well.  I like my crashing and burning Old Guard voodoo salesmen spicy, red chili hot spicy. 😋

[MikeyFresh]

8 minutes ago, crenca said:

They distort in accordance with their interests.

 

No question that this is the case, and those interests are not one in the same as those of consumers.

 

For this simple reason it is likely we'll have to just agree to disagree with them. 

[opus101]

3 hours ago, Lee Scoggins said:

2.  I am still studying how the deblurring works so more on that to come later.

 

 

Going back quite a few months you were going to be doing research into the ringing question. Is there anything of that going to come later perchance?

[crenca]

1 minute ago, MikeyFresh said:

 

No question that this is the case, and those interests are not one in the same as those of consumers.

 

For this simple reason it is likely we'll have to just agree to disagree with them. 

 

I think we can do more than that.  We can truthfully describe what they are.  We can point out their un-civility: how their interests, assumptions, hypocrisy - their distortions are not very nice, not very nice at all.  We disagree yes, but we are not even playing the same game or practicing the same hobby/musical interests.  We are not on the same team.

 

OR...we could simply keep passing them the sugar packets.

[manisandher]

7 hours ago, Lee Scoggins said:

Also, the file size is one quarter of the corresponding hirez file.

 

I'm not seeing this.

 

Here are the hires and undecoded-MQA file sizes for the track I used in the 'apples-to-apples final' thread:

 

Hires:

 

MQA:

 

So more like 55% than 25%.

 

Mani.

[manisandher]

5 hours ago, Jud said:

2 - So why might people like it?

 

Over my 3 'apples-to-apples' threads, the results through blind listening were pretty even between hires and decoded-MQA (in all cases from the same master).

 

For my own part, I preferred MQA in just one of the 3 cases (non-blind of course).

 

Mani.

[Thuaveta]

5 hours ago, The Computer Audiophile said:

I’d say my RMAF sugar showed exactly who MQA Ltd is. Believers weren’t going to believe me or facts any way. Now we have the MQA team on video acting like children. I’d say that’s a win. 

 

You actually argue my point well. Neither jud not tmtom etc... will change their minds and as you say if they don’t know already they ain’t ever going to know. Well, there you have it. They ain’t ever going to know if you beat them with a spicy hammer. 

 

Minds aren’t going to change with spice. Reasonable people are the only people who will change their minds and they usually accept sugar much better than spice. 

 

There seem to be multiple parts to the tactics we're seeing (with MQA and also elsewhere, eh). The way it seems to work is you've got a group of people either spouting inflammatory crap or straight-on gaslighting, getting called out on it, then going out and playing victim when they get called out.

 

Because gaslighting implies a weaponisation of common courtesy, as well as extreme contempt for the person in front of you, there's a point where I feel like you're wrong, and it's something you touch on: if your premise is that the person you're talking to isn't going to budge is correct, and they're trying to make use of the Brandolini's law, then they deserve all the tarring and feathering you can muster.

[k-man]

@manisandher I hope you don't mind me making some correction to your MQA file description:

 

 

 

[gdpr]

8 minutes ago, k-man said:

@manisandher I hope you don't mind me making some correction to your MQA file description:

 

 

Sorry but your message would make more sense if you had corrected the real MQA file description iso of the Hirez one.

Dirk

[k-man]

8 minutes ago, ddetaey said:

Sorry but your message would make more sense if you had corrected the real MQA file description iso of the Hirez one.

Dirk

Very true. 😅

Corrected.

[John Dyson]

The frightening thing about MQA is that it is NOT vaporware, but a very destructive thing.  From my view, I can imagine all of the leaked DolbyA encoded material being placed on MQA form, and not being able to decode it.  That isn't the only problem, but would be my own problem.

MQA only takes away freedom -- there is zero benefit from a quality standpoint.  Limiting 'compression' to the high frequencies/small details is exactly the opposite from the optimum behavior.  For optimum quality/minimum signal energy one would want to compress the high energy data -- where the precision is very crticial also.  By compressing the high energy material, and then removing it from the uncompressed stream -- it allows for less precision & signal energy in the uncompressed stream.  Also -- secondarliy different kinds of compression are more optimum for different kinds of data.  Data with high variance benefits more from a different algorithm than data with a small variance.  The benefit from compressing data with a large variance is MORE difficult to achieve.

MQA seems to be getting it absolutely opposite from optimum usage of resources vs quality.

 

Ahhh.  The compatibility issue -- compatible is of lower quality...  Not good either.

 

Additionally, the other benefits don't seem to manifest for the user side of things.  MQA is only bad for the user base.  Other kinds of compression/data reduction would be much more beneficial -- maybe a super-opus which is 100% optimized for quality, but not quite flac could be a good thing.   MQA simply seems to be optimally unoptimum.

 

John

[Jud]

38 minutes ago, John Dyson said:

The frightening thing about MQA is that it is NOT vaporware, but a very destructive thing.  From my view, I can imagine all of the leaked DolbyA encoded material being placed on MQA form, and not being able to decode it.  That isn't the only problem, but would be my own problem.

MQA only takes away freedom -- there is zero benefit from a quality standpoint.  Limiting 'compression' to the high frequencies/small details is exactly the opposite from the optimum behavior.  For optimum quality/minimum signal energy one would want to compress the high energy data -- where the precision is very crticial also.  By compressing the high energy material, and then removing it from the uncompressed stream -- it allows for less precision & signal energy in the uncompressed stream.  Also -- secondarliy different kinds of compression are more optimum for different kinds of data.  Data with high variance benefits more from a different algorithm than data with a small variance.  The benefit from compressing data with a large variance is MORE difficult to achieve.

MQA seems to be getting it absolutely opposite from optimum usage of resources vs quality.

 

Ahhh.  The compatibility issue -- compatible is of lower quality...  Not good either.

 

Additionally, the other benefits don't seem to manifest for the user side of things.  MQA is only bad for the user base.  Other kinds of compression/data reduction would be much more beneficial -- maybe a super-opus which is 100% optimized for quality, but not quite flac could be a good thing.   MQA simply seems to be optimally unoptimum.

 

John

 

I think @The Computer Audiophile hit this exactly right in his presentation.  In a world where we think nothing of streaming video on and through our mobile phones, where you can pick up 4-6 TB of storage for less than $150 at a local store or online, and where both these situations are going in the direction of faster and larger, cheaper - MQA is a solution whose time has come and gone.

 

Pay attention to the presentation video: Chris's slide about 5G and mobile/wi-fi speeds getting faster was something that really set off the MQA folks.  Now whether you know anything about the technical side of MQA or not, you surely see mobile/wi-fi/home internet speeds getting faster, and you surely know that streaming video, let alone streaming audio, is ubiquitous on the Web, home or mobile.  And you therefore can understand that there is no consumer value to MQA's file size minimization, whether via compression or the ADPCM (adaptive differential pulse code modulation) @Paul R mentioned.  But right there on the video you see MQA arguing against the faster cheaper future everyone knows is coming (really, for streaming audio purposes, is already here), because they realize it means at least half their supposed value proposition is quite evidently worthless.

[Jud]

4 hours ago, manisandher said:

 

Over my 3 'apples-to-apples' threads, the results through blind listening were pretty even between hires and decoded-MQA (in all cases from the same master).

 

For my own part, I preferred MQA in just one of the 3 cases (non-blind of course).

 

Mani.

 

I'm shocked (shocked!) that a blinded audio A/B comparison gave no better than chance results.  😉

[Lee Scoggins]

9 hours ago, Jud said:

Some additional points about MQA:

 

1 - Why can't MQA be done "better"? 

 

a -- To put what I said in my long post in different terms, every filter must strike some mathematical balance between "frequency distortions" (aliasing, imaging) and "time distortions" (ringing).  https://en.wikipedia.org/wiki/Conjugate_variables  This has been known for the better part of two centuries.  https://en.wikipedia.org/wiki/Fourier_transform#History

 

b -- It is the essence of MQA that it seeks to absolutely minimize time distortions.  If you don't have this, you don't have MQA.  This extreme approach to filtering inevitably means mathematically that it *must* maximize imaging and aliasing.  And this is indeed what we see in the frequency responses of all MQA filters.

 

c -- As noted by me and others, lossless compression cannot satisfactorily protect how MQA processing works, its intellectual property.  In order for MQA to make money in the way it is intended to do, by utilizing proprietary processing that protects its IP and allows the labels to hold back full resolution digital files (since they are still convinced, or at least trying to convince their bosses and investors, that piracy is the reason they aren't more successful), it must use lossy compression.

 

d -- As Miska has mentioned and mansr has confirmed, MQA uses part of the 24- or 16-bit word length to hold information about MQA decoding.  Thus, although the full dynamic range of 16 bit material is never used (let alone 24-bit, which is pretty well impossible, since you're into the heat noise of the electronics by that point), MQA does limit potential dynamic range versus RedBook material in particular.

 

e -- Combine filtering that must produce distortion with lossy compression and a higher noise floor with RedBook material, and the result *must* to a mathematical certainty be further from the original than better filtering and lossless (if any) compression, i.e., genuine hi res or RedBook played back with a reasonably good DAC or software player.

 

f -- Now you may not hear the distortion, or the losses due to compression, or you may not mind them if you do hear them, but MQA is not marketed as "We screwed up your music, but not so badly you can actually hear it."  It's marketed as closer to the original, which as we've just seen can't be true on a sheer mathematical basis.  No real room for argument, this is as plain a mathematical fact as 2+2=4 (though I do love the T-shirt that says "2+2=5, but only for higher values of 2").

 

2 - So why might people like it?

 

a -- We've gone through a variety of reasons - subjective predisposition, subconscious pressure of social expectations, a liking for a very slightly warmer, more exciting sound that distortion might produce.  But I want to add one more, that was first brought to our attention by @firedog.  It is that many MQA DACs do not stop using their MQA filters once MQA material is no longer being played.  Giving a DAC the ability to stop means extra expense and complexity.  So what many people who think they're comparing MQA and normal hi res are actually doing is comparing material that may have been prepared specifically for MQA's DAC filters (it may have the highs and low ultrasonics rolled off by the ADC) to material that doesn't play well with those same filters (real hi res, that has response in the low ultrasonic, causing intermodulation and harmonic distortion when used with MQA's filters).  In that situation, you would in fact be hearing more distortion with genuine hi res.

 

This is a good summary in many ways, but how do you explain when people like Peter McGrath and I who both have a background in recording acoustic music hear the MQA improvements in a straight comparison where the MQA filters are turned off so it's an apples to apples comparison?

 

In other words, couldn't Bob's filter actually be improving the aspects of the performance & recording that are more important than the tradeoffs of things like aliasing?  Bob has mentioned in the past that the timing elements are far more important based on published perceptual studies.  He seems to be making a conscious decision here on priority.

 

And finally, how do we know that the bits that Bob's algorithm is not using are audible at all?  It seems very low level noise so arguably nothing is lost.  So you have a technically lossy algorithm that captures all of the music but still has the benefits of the timing improvements.

[Lee Scoggins]

9 hours ago, The Computer Audiophile said:

I’d say my RMAF sugar showed exactly who MQA Ltd is. Believers weren’t going to believe me or facts any way. Now we have the MQA team on video acting like children. I’d say that’s a win. 

 

This really an accurate description of what happened at all Chris.  In talking to Ken and Mike, they were frustrated that your slides were so one-sided and ignored the materials that they had sent you that addressed the points you were making.  And as my ignored response showed, there were valid counterpoints to almost every slide.

 

If you had been more objective and presented both sides, then said, "I believe this side, because..." then you would have been more credible, even if you chose the anti-MQA viewpoint.  Instead the slides (as I detail herein) did not address many of the counterpoints.

 

That frustrated Ken and Mike and understandably so. 

[Lee Scoggins]

17 minutes ago, Jud said:

 

I think @The Computer Audiophile hit this exactly right in his presentation.  In a world where we think nothing of streaming video on and through our mobile phones, where you can pick up 4-6 TB of storage for less than $150 at a local store or online, and where both these situations are going in the direction of faster and larger, cheaper - MQA is a solution whose time has come and gone.

 

Pay attention to the presentation video: Chris's slide about 5G and mobile/wi-fi speeds getting faster was something that really set off the MQA folks.  Now whether you know anything about the technical side of MQA or not, you surely see mobile/wi-fi/home internet speeds getting faster, and you surely know that streaming video, let alone streaming audio, is ubiquitous on the Web, home or mobile.  And you therefore can understand that there is no consumer value to MQA's file size minimization, whether via compression or the ADPCM (adaptive differential pulse code modulation) @Paul R mentioned.  But right there on the video you see MQA arguing against the faster cheaper future everyone knows is coming (really, for streaming audio purposes, is already here), because they realize it means at least half their supposed value proposition is quite evidently worthless.

 

This is an analogy that is too simple and misleading.  You cannot look at this solely on an individual consumer's experience and make the claim when the whole thing is about distributing music at scale.

 

Running a streaming service requires lots of bandwidth and lots of storage.  While individually both are cheap, there are many issues that these services face at the vast scales they work in.  Look at amazon or netflix.  Many times the video will stutter at night when bandwidth is heaviest. As for downloads, Qobuz hangs up on me all the time although I have fiber-optic speed.  24/192 I think requires around 9 mbps download speeds.  It's provable that many areas in the country don't have that.

 

As long as these problems persist, you cannot argue that file size compression is an idea whose time "has come and gone."

[John Dyson]

25 minutes ago, Jud said:

 

I think @The Computer Audiophile hit this exactly right in his presentation.  In a world where we think nothing of streaming video on and through our mobile phones, where you can pick up 4-6 TB of storage for less than $150 at a local store or online, and where both these situations are going in the direction of faster and larger, cheaper - MQA is a solution whose time has come and gone.

 

Pay attention to the presentation video: Chris's slide about 5G and mobile/wi-fi speeds getting faster was something that really set off the MQA folks.  Now whether you know anything about the technical side of MQA or not, you surely see mobile/wi-fi/home internet speeds getting faster, and you surely know that streaming video, let alone streaming audio, is ubiquitous on the Web, home or mobile.  And you therefore can understand that there is no consumer value to MQA's file size minimization, whether via compression or the ADPCM (adaptive differential pulse code modulation) @Paul R mentioned.  But right there on the video you see MQA arguing against the faster cheaper future everyone knows is coming (really, for streaming audio purposes, is already here), because they realize it means at least half their supposed value proposition is quite evidently worthless.

Okay -- I do agree that data compression of audio is of less importance than today.  However, there are still cases where compression might be useful.  Maybe a LOT of compression of good quality might be more useful than some compression at the highest quality.  Where bandwidth reduction is needed -- then a lot of reduction is needed.  Data reduction for storage reasons is of less importance (or the need is less common.)

When I need data storage size reduction, I am happy with flac -- it does just enough to be worthwhile, and except for material that has data out of range, is good enough to store the data that I normally use floating point for manipulating.  Usually, flac is otherwise good enough.

I have to admit, sometimes I do play with flac's apodization options to maximize compression -- but the benefit is similar to that of twiddling thumbs.

 

MQA is optimally unoptimum in yet another way -- time has come and gone -- after digesting the informaion/thinking about it --  I do agree.

 

John

[Thuaveta]

14 minutes ago, Lee Scoggins said:

Instead the slides (as I detail herein) did not address many of the counterpoints.

 

That frustrated Ken and Mike and understandably so. 

 

So now, not only should these people control the way music is distributed, but they also have a right to control the discussion of said format, and if they're not given that, they're right in throwing all the hissy fits they want... good thing nobody's being saying there's a problem with MQA Ltd's behaviour, or aims.

[mansr]

16 minutes ago, Lee Scoggins said:

Peter McGrath and I who both have a background in recording acoustic music

Do you see what he did there?

[Shadders]

3 minutes ago, Lee Scoggins said:

In other words, couldn't Bob's filter actually be improving the aspects of the performance & recording that are more important than the tradeoffs of things like aliasing?

Hi,

You seem to believe what you are told by MQA Ltd about the deblurring, when on this forum it has been proven that the blurring does not actually exist. You seem to accept aliasing as inconsequential, because you are told to by MQA Ltd, yet aliasing is one of the most obvious basic examples of a failed audio system design.

 

We have current open standards for audio processing that are in use, which meets all the needs for high resolution in a correctly designed audio system, which includes existing physical formats and internet delivery.

 

Compared to the open standards for standard resolution and high resolution, MQA is a retrograde step in audio, in both performance, and audio system design.

 

So, given that the MQA Ltd claims have been disproved, that the MQA format is a retrograde design which performs worse than existing standards, why do you continually repeat the MQA Ltd false claims, promoting it as an optimal system ?

 

Regards,

Shadders.

[The Computer Audiophile]

16 minutes ago, Lee Scoggins said:

then you would have been more credible

I’m still searching for those people who think I wasn’t credible during that presentation. Let’s see, Ken, Mike, you, and Derek. That’s about it. 

 

You don’t have counterpoints Lee.  Your use of facts is on the same level as our President and it’s getting annoying. 

[mansr]

11 hours ago, shadowlight said:

I had asked previously but have yet to get a response.  Is the DRM functionality built in to the spec of MQA (there have been multiple mentions about protecting crown jewels)?  If yes, what's to prevent the studio's from waking up tomorrow morning and saying we want to flip that switch.

There is no published spec for MQA, but we have reverse engineered some of it. Whether it is/has/contains DRM depends on your definition of DRM. In the sense that the content is locked behind a secret decoding algorithm requiring a licence payment for access, it is most certainly DRM.  Some, mainly apologists, insist that the term DRM only applies to things using "cryptography." By this they mean, presumably, that it has to employ algorithms typically used for encryption (AES, RSA, etc).

 

The authentication in MQA is based on RSA. Of course, one can ignore the signatures and still play the file, so this aspect probably shouldn't qualify as DRM.

 

While reverse engineering the decoder, I discovered a provision for having the whole stream, including low frequencies, encrypted with the Salsa20 cipher. That should satisfy even the strictest definitions of DRM. However, the support for this feature appears to be incomplete, and MQA insist it was "accidentally" included in an early release of the decoder. Whatever the state of this may be in the latest decoder versions, the fact it ever existed, in any form, makes it pretty clear where their intentions are. If the labels requested it, MQA would likely finish the feature so it could be added by a software/firmware update.

[Shadders]

On 3/11/2019 at 8:49 PM, manisandher said:

 

For a while, I've been wondering what MQA actually does to 24/96 (and below) resolution files. I mean, here's the difference between an original 24/96 hires and the equivalent decoded MQA from the same master:

 

 

The difference is pretty low in level. I totally accept that it's accurate to call MQA 'lossy', but looking at this difference, is it fair to describe MQA as some sort of lossy compression scheme? It just doesn't look to me as if that's what it's doing. Those 'humps' in the upper regions seem to be intrinsic to what MQA is doing. I'd be interested in knowing what sort of processing could be causing them. Removing aliasing products due to the ADC used perhaps?

 

There's no doubt that for files above 24/96 resolution, MQA is definitely lossy. As has been shown before, at these resolutions, decoded and rendered MQA doesn't reproduce any signal above 88.2/96 kHz in the original hires faithfully. But generally, all there is above 88.2/96 kHz in the original hires is noise (often at quite high levels). Does it matter that this noise is not reproduced losslessly, but is replaced instead by imaging artefacts above 88.2/96 kHz, due to the non-ringing filters used by MQA renderers?

 

These are genuine questions that I'd like thoughts on.

 

Mani.

Hi,

In the spectrum, i can see some spikes - i think later you stated they were artefacts from the processing and not in the file.

 

Does anyone have the same plots from processing which show the disastrous ringing alluded to by MQA Ltd ?

 

Surely, if it was as ubiquitous as MQA Ltd state, then there will be a plethora of examples of spikes at the cut off frequency ????

 

Don't seem to be any examples on the MQA Ltd website of specific albums/tracks with ringing inherent. I wonder why that is ???

 

Regards,

Shadders.