John Atkinson: Yes, MQA IS Elegant...

[Miska]

41 minutes ago, John_Atkinson said:

"MQA, _if_ it operates as describes and as I investigated in my article, is the only commercially available end-to end solution. (Unless you consider very high-bitrate DSD implemented with complementary first-order low-pass filters.)"

 

Can you elaborate a bit more? Why would you need very high-bitrate DSD with first-order low-pass? That wouldn't comply with the D/A side specification. And even spec compliant DSD64 has better time domain performance than MQA. With DSD, at the ADC side sure, you don't need more than first order anti-alias filter, because Nyquist is already up at 1.4 MHz. At D/A side you need more, but less than with PCM. With PCM too you also need to take into account analog reconstruction filter that is always needed to remove images at multiples of the output PCM rate. One thing most PCM DACs miserably fail.

 

Of course, DSD is already established format with lot of material and not anymore encumbered with DRM or any other problematic "end-to-end" control thingies attempting to limit people's use of their content.

 

 

[mansr]

30 minutes ago, John_Atkinson said:

Please note that time-domain performance was the context for this posting to the Audio Asylum, not DRM, not the possibility of aliasing, not the file size, not the lossy vs lossless argument. In that context, I wrote "MQA, _if_ it operates as describes and as I investigated in my article, is the only commercially available end-to end solution. (Unless you consider very high-bitrate DSD implemented with complementary first-order low-pass filters.)"

You might as well be musing over what might be  if the Earth is flat. While that might be entertaining, it has no relation to actual reality. It has been abundantly clear for a long time that MQA does few or none of the things claimed of it. It is thus rather disingenuous to hold it up as the "only commercially available solution," no matter how many ifs you couch it in.

 

30 minutes ago, John_Atkinson said:

As you appear to be arguing with that assertion, what other combinations of commercially available A/D converter and D/A conversion, other than MQA, Ayre's experimental filters, or, possibly, high bit-rate DSD, give you perfect behavior in the time domain from analog original signal to the analog reconstruction of that signal?

Why would you need matched A/D and D/A converters? That isn't how sampling works.

[John_Atkinson]

3 minutes ago, mansr said:

Why would you need matched A/D and D/A converters? That isn't how sampling works.

 

This is explained in the article of mine on stereophile.com that initiated this thread. I acknowledge that some might not want to read that article and by doing so gift us page views, but so it goes.

 

John Atkinson

Editor, Stereophile

[crenca]

11 minutes ago, mansr said:

Why would you need matched A/D and D/A converters? That isn't how sampling works.

 

That is the idea that John A and the like have lapped up from MQA's marketing speak is it not (i.e. this alleged link between A/D to D/A - "end to end")?  On the one hand, John A, Jim A, and the like seem to understand that the nature of sampling in that there is a trade off between time and frequency domains, and on the other hand they appear to be trying to justify and support the idea that MQA has "fixed" this reality through some fundamental new discovery in signal processing, one in which the company that owns this discovery (hidden as it is behind IP) has mysteriously bypassed the opportunity to make $billions$ from sectors such as the military industrial complex and has instead focused on a niche of a niche of the audio industry....

 

As Doug S says, they have been "duped" and they simply don't know what to do except keep digging...

[mansr]

12 minutes ago, John_Atkinson said:

This is explained in the article of mine on stereophile.com that initiated this thread. I acknowledge that some might not want to read that article and by doing so gift us page views, but so it goes.

There can be no explanation for that which is not true.

[rickca]

6 minutes ago, mansr said:

There can be no explanation for that which is not true.

Maybe MQA needs to hire Rudy Giuliani.  He just said truth isn't truth.  Nobody can split hairs like a lawyer/politician.

[Miska]

43 minutes ago, John_Atkinson said:

This is explained in the article of mine on stereophile.com that initiated this thread. I acknowledge that some might not want to read that article and by doing so gift us page views, but so it goes.

 

I went through the original published article, but couldn't locate that.

 

The article in itself is praise for aliasing distortion.

 

And self-contradictory just like MQA itself too. Excessively slow roll-off filter like MQA's that begins to roll-off at least octave before the Nyquist "to preserve transients" actually does exactly the opposite. Since transients are defined primarily by the high frequency harmonics (consider for example harmonic structure of sine vs square wave), those filters instead of ringing slow down the rise time of the transient edge. In effect they blur the transients by slowing down the rise time.

 

For example in Fig 10 the main lobe of transient is narrower than in Fig 11. So the "Listen" filter slows down the transient compared to "Measure" filter. I'd be also curious to see spectrogram of 0 - 100 kHz full level linear sweep with both filters at 96 kHz sampling rate.

 

From signal processing point of view, I just find it very funny that someone describes an all-pass filter as one having the perfect time domain response - because it doesn't filter/do anything! Function of a filter that does essentially nothing is worthless in this context.

 

From amplifier side, I personally prefer sound of amplifiers with faster rise times instead of the slow ones. (one could take this to extreme and compare something like Spectral gear to Naim)

 

[NOMBEDES]

On 8/15/2018 at 8:43 AM, Brinkman Ship said:

why are we surprised when Stereophile employs one of the kings of pseudo science, Fremer? there has never been a fantastical claim by manufacturers he did not parrot. Cable with internal vacuums anybody>?

 

 

Fremer?  Are they cutting vinyl with MQA now?  Mike should stick to the 20th Century. 

[FredericV]

The end to end argument requires drivers which do not smear the time domain:

Example: bending wave driver:

 

Ribbons:

 

vs dynamic drivers (e.g. Wilson (source ) )

 

Most dynamic drivers can't stop the motion immediately, therefore never being able to recreate the time domain accurately. Most tweeters are just not fast enough to correctly playback non-periodic sounds & events.

Bass reflex also has an added time delay, therefore smearing the time domain.

Fix the speakers first, the errors here are much larger than what MQA is trying to solve.

Most speakers fail this test:

 

They can't reproduce the point source of the clapper hitting the bell, while they can reproduce the resonance of the bell. Frequency domain is not the issue here, but the time domain is.

 

[sphinxsix]

1 hour ago, FredericV said:

Fix the speakers first, the errors here are much larger than..

...any other errors in the audio system.

 

BTW I think that the claim that MQA is elegant is an understatement.

                                              

                                               IMO MQA is :

 

                                       

 

                                                          

[Ryan Berry]

On ‎8‎/‎15‎/‎2018 at 10:38 AM, Brinkman Ship said:

With all due respect to Ayre, the fact their ADC did not sell, proves it provided a solution the industry did not need, or was not willing to pay for. And that applies 10,000 x more to MQA.

 

Perhaps more accurately is that Ayre has never believed in forcing people to use our technology, which would have been required to implement such a filter.  To take advantage of the complimentary filter with the QA-9, every recording would have to have been done with the QA-9.  The filter on its own was pretty unappealing when used with recordings done on a different A/D converter, so its usefulness was mainly in a closed system, where the user could record their LP's using a QA-9 and playing them back through an Ayre DAC.  But doing so really reduced the usefulness of the DAC and toggling filters on and off isn't something we've ever really felt was enjoyable.  We'd prefer to just listen to the music, not fumble around with this filter vs. that filter.  In fact, the only reason we even have the Measure filter on our products is for reviewers to keep the Minimum-Phase filter from getting in the way of measurements.  I posted the below on another forum, but in respect to avoiding preferring one forum over another, I'll post it here too:

 

Interesting article by John. While we at Ayre (myself especially) have been very diligent in staying silent regarding MQA since Charley's passing, who was quite...outspoken on the subject; out of respect to Charley, I cannot help but to comment now.

 

Charley's objections about MQA, despite his focus on the topic, was never primarily about the technology. We've known for quite some time that the filters are similar, which, coupled with the other things Ayre does with the DAC, is likely why Charley and I were never leaping at the opportunity to usher in a new era of music and felt that PCM simply sounded better. Where Charley and I disagreed was Charley's opinion that everyone hearing otherwise must be "shills". I believe to this day that people reported what they hear, but there were a number of factors in play that altered their perspective. Having a version of a minimum-phase filter, for example, could make many DACs suddenly sound what is in our opinion better.

 

Charley's objections were the nature in which he perceived that the technology was being handled. The reason Charley spent hours on end explaining what we do and how we do it, John, is because Charley (and by extension, Ayre) believed in making music accessible to as much of the world as possible. Charley strongly believed that music made the world a better place and many of the conflicts around us would have been reduced if people spent more time enjoying music. He would talk at length about why we used this transformer, how our filters work, why we chose solid state over tubes, why we use zero-feedback and fully-balanced designs with discrete components not because he was bragging about it, but because he wanted others -- both listeners and manufacturers -- to compare what we've heard to see the difference for themselves. His hope was that if music sounded better, people would want to listen to it more and the world would be a better place.

 

I don't know if I could see all of that, but I can absolutely agree that we need more music in our lives and that the way to do it is to draw more people to music by making it sound better than it had degraded to during the MP3 era. Ayre never was about getting huge and rich...Charley and I would always disagree about what the "maximum amount" Ayre should make without losing its soul and reason it exists. Instead, Ayre was about bring people music through any means possible.

 

Charley was always outspoken about technologies like DSD and MQA because he felt that they threaten this freedom of music. Embracing one could jeopardize being able to get music in PCM format, for example, and ultimately the world would be the less for it. So he railed and gnashed his teeth and screamed until people would listen to him, because in many ways, Charley was extraordinarily frustrating...but he was almost always right.

 

While you may feel that MQA's implementation of a similar technology to Ayre's Minimum-Phase filter is an unexpected tribute to Charley, John, I in all due respect must disagree. If MQA threatens the availability of music in any other format, then the technology remains a direct slap in the face of everything Charley spent the vast majority of his life working towards. It undermines the hours he spent in pain at a computer desk typing out post after post talking about why Ayre does X or how to improve someone's system through this or that simple step. I cannot predict what the end goal for MQA is and will not speculate on such; but I do know that was how Charley felt right up until his passing and that nothing in the implementation would ever feel like a compliment or tribute to him. My hope is that Charley's feelings in the matter prove to be wrong, but only time will tell there.

[mansr]

6 hours ago, Ryan Berry said:

complimentary filter

With or without spelling mistakes, that concept still doesn't make sense to me in the context of sampling and reconstruction.

[Jud]

1 hour ago, mansr said:

With or without spelling mistakes, that concept still doesn't make sense to me in the context of sampling and reconstruction.

 

Maybe it's right and maybe it's wrong.  Maybe "It doesn't make sense to me" is a reason to assume it's wrong, or maybe it's a reason to learn more about why it might be right.

 

Anyway: I agree with Charles and Ryan about MQA and about SACD, though not about DSD, which doesn't threaten the availability of PCM in any way (except to the extent sigma delta modulation is used in converters, and that battle was basically over decades ago).

[Shadders]

1 hour ago, mansr said:

With or without spelling mistakes, that concept still doesn't make sense to me in the context of sampling and reconstruction.

Hi,

Is this not referring to matched filters - root raised cosine at each modem which when multiplied together is a raised cosine filter - so this is an analogy to the filters in the audio chain, at recording, and playback ?

Regards,

Shadders.

[mansr]

3 minutes ago, Jud said:

Maybe it's right and maybe it's wrong.  Maybe "It doesn't make sense to me" is a reason to assume it's wrong, or maybe it's a reason to learn more about why it might be right.

If I've misunderstood something, I'd like to be educated. The so-called explanations have done nothing of the sort. In fact, I have not seen anything that even attempts to explain what they mean, let alone why it would be useful, only reiterations of the same assertions over and over again.

[mansr]

3 minutes ago, Shadders said:

Is this not referring to matched filters - root raised cosine at each modem which when multiplied together is a raised cosine filter - so this is an analogy to the filters in the audio chain, at recording, and playback ?

I'm familiar with matched filters in other contexts. I just don't see how the concept applies here.

[Shadders]

49 minutes ago, mansr said:

I'm familiar with matched filters in other contexts. I just don't see how the concept applies here.

Hi,

An initial thought is that the filter has a less impact on the recording, and the playback, but once the two are used together in the chain, the result is the combination - which is heard. The combined filter has less impact, than the playback filter being of any design of another manufacturers DAC - which could be any design.

 

All it is, is Ayre presenting the proposal as per MQA, but it is free. Does not have to be Ayre's filter - could be any filter in the recording, but if it is known, then the DAC can be designed appropriately.

 

I do not see the DAC IC (ADC/DAC semiconductor) manufacturers responding to MQA or taking any interest in the audio field - apart from the smaller, more bespoke/esoteric manufacturers.

 

Regards,

Shadders.

[mansr]

18 minutes ago, Shadders said:

An initial thought is that the filter has a less impact on the recording, and the playback, but once the two are used together in the chain, the result is the combination - which is heard. The combined filter has less impact, than the playback filter being of any design of another manufacturers DAC - which could be any design

Which aspects of the filters should be matched, and how, in order to achieve this?

 

The thing is, reconstruction is not really the inverse of sampling. The perfect anti-aliasing filter used during sampling is an ideal low-pass filter. The perfect anti-imaging filter during reconstruction is also an ideal low-pass filter. Obviously, ideal filters are impossible to realise. However, nothing says they need to be matched in any way. If both have a stopband attenuation of at least 200 dB and a passband ripple of at most 0.01 dB, which is easily done with digital filters, there is nothing to worry about.

 

Any matching of the filters would have to be related to their deviations from ideal low-pass filters, so let us consider the effects of imperfections. Lacking stopband attenuation in the anti-aliasing filter will result in high frequencies aliasing irreparably into the audible range. No amount of "matching" in the reconstruction filter can make up for this. Similarly, poor stopband attenuation in the anti-imaging filter creates images of the low frequencies regardless what the anti-aliasing filter looked like. Unlike aliasing, however, images fall outside the audible range and are thus mostly harmless unless they are strong enough to cause audible intermodulation distortion or put excessive power demands on amps and tweeters.

 

In the passband, final output signal contains the combined imperfections (ripple and phase shifts) of the two filters. If the anti-aliasing filter is so poor as to have audible anomalies here, a reconstruction filter carefully chosen to have precisely the opposite ripple, a peak in one wherever the other dips and conversely. A much more practical approach, however, would be to simply correct any such effects before distributing the digital streams rather than trying to get a reconstruction filter with matched horridness into every DAC. Or use a decent ADC in the first place.

 

Now I want to emphasise that the anti-aliasing filters commonly used do not have any passband anomalies that need correcting. It is a non-problem. Moreover, there is no indication, from MQA or Ayre, that this is the kind of matching they're talking about. Indeed, their filter designs, near as I can tell, do quite the opposite, the reconstruction filters exacerbating early roll-off and phase shifts already present from the recording filters.

[Shadders]

27 minutes ago, mansr said:

Which aspects of the filters should be matched, and how, in order to achieve this?

 

The thing is, reconstruction is not really the inverse of sampling. The perfect anti-aliasing filter used during sampling is an ideal low-pass filter. The perfect anti-imaging filter during reconstruction is also an ideal low-pass filter. Obviously, ideal filters are impossible to realise. However, nothing says they need to be matched in any way. If both have a stopband attenuation of at least 200 dB and a passband ripple of at most 0.01 dB, which is easily done with digital filters, there is nothing to worry about.

 

Any matching of the filters would have to be related to their deviations from ideal low-pass filters, so let us consider the effects of imperfections. Lacking stopband attenuation in the anti-aliasing filter will result in high frequencies aliasing irreparably into the audible range. No amount of "matching" in the reconstruction filter can make up for this. Similarly, poor stopband attenuation in the anti-imaging filter creates images of the low frequencies regardless what the anti-aliasing filter looked like. Unlike aliasing, however, images fall outside the audible range and are thus mostly harmless unless they are strong enough to cause audible intermodulation distortion or put excessive power demands on amps and tweeters.

 

In the passband, final output signal contains the combined imperfections (ripple and phase shifts) of the two filters. If the anti-aliasing filter is so poor as to have audible anomalies here, a reconstruction filter carefully chosen to have precisely the opposite ripple, a peak in one wherever the other dips and conversely. A much more practical approach, however, would be to simply correct any such effects before distributing the digital streams rather than trying to get a reconstruction filter with matched horridness into every DAC. Or use a decent ADC in the first place.

 

Now I want to emphasise that the anti-aliasing filters commonly used do not have any passband anomalies that need correcting. It is a non-problem. Moreover, there is no indication, from MQA or Ayre, that this is the kind of matching they're talking about. Indeed, their filter designs, near as I can tell, do quite the opposite, the reconstruction filters exacerbating early roll-off and phase shifts already present from the recording filters.

Hi,

It was an initial thought by me as to what Ayre are referring to.

 

The current ADC's from what has been stated on this forum (by others), sample at a much higher rate than 44.1kHz. Therefore the ideal filter is not required to meet the restricted bandwidth to ensure no aliasing. The filter for ADC can be relaxed - however that is to be designed and implemented. The ideal filter is not required as we are not using a low sample rate.

 

Given this, my interpretation is that Ayre are proposing an end to end system which has minimal impact on the music, but their system is free, in that is not patented or protected by IP.

 

DAC IC manufacturers can follow the proposal, or not.

 

Let us wait and see how @Ryan Berry responds to what he was actually referring to.

 

Regards,

Shadders.

[Brinkman Ship]

11 hours ago, Ryan Berry said:

 

Perhaps more accurately is that Ayre has never believed in forcing people to use our technology, which would have been required to implement such a filter.  To take advantage of the complimentary filter with the QA-9, every recording would have to have been done with the QA-9.  The filter on its own was pretty unappealing when used with recordings done on a different A/D converter, so its usefulness was mainly in a closed system, where the user could record their LP's using a QA-9 and playing them back through an Ayre DAC.  But doing so really reduced the usefulness of the DAC and toggling filters on and off isn't something we've ever really felt was enjoyable.  We'd prefer to just listen to the music, not fumble around with this filter vs. that filter.  In fact, the only reason we even have the Measure filter on our products is for reviewers to keep the Minimum-Phase filter from getting in the way of measurements.  I posted the below on another forum, but in respect to avoiding preferring one forum over another, I'll post it here too:

 

Interesting article by John. While we at Ayre (myself especially) have been very diligent in staying silent regarding MQA since Charley's passing, who was quite...outspoken on the subject; out of respect to Charley, I cannot help but to comment now.

 

Charley's objections about MQA, despite his focus on the topic, was never primarily about the technology. We've known for quite some time that the filters are similar, which, coupled with the other things Ayre does with the DAC, is likely why Charley and I were never leaping at the opportunity to usher in a new era of music and felt that PCM simply sounded better. Where Charley and I disagreed was Charley's opinion that everyone hearing otherwise must be "shills". I believe to this day that people reported what they hear, but there were a number of factors in play that altered their perspective. Having a version of a minimum-phase filter, for example, could make many DACs suddenly sound what is in our opinion better.

 

Charley's objections were the nature in which he perceived that the technology was being handled. The reason Charley spent hours on end explaining what we do and how we do it, John, is because Charley (and by extension, Ayre) believed in making music accessible to as much of the world as possible. Charley strongly believed that music made the world a better place and many of the conflicts around us would have been reduced if people spent more time enjoying music. He would talk at length about why we used this transformer, how our filters work, why we chose solid state over tubes, why we use zero-feedback and fully-balanced designs with discrete components not because he was bragging about it, but because he wanted others -- both listeners and manufacturers -- to compare what we've heard to see the difference for themselves. His hope was that if music sounded better, people would want to listen to it more and the world would be a better place.

 

I don't know if I could see all of that, but I can absolutely agree that we need more music in our lives and that the way to do it is to draw more people to music by making it sound better than it had degraded to during the MP3 era. Ayre never was about getting huge and rich...Charley and I would always disagree about what the "maximum amount" Ayre should make without losing its soul and reason it exists. Instead, Ayre was about bring people music through any means possible.

 

Charley was always outspoken about technologies like DSD and MQA because he felt that they threaten this freedom of music. Embracing one could jeopardize being able to get music in PCM format, for example, and ultimately the world would be the less for it. So he railed and gnashed his teeth and screamed until people would listen to him, because in many ways, Charley was extraordinarily frustrating...but he was almost always right.

 

While you may feel that MQA's implementation of a similar technology to Ayre's Minimum-Phase filter is an unexpected tribute to Charley, John, I in all due respect must disagree. If MQA threatens the availability of music in any other format, then the technology remains a direct slap in the face of everything Charley spent the vast majority of his life working towards. It undermines the hours he spent in pain at a computer desk typing out post after post talking about why Ayre does X or how to improve someone's system through this or that simple step. I cannot predict what the end goal for MQA is and will not speculate on such; but I do know that was how Charley felt right up until his passing and that nothing in the implementation would ever feel like a compliment or tribute to him. My hope is that Charley's feelings in the matter prove to be wrong, but only time will tell there.

Hi Ryan:

 

thanks very much for this post. Very heartfelt, and I personally found it very informative.

 

I see you posted the same at AA, and the responses and your follow ups are also eye opening.

 

Your insights are very much appreciated.

[mansr]

4 minutes ago, Shadders said:

The current ADC's from what has been stated on this forum (by others), sample at a much higher rate than 44.1kHz. Therefore the ideal filter is not required to meet the restricted bandwidth to ensure no aliasing. The filter for ADC can be relaxed - however that is to be designed and implemented. The ideal filter is not required as we are not using a low sample rate.

Sigma-delta ADCs generally sample at 5.6 MHz or higher. This means that they don't need much, or even any, analogue anti-alias filtering. The PCM output is created by digitally low-pass filtering the high-rate output of the sigma-delta converter. This filter has to obey all the usual rules. The advantage of this approach is that a digital filter can get much closer to perfection than an analogue one. On the flip side, the presence of ultrasonic noise from the modulator makes good stopband attenuation even more important. As for commercially available ADC chips, simply recording silence and observing the (mostly lack of) residual noise in the audible band shows that they are perfectly adequate in this regard.

 

27 minutes ago, Shadders said:

Given this, my interpretation is that Ayre are proposing an end to end system which has minimal impact on the music, but their system is free, in that is not patented or protected by IP.

Firstly, this presupposes that there is something lacking in the currently available systems. Secondly, it is simply stated as fact that these unarticulated ailments can be cured by doing something end to end. I see only hand-waving in support of either assumption.

[crenca]

8 minutes ago, mansr said:

Sigma-delta ADCs generally sample at 5.6 MHz or higher. This means that they don't need much, or even any, analogue anti-alias filtering. The PCM output is created by digitally low-pass filtering the high-rate output of the sigma-delta converter. This filter has to obey all the usual rules. The advantage of this approach is that a digital filter can get much closer to perfection than an analogue one. On the flip side, the presence of ultrasonic noise from the modulator makes good stopband attenuation even more important. As for commercially available ADC chips, simply recording silence and observing the (mostly lack of) residual noise in the audible band shows that they are perfectly adequate in this regard.

 

Firstly, this presupposes that there is something lacking in the currently available systems. Secondly, it is simply stated as fact that these unarticulated ailments can be cured by doing something end to end. I see only hand-waving in support of either assumption.

 

1 hour ago, mansr said:

Which aspects of the filters should be matched, and how, in order to achieve this?

 

The thing is, reconstruction is not really the inverse of sampling. The perfect anti-aliasing filter used during sampling is an ideal low-pass filter. The perfect anti-imaging filter during reconstruction is also an ideal low-pass filter. Obviously, ideal filters are impossible to realise. However, nothing says they need to be matched in any way. If both have a stopband attenuation of at least 200 dB and a passband ripple of at most 0.01 dB, which is easily done with digital filters, there is nothing to worry about.

 

Any matching of the filters would have to be related to their deviations from ideal low-pass filters, so let us consider the effects of imperfections. Lacking stopband attenuation in the anti-aliasing filter will result in high frequencies aliasing irreparably into the audible range. No amount of "matching" in the reconstruction filter can make up for this. Similarly, poor stopband attenuation in the anti-imaging filter creates images of the low frequencies regardless what the anti-aliasing filter looked like. Unlike aliasing, however, images fall outside the audible range and are thus mostly harmless unless they are strong enough to cause audible intermodulation distortion or put excessive power demands on amps and tweeters.

 

In the passband, final output signal contains the combined imperfections (ripple and phase shifts) of the two filters. If the anti-aliasing filter is so poor as to have audible anomalies here, a reconstruction filter carefully chosen to have precisely the opposite ripple, a peak in one wherever the other dips and conversely. A much more practical approach, however, would be to simply correct any such effects before distributing the digital streams rather than trying to get a reconstruction filter with matched horridness into every DAC. Or use a decent ADC in the first place.

 

Now I want to emphasise that the anti-aliasing filters commonly used do not have any passband anomalies that need correcting. It is a non-problem. Moreover, there is no indication, from MQA or Ayre, that this is the kind of matching they're talking about. Indeed, their filter designs, near as I can tell, do quite the opposite, the reconstruction filters exacerbating early roll-off and phase shifts already present from the recording filters.

 

Thank you @mansr for these cogent replies.  Also, thank you @Ryan Berryfor your reply and insight into Charley's posititions.  Ryan, as you can tell part of the context of @John_Atkinson selling support of MQA is this alleged "end to end" ADC to DAC linking, hand off, and design "synergy".  Yet as mansr is pointing out there appears to be little (any?) technical explication behind it.  In the end, we are dealing with PCM and PCM is a mature algorithmic process for digital sampling/encoding of audio.  As mansr explains the relationship between the implementation of the ADC and the DAC is not normally described as "end to end" as if it is somehow efficacious (or even necessary) for one to directly account for the other or be sold in a product bundle like MQA tries (but fails) at doing.  As long as each "do their jobs" so to speak, there is no real "end to end" relationship.

 

I suspect that all you/Charlie/Ayre mean to say is the QA-9 samples at 192khz, and thus the "end to end" implication is thus the reconstruction filter on the DAC end has plenty of "headroom" for it's design choices.  If something is implied beyond the sampling rate and/or the good internal design choices of the QA-9, I along with mansr would like to see a technical explanation.

[Le Concombre Masqué]

the one thing I like about mqa is Miska's sweet mqa mp filter 

[Ryan Berry]

5 hours ago, Jud said:

 

Maybe it's right and maybe it's wrong.  Maybe "It doesn't make sense to me" is a reason to assume it's wrong, or maybe it's a reason to learn more about why it might be right.

 

Anyway: I agree with Charles and Ryan about MQA and about SACD, though not about DSD, which doesn't threaten the availability of PCM in any way (except to the extent sigma delta modulation is used in converters, and that battle was basically over decades ago).

 

Ultimately, Charley agreed, which is why DSD appeared in Ayre products.  Before then, and perhaps a bit longer than necessary, Charley held out on the principle of the technology. He never really heard any appear in listening to the files, but once the format was more or less "harmless" in his opinion and a sufficient number of files existed to justify the time spent developing the capability, we began preparing the QB-9s to accept DSD file and released the DSD upgrade about a year later.

[Jud]

6 hours ago, mansr said:

If I've misunderstood something, I'd like to be educated. The so-called explanations have done nothing of the sort. In fact, I have not seen anything that even attempts to explain what they mean, let alone why it would be useful, only reiterations of the same assertions over and over again.

 

3 hours ago, mansr said:

Which aspects of the filters should be matched, and how, in order to achieve this?

 

In the passband, final output signal contains the combined imperfections (ripple and phase shifts) of the two filters. If the anti-aliasing filter is so poor as to have audible anomalies here, a reconstruction filter carefully chosen to have precisely the opposite ripple, a peak in one wherever the other dips and conversely. A much more practical approach, however, would be to simply correct any such effects before distributing the digital streams rather than trying to get a reconstruction filter with matched horridness into every DAC. Or use a decent ADC in the first place.

 

I was in a similar position (actually worse, being a layperson unfamiliar with the technical details) regarding the basis of your objections to the "complementary filter" notion as mentioned in this thread, until I saw your understandable explanation that I've excerpted above.  Thanks.