MQA is Vaporware

[phosphorein]

1 hour ago, mansr said:

Time and frequency are interchangeable via the Fourier transform. If one is correct, so is the other.

 

He's not asserting anything at all as long as he's not making sense.

 

I've always assumed that Bob S. doesn't understand the Fourier Transform and the same goes for Austin, which is peculiar for a credentialed physicist.

[crenca]

1 hour ago, phosphorein said:

 

I've always assumed that Bob S. doesn't understand the Fourier Transform and the same goes for Austin, which is peculiar for a credentialed physicist.

 

I appears that way does it not?  Actually it is explicit:

 

"Post-Shannon sampling theory relaxes Shannon's requirement that a signal to be sampled—eg, a recording of music—be band-limited to half the sample rate. Relaxing that constraint restores the symmetry between the time and frequency domains that was missing from Shannon's theory."

 

What is really happening?  As @Jim Austinand others like to point out, he is a Fellow, and audio savant, a credentialed and respected member of the community, blah blah blah.  Yet, he is peddling what appears to be a sophisticated version of the laypersons stair step understanding of digital.

[miguelito]

3 hours ago, mansr said:

This isn't him, but I imagine his hands looking something this:

Bob explaining MQA and the audio press/music industry:

[Jim Austin]

What are you folks still on about? Still assuming that all sampling theory ended with Shannon? This place is obviously a massive circle-jerk, but your posts are visible to the outside world, so you run the risk of embarrassing yourselves (those of you not hiding behind a pseudonym, that is). So I'll do you a favor--I'll present some quotes from scientific literature. Then you can dig them out yourself if you care to educate yourself; I don't expect that, but maybe it'll at least convince you to think twice before posting ignorant things. 

 

From an article by Yonina C. Eldar and Tomer Michaeli, published in IEEE Signal Processing Magazine. Both are at Technion, an the senior author--Eldar--has a PhD from MIT.
 

Quote

... Many

of the limitations encountered in current converters is due

to a traditional assumption that the sampling stage needs to

acquire the data at the Shannon-Nyquist rate, corresponding

to twice the signal bandwidth [6]–[8]. To avoid aliasing, a

sharp low-pass filter (LPF) must be implemented prior to

sampling. The reconstructed signal is also a bandlimited

function, generated by integer shifts of the sinc interpolation

kernel.

 

A major drawback of this paradigm is that many natural

signals are better represented in alternative bases other

than the Fourier basis [9]–[11], or possess further structure

in the Fourier domain. In addition, ideal pointwise sampling,

as assumed by the Shannon theorem, cannot be

implemented. More practical ADCs introduce a distortion

that should be accounted for in the reconstruction process.

Finally, implementing the infinite sinc interpolating kernel

is difficult, since it has slow decay. In practice, much simpler

kernels are used, such as linear interpolation. Therefore,

there is a need to develop a general sampling theory that

will accommodate an extended class of signals beyond bandlimited

functions and will account for the nonideal nature

of the sampling and reconstruction processes.

Sampling theory has benefited from a surge of research

in recent years, due in part to the intense research in wavelet

theory and the connections made between the two fields.

In this survey, we present several extensions of the Shannon

theorem that have been developed primarily in the past two

decades, which treat a wide class of input signals as well

as nonideal sampling and nonlinear distortions. ...

 

If you're feeling ambitious, look up "Sampling Moments and Reconstructing Signals of Finite Rate of Innovation: Shannon Meets Strang–Fix," by Pier Luigi Dragotti et al.; you can find the full citation in the MQA literature. The gist is that under appropriate conditions, signals that have a finite number of degrees of freedom per unit of time, "such as, for example, nonuniform splines or piecewise polynomials," can be perfectly reconstructed even if they are not band-limited ... but you must use appropriate sampling kernels--not the usual sincx function. The class of kernels "that we can use is very rich and includes functions satisfying Strang–Fix conditions, exponential splines and functions with rational Fourier transform." I'm not sure this strictly applies to a music signal, but in practical terms that may not matter--especially since there's no real need to recreate the waveform perfectly. 

 

Finally there's this, by Michael Unser, from IEEE Signal Processing Magazine. My copy of this is in pixels, not characters, so I'll paste in the image: 

...

There's more I could quote. 

 

None of this proves that MQA is a good idea, that it's valid mathematically, or that its application to music makes sense, and it certainly doesn't prove that MQA sounds better. It does however show what some of you still seem to be questioning: sampling theory didn't end with Shannon. The body of theoretical work referred to in the latest MQA article/interview is real. 

 

So stop making fools of yourselves by pretending it doesn't exist. Or, just keep making fools of yourselves. Makes no difference to me. 

 

Oh, and no, I won't provide the citations. If you want to read them you can find them on your own. 

 

jca

[Brinkman Ship]

3 minutes ago, Jim Austin said:

What are you folks still on about? Still assuming that all sampling theory ended with Shannon? This place is obviously a massive circle-jerk, but your posts are visible to the outside world, so you run the risk of embarrassing yourselves (those of you not hiding behind a pseudonym, that is). So I'll do you a favor--I'll present some quotes from scientific literature. Then you can dig them out yourself if you care to educate yourself; I don't expect that, but maybe it'll at least convince you to think twice before posting ignorant things. 

 

From an article by Yonina C. Eldar and Tomer Michaeli, published in IEEE Signal Processing Magazine. Both are at Technion, an the senior author--Eldar--has a PhD from MIT.
 

 

If you're feeling ambitious, look up "Sampling Moments and Reconstructing Signals of Finite Rate of Innovation: Shannon Meets Strang–Fix," by Pier Luigi Dragotti et al.; you can find the full citation in the MQA literature. The gist is that under appropriate conditions, signals that have a finite number of degrees of freedom per unit of time, "such as, for example, nonuniform splines or piecewise polynomials," can be perfectly reconstructed even if they are not band-limited ... but you must use appropriate sampling kernels--not the usual sincx function. The class of kernels "that we can use is very rich and includes functions satisfying Strang–Fix conditions, exponential splines and functions with rational Fourier transform." I'm not sure this strictly applies to a music signal, but in practical terms that may not matter--especially since there's no real need to recreate the waveform perfectly. 

 

Finally there's this, by Michael Unser, from IEEE Signal Processing Magazine. My copy of this is in pixels, not characters, so I'll paste in the image: 

...

There's more I could quote. 

 

None of this proves that MQA is a good idea, that it's valid mathematically, or that its application to music makes sense, and it certainly doesn't prove that MQA sounds better. It does however show what some of you still seem to be questioning: sampling theory didn't end with Shannon. The body of theoretical work referred to in the latest MQA article/interview is real. 

 

So stop making fools of yourselves by pretending it doesn't exist. Or, just keep making fools of yourselves. Makes no difference to me. 

 

Oh, and no, I won't provide the citations. If you want to read them you can find them on your own. 

 

jca

Welcome back. You telling folks not to make fools of them selves..oh that is JUST too rich. And not a good way to be taken seriously.

 

 

[Brinkman Ship]

9 minutes ago, miguelito said:

Bob explaining MQA and the audio press/music industry:

LOL...I need that..!

[phosphorein]

3 minutes ago, Jim Austin said:

What are you folks still on about? Still assuming that all sampling theory ended with Shannon? This place is obviously a massive circle-jerk, but your posts are visible to the outside world, so you run the risk of embarrassing yourselves (those of you not hiding behind a pseudonym, that is). So I'll do you a favor--I'll present some quotes from scientific literature. Then you can dig them out yourself if you care to educate yourself; I don't expect that, but maybe it'll at least convince you to think twice before posting ignorant things. 

 

From an article by Yonina C. Eldar and Tomer Michaeli, published in IEEE Signal Processing Magazine. Both are at Technion, an the senior author--Eldar--has a PhD from MIT.
 

 

If you're feeling ambitious, look up "Sampling Moments and Reconstructing Signals of Finite Rate of Innovation: Shannon Meets Strang–Fix," by Pier Luigi Dragotti et al.; you can find the full citation in the MQA literature. The gist is that under appropriate conditions, signals that have a finite number of degrees of freedom per unit of time, "such as, for example, nonuniform splines or piecewise polynomials," can be perfectly reconstructed even if they are not band-limited ... but you must use appropriate sampling kernels--not the usual sincx function. The class of kernels "that we can use is very rich and includes functions satisfying Strang–Fix conditions, exponential splines and functions with rational Fourier transform." I'm not sure this strictly applies to a music signal, but in practical terms that may not matter--especially since there's no real need to recreate the waveform perfectly. 

 

Finally there's this, by Michael Unser, from IEEE Signal Processing Magazine. My copy of this is in pixels, not characters, so I'll paste in the image: 

...

There's more I could quote. 

 

None of this proves that MQA is a good idea, that it's valid mathematically, or that its application to music makes sense, and it certainly doesn't prove that MQA sounds better. It does however show what some of you still seem to be questioning: sampling theory didn't end with Shannon. The body of theoretical work referred to in the latest MQA article/interview is real. 

 

So stop making fools of yourselves by pretending it doesn't exist. Or, just keep making fools of yourselves. Makes no difference to me. 

 

Oh, and no, I won't provide the citations. If you want to read them you can find them on your own. 

 

jca

 

Mr. Austin,

 

If you are referring to me, you are misrepresenting my post. Furthermore your attitude is condescending and mean-spirited and in my opinion violates the terms of this discourse.

 

[james45974]

11 minutes ago, Jim Austin said:

So stop making fools of yourselves by pretending it doesn't exist. Or, just keep making fools of yourselves. Makes no difference to me. 

When are the manufacturer's who jumped in with MQA going to realize what an embarrassment it is becoming to be associated with it?  Is is really a farce!

[The Computer Audiophile]

32 minutes ago, Jim Austin said:

What are you folks still on about? Still assuming that all sampling theory ended with Shannon? This place is obviously a massive circle-jerk, but your posts are visible to the outside world, so you run the risk of embarrassing yourselves (those of you not hiding behind a pseudonym, that is). So I'll do you a favor--I'll present some quotes from scientific literature. Then you can dig them out yourself if you care to educate yourself; I don't expect that, but maybe it'll at least convince you to think twice before posting ignorant things. 

 

From an article by Yonina C. Eldar and Tomer Michaeli, published in IEEE Signal Processing Magazine. Both are at Technion, an the senior author--Eldar--has a PhD from MIT.
 

 

If you're feeling ambitious, look up "Sampling Moments and Reconstructing Signals of Finite Rate of Innovation: Shannon Meets Strang–Fix," by Pier Luigi Dragotti et al.; you can find the full citation in the MQA literature. The gist is that under appropriate conditions, signals that have a finite number of degrees of freedom per unit of time, "such as, for example, nonuniform splines or piecewise polynomials," can be perfectly reconstructed even if they are not band-limited ... but you must use appropriate sampling kernels--not the usual sincx function. The class of kernels "that we can use is very rich and includes functions satisfying Strang–Fix conditions, exponential splines and functions with rational Fourier transform." I'm not sure this strictly applies to a music signal, but in practical terms that may not matter--especially since there's no real need to recreate the waveform perfectly. 

 

Finally there's this, by Michael Unser, from IEEE Signal Processing Magazine. My copy of this is in pixels, not characters, so I'll paste in the image: 

...

There's more I could quote. 

 

None of this proves that MQA is a good idea, that it's valid mathematically, or that its application to music makes sense, and it certainly doesn't prove that MQA sounds better. It does however show what some of you still seem to be questioning: sampling theory didn't end with Shannon. The body of theoretical work referred to in the latest MQA article/interview is real. 

 

So stop making fools of yourselves by pretending it doesn't exist. Or, just keep making fools of yourselves. Makes no difference to me. 

 

Oh, and no, I won't provide the citations. If you want to read them you can find them on your own. 

 

jca

 

Keep digging Jim. The hole is only getting deeper.

 

I won't censure you post. I'd rather it be available for everyone to see for a very long time. 

[Jim Austin]

Just now, The Computer Audiophile said:

 

Keep digging Jim. The hole is only getting deeper.

 

 

So, quite obviously, you're now abandoning any pretense of objectivity and even-handedness. 

[Mordikai]

15 minutes ago, Jim Austin said:

 

So, quite obviously, you're now abandoning any pretense of objectivity and even-handedness. 

Jim- you're the "journalist", there have been plenty of names of very qualified people mentioned who think the MQA science is B.S. You should seek those people out and ask them.

[The Computer Audiophile]

9 minutes ago, Jim Austin said:

 

So, quite obviously, you're now abandoning any pretense of objectivity and even-handedness. 

When you address the CA Community as a circle jerk and pull a Trump move by telling people who use facts that they are incorrect and your wild speculation, backed by nobody willing to comment in public today about MQA claims, you need to be addressed a bit differently. 

 

My comments are about what you said, not about MQA.

 

I'm still waiting for a single expert to address MQA's claims and agree with Bob. Just one. I highly recommend you roam the halls of an audio show and talk to the leading experts in the industry about MQA. Everyone of them will disagree with Bob. Reasons for implementing MQA are a different story, but nobody agrees with his statements about MQA.

[Brinkman Ship]

50 minutes ago, The Computer Audiophile said:

 

Keep digging Jim. The hole is only getting deeper.

 

I won't censure you post. I'd rather it be available for everyone to see for a very long time. 

...take note, he may be going down the path that "everyone with a different view on MQA gets banned" by provoking, and making a self fulfilling prophecy come true so he can tell every one he was banned for being pro MQA>

[Brinkman Ship]

50 minutes ago, Jim Austin said:

 

So, quite obviously, you're now abandoning any pretense of objectivity and even-handedness. 

It seems you decided to take on the tactic of provoking...a way to get banned so you can have "proof" of your theory that pro MQA opinions are not allowed?

 

Is this all you have left?

[mansr]

The Shannon-Nyquist sampling theorem provides a sufficient condition for fixed-interval sampling to fully capture a signal and enable subsequent reconstruction. Later research has defined other conditions allowing certain signals to be accurately captured without fulfilling the Shannon-Nyquist criterion. A search for terms like sparse signal, sparse sampling, compressed sensing, and finite rate of innovation will turn up hundreds of papers spanning decades.

 

None of this is new. The reason it hasn't been applied to audio is that it is unnecessary. An audio signal has such a low bandwidth to begin with that the Shannon-Nyquist requirement is easily met. The data rates involved also pose no problems for processing, transmission, or storage systems. Even if some form of sparse sampling of audio could cut the data rate in half, say, there are good reasons not to do this outside very specific applications. Traditional sampling produces a signal that is easy to process in a multitude of ways (think of all the operations a DAW can do). That isn't necessarily true of sparse sampling. Why should we complicate everything only for the sake of a data rate reduction we don't need? I can't think of a single reason.

 

With MQA it gets worse. Even if it is based on some clever sampling method that really does reduce the data rate, it is still a proprietary system. Nobody can do anything without Bob Stuart's blessing (and the handing over of a tidy sum of money). On top of that, the distribution format has various DRM features that can never be in the interest of consumers.

 

It is clear to me that MQA is a collection of (alleged) features, each designed to appeal to a particular step in the music production and distribution chain. Sound quality for the listener. Data savings for the distributors. DRM and control for the content owners. Money for Bob from all of the above.

 

More damning still is the fact that every time a claim about MQA is poked, it falls apart like a house of cards. Bandwidth reduction? Standard FLAC does better at comparable quality. Sound quality? Accurate comparisons are made extremely difficult (why?), and with the material available, listeners are anything but united in favour of MQA. DRM? We don't know what the labels have been promised, so assessing this is impossible at this time. We do know that every DRM system to date has been broken. The odds are not in MQA's favour. Authentication (a subset of DRM)? Shoddy at best. Mastering engineers report that their work has been MQA'd without their knowledge, let alone approval. Based on my analysis of MQA firmware, tricking it into turning on the blue light is probably not difficult.

 

In the end, what have we got? By the looks of it, MQA either fails to deliver or is outright harmful in every aspect it purports to improve.

[crenca]

@Jim Austin says:

 

"....None of this proves that MQA is a good idea, that it's valid mathematically, or that its application to music makes sense, and it certainly doesn't prove that MQA sounds better. It does however show what some of you still seem to be questioning: sampling theory didn't end with Shannon. The body of theoretical work referred to in the latest MQA article/interview is real."

 

Did you say any of this to Bob S?  Does Bob explain how any of of the theory you cite is applied in MQA?  It could only be on the MQA DSP treatment side, not on the end user DAC side unless some little nuget is still hiding from the reverse engineers.  Why do you and your publication give Bob S coverage (almost all of it uniformly positive) if the theory to application is so tenuous, so ephemeral, so questionable?  Why is there no corroboration from the many other capable Fellows and audio savants in the field?  Why is what Chris C. reporting the actual situation - where is Bob S's and MQA's applied technical collaboration, support, and peer review?

 

Perhaps one question will do:  Do you, Jim Ausin, believe that Bob S is a Galilean figure - persecuted, who is doing his best to mutter "And yet it moves"?

[mansr]

22 minutes ago, Brinkman Ship said:

...take note, he may be going down the path that "everyone with a different view on MQA gets banned" by provoking, and making a self fulfilling prophecy come true so he can tell every one he was banned for being pro MQA>

It should be noted that anti-MQA people have also been banned. Not because of their views on MQA, but because they failed to maintain a civil tone of discourse. Just like everybody who has been banned here.

[Thuaveta]

1 hour ago, Jim Austin said:

 

None of this proves that MQA is a good idea, that it's valid mathematically, or that its application to music makes sense, and it certainly doesn't prove that MQA sounds better. It does however show what some of you still seem to be questioning: sampling theory didn't end with Shannon. The body of theoretical work referred to in the latest MQA article/interview is real. 

 

If I may - if the body of theoretical work that's referred to is, to put it gently, questionably applicable to the point being made, why bother your readers with it, and, assuming it is not applicable, in what way is what Bob S. doing any different then explaining swimming pool accidents in the United States through Nicolas Cage's employment history ?

[crenca]

4 minutes ago, mansr said:

The Shannon-Nyquist sampling theorem provides a sufficient condition for fixed-interval sampling to fully capture a signal and enable subsequent reconstruction. Later research has defined other conditions allowing certain signals to be accurately captured without fulfilling the Shannon-Nyquist criterion. A search for terms like sparse signal, sparse sampling, compressed sensing, and finite rate of innovation will turn up hundreds of papers spanning decades.

 

None of this is new. The reason it hasn't been applied to audio is that it is unnecessary. An audio signal has such a low bandwidth to begin with that the Shannon-Nyquist requirement is easily met. The data rates involved also pose no problems for processing, transmission, or storage systems. Even if some form of sparse sampling of audio could cut the data rate in half, say, there are good reasons not to do this outside very specific applications. Traditional sampling produces a signal that is easy to process in a multitude of ways (think of all the operations a DAW can do). That isn't necessarily true of sparse sampling. Why should we complicate everything only for the sake of a data rate reduction we don't need? I can't think of a single reason.

 

With MQA it gets worse. Even if it is based on some clever sampling method that really does reduce the data rate, it is still a proprietary system. Nobody can do anything without Bob Stuart's blessing (and the handing over of a tidy sum of money). On top of that, the distribution format has various DRM features that can never be in the interest of consumers.

 

It is clear to me that MQA is a collection of (alleged) features, each designed to appeal to a particular step in the music production and distribution chain. Sound quality for the listener. Data savings for the distributors. DRM and control for the content owners. Money for Bob from all of the above.

 

More damning still is the fact that every time a claim about MQA is poked, it falls apart like a house of cards. Bandwidth reduction? Standard FLAC does better at comparable quality. Sound quality? Accurate comparisons are made extremely difficult (why?), and with the material available, listeners are anything but united in favour of MQA. DRM? We don't know what the labels have been promised, so assessing this is impossible at this time. We do know that every DRM system to date has been broken. The odds are not in MQA's favour. Authentication (a subset of DRM)? Shoddy at best. Mastering engineers report that their work has been MQA'd without their knowledge, let alone approval. Based on my analysis of MQA firmware, tricking it into turning on the blue light is probably not difficult.

 

In the end, what have we got? By the looks of it, MQA either fails to deliver or is outright harmful in every aspect it purports to improve.

 

@Jim Austin, @John_Atkinson, @ARQuint, why don't you take this very reasonable (in content, style, word length) summation of the "cons" of MQA, clean it up a bit and publish it?  

[tmtomh]

41 minutes ago, mansr said:

The reason [sparse sampling] hasn't been applied to audio is that it is unnecessary. An audio signal has such a low bandwidth to begin with that the Shannon-Nyquist requirement is easily met. The data rates involved also pose no problems for processing, transmission, or storage systems.

 

Mainly quoting for emphasis: what @mansr writes here is really key to the whole discussion of so-called "post-Shannon sampling theory" that gets trotted out in some MQA discussions.

 

"sparse sampling," "undersampling," "post-Shannon" - this is all a misapplication of digital sampling techniques used when dealing with much higher-frequency waves (for example radio waves), far up in the ultrasonic range.

 

The key thing is that if you have to sample a relatively narrow range of frequencies, but in a high frequency range, you get a lot of waste if you do it the way digital audio is sampled. For example, one common cellular phone standard transmits and receives roughly from 1850 to 1990MHz. The problem with applying traditional Nyquist-Shannon sampling there is that you're wasting data by massively oversampling, since frequencies from zero Hz up to 1849.99999MHz don't need to be sampled for that application.

 

But with audio, not only are the frequencies themselves much lower (and therefore require much lower sampling rates), but the "wasted" frequency band is only the range from 0 to 19.999999999Hz. - practically nothing, in terms of inefficiency of sampling.

 

Even if you want to split up a high-sample-rate audio signal into multiple bands - say up to 22.05kHz, and then the ultrasonics from there up to 48kHz or 96kHz - you're still saving very little data by using undersampling on that higher band, and there's nothing "post-Shannon" about it - you're just using an algorithm to encode the ultrasonics as if they were lower frequencies, and then shift them all up by a fixed adjustment amount during decoding.

 

To put it perhaps more simply, sparse/under-sampling basically just means, "We can mis-encode our frequencies because the we know the precise frequency range they're supposed to be in, and therefore we can precisely quantify the error in our mis-encoding and compensate for it."

[Ralf11]

2 hours ago, Jim Austin said:

 

Oh, and no, I won't provide the citations. If you want to read them you can find them on your own. 

 

jca

 

how about just answering the question asked of you above?

[Mordikai]

sigh... what's the use? Serious questions are raised= MQA proxy disappears. Frustrating.

[Brinkman Ship]

39 minutes ago, mansr said:

It should be noted that anti-MQA people have also been banned. Not because of their views on MQA, but because they failed to maintain a civil tone of discourse. Just like everybody who has been banned here.

Quite. true. I have come to believe that Mr. Connaker  in general bans people not for what they say, but how they say it. Pro this or Anti that.

[Brinkman Ship]

Just now, Mordikai said:

sigh... what's the use? Serious questions are raised= MQA proxy disappears. Frustrating.

proxy=shill shill=proxy proxy=shill shill=proxy

 

 

[mansr]

4 minutes ago, tmtomh said:

For example, one common cellular phone standard transmits and receives roughly from 1850 to 1990MHz. The problem with applying traditional Nyquist-Shannon sampling there is that you're wasting data by massively oversampling, since frequencies from zero Hz up to 1849.99999MHz don't need to be sampled for that application.

Even this is handled by Shannon. The sample rate needs to exceed twice the bandwidth, not twice the maximum frequency. If the signal is known to have a lower as well as upper bound, a matching band-pass filter is all you need for sampling and reconstruction. The spectrum of the sampled version of the signal might be mirrored, but this can be dealt with.