MQA discussion

rbbert · Sep 26, 2017

If nothing else would convince doubters that MQA is a scam, this should

http://www.theabsolutesound.com/articles/let-the-revolution-begin/

Even implicitly equating Bob Stuart, et al, with Planck, Kelvin and Einstein is so far beyond the pale that one wonders, why?

adyc · Sep 26, 2017

rbbert said:
If nothing else would convince doubters that MQA is a scam, this should

http://www.theabsolutesound.com/articles/let-the-revolution-begin/

Even implicitly equating Bob Stuart, et al, with Planck, Kelvin and Einstein is so far beyond the pale that one wonders, why?

What a joke.

Ken Newton · Sep 26, 2017

I wouldn't penalize Bob Stuart or MQA too harshly for Robert Harley's typically over-the-top hyper-ventilating style.

Ken Newton · Sep 26, 2017

ack said:
The result was that MQA came within spitting distance of what 192kHz, 24-bit PCM is capable of using optimum A/D and D/A conversion filtering."

This seems the key issue. What, exactly, was determined to be the optimum conversion filtering? Which, I presume, refers to the A/D anti-alias, and the D/A anti-image filters. If optimum filtering means filters which prioritize the time-domain, then 192kHz will deliver the same time-domain performance as full MQA, and with lower quantization noise (loss) to boot.

The primary performance benefit of MQA is that it provides an commercial system SPECIFICATION for prioritizing the digital channel for the time-domain. High-Rez is equally capable of being oprimized for the time-domain but High-Rez lacks an commercial/industry specification for this, instead typically being oprimized for the frequency-domain. Which is to say, optimized for maximum flat ultrasonic bandwidth.

rbbert · Sep 26, 2017

I’m not “penalizing” Bob Stuart per se, but rather Meridian’s marketing team and their apparent desperation to get MQA accepted as the default digital delivery system for audiophiles. If MQA is even half as good as its proponents suggest, it should succeed on its own merits; if not, then it shouldn’t.

Your second post alludes to something else which I think should be obvious (but obviously I could be wrong); since whatever filtering and processing MQA is doing is adjusted and calibrated by ear, it is probably something that could be duplicated or even improved on by other digital audio designers and engineers, were they given the apropriate incentives.

GMKF · Sep 26, 2017

Good morning

I use MQA - But I'm not pro MQA or against MQA.

However

"MQA fits the definition of a paradigm shift; the ideas on which it is based are not advances within an existing framework of knowledge"

this sentence is highly debatable.... you are still using the same surroundings in audio. No new magic amp here. He shouldn't have compared it to Einstein. Because he caused a paradigm shift in most areas of physics. A paradigm shift would be finding a perfect amplification method, that would rule all others obsolete in a year. But all of this has never happend in audio... tubes are still around and meaningfull... I also quite like tubes...

What about tape or vinyl or even other digital formats ? Comparisons ? Isnt The ABSOLUTE Sound all about finding it out by listening?
Why not use one of their reference listening setups (which we can acknowledge are quite stout...) and do a full write-up on the differences between MQA and 192kHz and throw in vinyl ? And invite some audio company owners and let them also have a listen... and interview them afterwards

Or maybee I'm beeing just stupid

Have a good one
Max !

Ken Newton · Sep 26, 2017

rbbert,

Marketing teams exist to promote the market success of their company's products. Many new product introduction efforts involve the public hyping of the product to gain consumer attention. If only the benefits and merits of a new product were inherently and widely self-apparent. Any number of otherwise worthy new products fail largely because their associated marketing programs were lacking. This is one of the major functions of sales and marketing, to make clear the benefits and merits of a product to it's intended customers.

As I'd wrote, the performance benefits of MQA can easily be offered via the existing High-Rez PCM format. Well, half of such benefits anyway, as full MQA provides for time-domain optimizing the playback end of the reproduction chain as well, something which High-Rez does not provide for.

Al M. · Sep 26, 2017

Here is my response to Robert Harley's article that I just posted there:

Regarding MQA hyperbole: The question if MQA solves a real problem, or does not do so, is not unequivocally answered in Robert Harley's writings themselves. In his article "Master Quality Authenticated (MQA): The View From 30,000 Feet" he states:

"The other problem with “high-resolution” digital audio is that it didn’t really solve the fundamental problem of why digital sounds the way it does—flat, congested, hard, and glassy. Digital audio requires low-pass “brickwall” filters to prevent a type of distortion called “aliasing.” But these filters introduce ringing, or a smearing of musical signals over time. Despite attempts to minimize this distortion through faster and faster sample rates (the filters for which are less sonically detrimental), digital audio was constrained by the very fundamentals of sampling theorem codified more than fifty years ago."

Compare this questionable statement about digital sound quality with his excellent and in my view largely correct review of the Schiit Yggdrasil DAC which does not decode MQA (I own one myself, and also have repeatedly heard it in a top system, considerably better than mine):

"One of the qualities that makes the Yggy special is its ability to reveal, with startling clarity, individual musical lines within complex arrangements. Every instrument, voice, and sound is spatially and timbrally distinct. This had the effect of revealing each musical line with great precision, and with that precision comes a fuller, richer, and more complex presentation of the composition and arrangement, as well as the intent of each musician. The Yggy is the antithesis of congealed, homogenized, flat, confused, or thick."

[...]

"Although the Yggy has a bold and assertive character, it was never overbearing. In fact, the Yggy encouraged high playback levels, in part because of the smoothness of its upper midrange and its lack of glare in the treble. The top end was extremely clean and well rendered; cymbals had a full measure of energy and verve, yet the sound wasn’t bright. I loved the way the Yggy revealed cymbal work by great drummers; the combination of the dynamic alacrity mentioned earlier with the treble’s pristine quality made such detail especially engaging."

No wonder that Robert Harley does not find it a problem that the DAC does not decode MQA:

"If you’re looking for a DAC that does quad-rate DSD, decodes MQA, offers a volume control, and includes a headphone amp, look elsewhere. But if the very best reproduction of PCM sources is your goal, the Yggdrasil is the ticket. It’s a spectacular performer on an absolute level, and an out-of-this world bargain. The Yggy is not just a tremendous value in today’s DACs, it’s one of the greatest bargains in the history of high-end audio."

***

So which is it? Non-MQA digital sounds flat, congested, hard, and glassy (above cited MQA article by Robert Harley)? Or can the sound of conventional digital be of startling clarity, and the antithesis of congealed, homogenized and flat, with a clean, pristine treble that has no glare, and with no apparent need for MQA since it sounds spectacular on an absolute level (Yggdrasil review by the same author)?

rbbert · Sep 26, 2017

Ken Newton said:
rbbert,

Marketing teams exist to promote the market success of their company's products. Many new product introduction efforts involve the public hyping of the product to gain consumer attention. If only the benefits and merits of a new product were inherently and widely self-apparent. Any number of otherwise worthy new products fail largely because their associated marketing programs were lacking. This is one of the major functions of sales and marketing, to make clear the benefits and merits of a product to it's intended customers.

As I'd wrote, the performance benefits of MQA can easily be offered via the existing High-Rez PCM format. Well, half of such benefits anyway, as full MQA provides for time-domain optimizing the playback end of the reproduction chain as well, something which High-Rez does not provide for.

Yes, poor marketing is certainly largely to blame for the marketplace failure of previous attempts at better digital sound (SACD, DVD-A, Bluray Audio) although each of those have technological obstacles as well (no digital out for DSD, no backwards compatibility with CD for DVD-A, etc). And it is probably unfair to blame Meridian’s marketing team for the inappropriately enthusiastic editorial by Hartley, but it is somewhat amazing to me that anyone still pays any attention to him, not least because of inconsistencies like the one noted by Al above.

Don’t you think almost any DAC can be designed to optimize time domain, not just MQA DAC’s? With the attention often paid to accurate clocking, and the option of selectable filters, it seems likely to me.

microstrip · Sep 26, 2017

rbbert said:
If nothing else would convince doubters that MQA is a scam, this should

http://www.theabsolutesound.com/articles/let-the-revolution-begin/

Even implicitly equating Bob Stuart, et al, with Planck, Kelvin and Einstein is so far beyond the pale that one wonders, why?

Well, I read the article and could not find any "implicit equating" of Bob Stuart to anyone. IMHO Robert Hartley just picks some well known quotes, very often used in the net, to illustrate his points.

People can dislike his style, but he refers to arguments that have often been used by the high-end industry and were often presented and discussed in the old days of the subjective/objective debates of WBF.

And yes, it is as hyperbolic as 99% of what we usually write about audio.

Ken Newton · Sep 26, 2017

rbbert said:
Don’t you think almost any DAC can be designed to optimize time domain, not just MQA DAC’s? With the attention often paid to accurate clocking, and the option of selectable filters, it seems likely to me.

That isn't a simple question to answer. The relevant DAC circuit block determining the DAC's time-domain behavior is it's anti-image filter. Which is also commonly referred to as an 'reconstruction' filter, or as an 'oversampling' filter. This filter is what determines the impulse response (the time-domain behavior) of the DAC.

As you alluded to, many DACs feature selectable filter functions. Usually, such functions include a brickwall sloped filter plus an softer sloped filter. For RBCD playback, a soft sloping filter has the following performance consequences, which are due to the system Nyquist frequency being very closely located to the audio band:
1.) It prioritizes the DAC impulse response for the time-domain.
2.) It suffers a slight in-band high frequency response roll-off.
3.) It opens the door for undesired ultrasonic image products to pass out of the DAC.

High-Rez (native 96kHz or 192kHz) soft sloping anti-image filters, however, retain consequence no. 1, but do not suffer consequences no. 2 and 3 above, because the system Nyquist frequency is much more distantly located from the audio band. In short, yes, non-MQA DACs can be prioritized for the time-domain. Even so, there is an additional benefit brought by MQA, which is that the DAC impulse response is designed as a half of an end-to-end system targeting optimization of the time-domain.

Assuming MQA functions as it's concept was described by Bob Stuart in his AES paper dated 2014, the anti-alias filter impulse response at mastering combines with the anti-image filter impulse response at playback to produce an specific overall total system impulse response. This holistic approach is also possible with High-Rez, but not without an industry standard which optimizes the system time-domain at both the mastering and playback ends first being established and supported by commercial vendors.

rbbert · Sep 26, 2017

However, according to more recent comments by Stuart, the anti-image playback filter characteristics are determined by ear, implying (to me at least) that a similar process could be incorprated into a non-MQA DAC for the end user to choose or not.

Ken Newton · Sep 26, 2017

rbbert said:
However, according to more recent comments by Stuart, the anti-image playback filter characteristics are determined by ear, implying (to me at least) that a similar process could be incorprated into a non-MQA DAC for the end user to choose or not.

I suspect that even though the final filter function was apparently determined by ear, Stuart intends it to define a single end-to-end system impulse response which performs consistently across all MQA enabled gear. Without an industry standard, any non-MQA DAC, though set for proprietary time-domain filter optimization, could deliver an unpredictable sounding result in conjunction with proprietary mastering filters.

Al M. · Oct 6, 2017

Interesting article in Soundstage! HiFi just out this month:

Mismatched Masters and False Frequencies -- Is MQA Better, Worse, or Just Different?

ack · Oct 7, 2017

Al M. said:
Interesting article in Soundstage! HiFi just out this month:

Mismatched Masters and False Frequencies -- Is MQA Better, Worse, or Just Different?

Very well written article Al - thanks. Of particular note is this:

In his article, Tarr summarized what aliasing is on the playback side:
A conventional D-to-A converter for audio will only reproduce signals within a specific frequency range that is determined by the sampling rate.
If the D-to-A converter is fed any recorded frequencies outside of this range, they are interpreted by the converter and mapped to frequencies within this range. This is aliasing -- when one frequency is coded as a different frequency.

...

MQA doesn’t use a brickwall filter. Instead, it has a filter with a very slow rolloff that, MQA says, optimizes the time-domain accuracy, for more accurate reproduction of transients. As with most things in audio, this approach requires tradeoffs. The tradeoff with a slow-rolloff filter like MQA’s is that aliasing can leak through, which is why, in some articles and on some forums, MQA has been described as having a “leaky filter.” This should be no surprise; in MQA patent #WO2015189533 A1, “Digital Encapsulation of Audio Signals,” they state: “We accept that aliasing may take place and are proposing to balance aliasing against ‘time-smear’ of transients due to the lengthening of the system’s impulse response caused by filtering.”

Is the tradeoff worthwhile? It depends on whom you ask. According to John Siau, VP and director of engineering at Benchmark Media Systems, it’s not. He not only disputes MQA’s claim of improving time-domain accuracy, he says it makes it worse, as he wrote in “Is MQA DOA?”: “With MQA’s adaptive filtering, transients are subjected to a distortion mechanism known as aliasing. This aliasing causes tonal shifts of high-frequency content and it causes small errors in the timing of fast transients.”

This "lengthening of the system's impulse response" is disturbing to me.

awsmone · Oct 7, 2017

Al M. said:
Regarding MQA hyperbole: The question if MQA solves a real problem, or does not do so, is not unequivocally answered in Robert Harley's writings themselves. In his article "Master Quality Authenticated (MQA): The View From 30,000 Feet" he states:

"The other problem with “high-resolution” digital audio is that it didn’t really solve the fundamental problem of why digital sounds the way it does—flat, congested, hard, and glassy. Digital audio requires low-pass “brickwall” filters to prevent a type of distortion called “aliasing.” But these filters introduce ringing, or a smearing of musical signals over time. Despite attempts to minimize this distortion through faster and faster sample rates (the filters for which are less sonically detrimental), digital audio was constrained by the very fundamentals of sampling theorem codified more than fifty years ago."

Common mistake made here

Nyquist says band limited

This in the past was often done by brick wall linear phase filters

As the bandwidth 44.1khz was limited

However with high resolution recoding then naturally bandwidth limited ie less than half the sampling frequency ie for 192khz that’s 96khz

There is no content musically, even with the best microphones which go to near 100khz

Therefore there is no preringing

A Dirac impulse is infinite in bandwidth and filters are often shown in relation to this

Real world content is naturally bandwidth limited, and provided there is no content there will be no ringing

The amount of Sinc reconstruction of after ringing is debatable

The MQA filter specificies up to 32db of aliasing leakage through the 0-7khz region

When listened to tracks of MQA compared to hires, I had the distinct feeling of some psychoacoustic manipulation to blur what was otherwise not very hires tracks
The 16/44.1 was worse, and the higher res 24/96 better without the strange manipulation sonics......

Ken Newton · Oct 7, 2017

ack said:
Very well written article Al - thanks. Of particular note is this:

This "lengthening of the system's impulse response" is disturbing to me.

The length of a filter's impulse response (it's ringing) is a function of how sharp is the filtering. The sharper the filtering in the frequency-domain, the longer is the impulse response in the time-domain. The two are directly related mathematically via the Fourier transform. Therefore, a brickwall filter necessarily rings longer than some less sharp filter would.

ack · Oct 7, 2017

Ken Newton said:
The length of a filter's impulse response (it's ringing) is a function of how sharp is the filtering. The sharper the filtering in the frequency-domain, the longer is the impulse response in the time-domain. The two are directly related mathematically via the Fourier transform. Therefore, a brickwall filter necessarily rings longer than some less sharp filter would.

Thanks, that makes sense; I now recall seeing the graphs

Al M. · Oct 9, 2017

ack said:
Very well written article Al - thanks.

You're welcome, Ack, I also thought this article by Doug Schneider was a good find.

As for the ringing, I just don't hear it. I just don't understand the claims of 'temporal blur' and transient smearing as being a clearly audible issue. Piano music seems particularly critical regarding transients, and on very high-end systems I have heard enormously realistic, dynamic and clean rendition of piano from RBCD -- including hard attacks in the very high register (Stockhausen).

This 'temporal blur' thing seems just clever marketing to make audiophiles panic into thinking they must have MQA. A solution to a non-existing problem. Very clever.

Also, the pre-ringing is only on an impulse signal; as engineer Dan Lavry, a siginificant player in the field (Lavry DACs), states, this is not relevant:

"Your signal (data) is NEVER a single unity impulse preceded and followed by zeros. That is not a real world signal, such signal represents infinite bandwidth."

Link:
https://www.head-fi.org/threads/digital-filters-and-pre-post-ringing.447372/

As one guy (poster KeithEmo) puts it well here:

https://www.head-fi.org/threads/tho...-get-you-to-think-about-stuff.693798/page-379

"while all digital filters ring at least a little - so do microphones and speakers. In other words, many of the signals that DACs can't seem to reproduce perfectly are also the sort of signals that don't exist at all in real life and, if they did, microphones wouldn't be able to record them accurately, speakers couldn't play them, and vinyl and analog tape couldn't reproduce them accurately either."

adyc · Oct 9, 2017

Al M. said:
You're welcome, Ack, I also thought this article by Doug Schneider was a good find.

As for the ringing, I just don't hear it. I just don't understand the claims of 'temporal blur' and transient smearing as being a clearly audible issue. Piano music seems particularly critical regarding transients, and on very high-end systems I have heard enormously realistic, dynamic and clean rendition of piano from RBCD -- including hard attacks in the very high register (Stockhausen).

This 'temporal blur' thing seems just clever marketing to make audiophiles panic into thinking they must have MQA. A solution to a non-existing problem. Very clever.

Also, the pre-ringing is only on an impulse signal; as engineer Dan Lavry, a siginificant player in the field (Lavry DACs), states, this is not relevant:

"Your signal (data) is NEVER a single unity impulse preceded and followed by zeros. That is not a real world signal, such signal represents infinite bandwidth."

Link:
https://www.head-fi.org/threads/digital-filters-and-pre-post-ringing.447372/

As one guy (poster KeithEmo) puts it well here:

https://www.head-fi.org/threads/tho...-get-you-to-think-about-stuff.693798/page-379

"while all digital filters ring at least a little - so do microphones and speakers. In other words, many of the signals that DACs can't seem to reproduce perfectly are also the sort of signals that don't exist at all in real life and, if they did, microphones wouldn't be able to record them accurately, speakers couldn't play them, and vinyl and analog tape couldn't reproduce them accurately either."

+1. Couldn’t agree more.

MQA discussion

Well-Known Member

VIP/Donor

Well-Known Member

Well-Known Member

Well-Known Member

VIP/Donor

Well-Known Member

VIP/Donor

Well-Known Member

VIP/Donor

Well-Known Member

Well-Known Member

Well-Known Member

VIP/Donor

VIP/Donor & WBF Founding Member

Well-Known Member

Well-Known Member

VIP/Donor & WBF Founding Member

VIP/Donor

VIP/Donor