Warner Music Group Goes with MQA

MQA's game-changing breakthrough has arrived. On 9am UK time on May 6, 2016, Warner Music Group (WMG), whose vast catalog includes everything from the Beatles to Maria Callas, announced a long-term licensing deal with MQA (Master Quality Authenticated). The agreement makes it possible to digitize the entire WMG catalogue in the superior MQA-encoded format of various resolutions, and disseminate the files via download and streaming services in a far more efficient and user-friendly manner.

Up until now, MQA-encoded music files have only been available from a handful of small labels. These include Acoustic Music Records, Bauer Studios/Neuklang, Eudora Records, Jazz Arts, Personality Records, Mons Records, Ozella Music and Triplet Records, all available exclusively through HighRes Audio. In addition, 2L, Onkyo Music, e-Onkyo, 7digital, and Technics Tracks now make MQA-encoded music files available through their services. With the addition of Warner, one of the "Big Three" record labels in the world, music lovers can at last expect to see a huge amount of quality music in all genres available in MQA.

Bob Stuart, creator of MQA (above), received the news of the WMG signing just minutes before appearing at a press conference at Munich High End to introduce Brinkmann Audio's new Nyquist DAC (approx. $12,000–$13,000 US). Brinkmann's is one of the first DACs optimized for MQA streaming and playback. Standing before a roomful of representatives from publications around the world, Stuart barely raised an eyelash as he announced, in his characteristically soft-spoken manner, that the deal had just been sealed.

Stuart noted that as much as MQA is a technology, "it is also a mission." To paraphrase what he said rather quickly, our capacity to hear minute timing differences is central to our survival, because the sounds of nature give us cues as to what is going on in the environment around us. When those sounds are smeared, vital information is lost.

Stuart asserted that MP3s and even CDs push people into a corner where they are no longer listening to natural sounds. MQA changes all that. With MQA, engineers go back to the original recordings and ensure the timing information on them arrives at the ears in a natural manner. MQA removes the pollution of time smearing, giving us sounds as artists intended them to be heard.

In a press release issued by MQA, Craig Kallman, Chairman & CEO of WMG's Atlantic Records division, cited the other important aspect of MQA: the ability to convey PCM music files of all resolutions over the internet in a far more efficient and rapid manner. "The digital music era has been all about convenience," he said. "It is fantastic that we can listen to virtually any song, anywhere, any time. In that process, however, convenience has trumped sound quality, and we have gotten further away from the sound that artists work so hard to create. MQA makes hi-resolution music easy to stream or download to any device. Music fans will love it when they hear it, and WMG is thrilled to be partnering with MQA to take the next step in bringing hi-resolution music to consumers across the globe."

MQA's Bob Stuart (center) flanked by Helmut Brinkmann (left) and Matthias Lück (right) of Brinkmann Audio, shows off the new Nyquist DAC.

As of Munich High End 2016, we now have MQA-decoding available on Pioneer XDP-100R and Onkyo DP-X1 portable players, as well as DACs from Mytek, Brinkmann, Meridian, and all Bluesound BluOS-enabled wireless music systems. With Warner on board, it is far more likely that MQA-encoded streaming via Tidal, and MQA downloads from HDTracks, will come sooner rather than later.

COMMENTS
GuillaumeLN's picture

This is a sad day for music lovers worldwide. This is the epitome of boomers not looking further than their nose. MQA is for desensitized old dudes with aseptic tastes. This deal marks a sad day for music.

John Atkinson's picture
GuillaumeLN wrote:
This is a sad day for music lovers worldwide.

Would you mind explaining why you feel that way, please.

John Atkinson
Editor, Stereophile

GuillaumeLN's picture

The influence of MQA grew another step further with WMG, they seem to enlist so many organization it's a disaster. MQA makes music flatter and lifeless. It doesn't even sound good in their own gear. As long as MQA is offered for music solely appreciated by boomers, I will remain hopeful.

eriks's picture

I'm all for undermining the dominant paradigm, and at least one reviewer at SoundStage in Munich has said he was disappointed by an MQA demo, but I have to ask you if you can reply in a purely technical description. What demo's have you heard? Have you been able to hear FLAC or ALAC in direct comparison, or just MQA by itself?

As far as I knew, MQA was a lossless, but very tricky, compression algorithm. If it's causing audible effects it would be more likely to be due to processing and buffering issues than the compression itself. If you have an experience which undermines my current thinking, I'd love to learn about it. Maybe it's not lossless, I've not been following it that closely.

Of course, it should also be possible to test whether MQA is lossless and bit-perfect or not, which FLAC and ALAC can be proven to be, much like Zip compression. If it wasn't bit-perfect all Java code would cease to run. :)

Best,

Erik

GuillaumeLN's picture

I heard Meridian's demo in Montreal in a full Meridian system. They played various file formats, comparing them with MQA. I have to say I don't like the sound of Meridian products, so there wasn't much to be won over, but from the formats they played, MQA was the flatter. I don't care about technology, I rely on my ears to judge potential.

eriks's picture

That's fine. I'll have to try to listen sometime.

John Atkinson's picture
eriks wrote:
What demo's have you heard? Have you been able to hear FLAC or ALAC in direct comparison, or just MQA by itself?

I have heard multiple demos, both of MQA files on their own, and comparisons between MQA files and the original hi-rez versions. This includes recordings that I had made and had been remastered in MQA.

eriks wrote:
Of course, it should also be possible to test whether MQA is lossless and bit-perfect or not...

Read my discussion at www.stereophile.com/content/ive-heard-future-streaming-meridians-mqa and Jim Austin's and my evaluations in the June issue of Stereophile.

John Atkinson
Editor, Stereophile

divasson's picture

I am an e-subscriber of Stereophile. Very interested in MQA, after having heard an Explorer2 with MQA files run circles around a Benchmark DAC1 Pre with the same titles' original files, in my own system. Amazing what it does with soundstage, and layering and separation of instruments. Also strings much sweeter and relaxing to hear (while still full of detail) Look forward to hear it with more varied music.

(I had always thought that the space rendering of a replay chain, while not linked to musical enjoyment per se, is a very good indicator of the intrinsic quality of the chain. The deeper, wider, higher, the better.)

Also was quite opposed to the concept of streaming, but latest version of Audirvana offered a three month trial period of Tidal and now I am hooked. Look forward to Tidal's switching on MQA!

Audiophools's picture

You sir are talking nonsense and have clearly never heard it!

crenca's picture

What the "Audiophile press" does not understand about the digital world is the role format/encoding and software patents (aka "intellectual property") plays in all things digital (across the board, from the internet/computer networks to digital music players/DAC's and music encoded with digital formats such as PCM or MQA). These things are sort of the ground you walk in in the digital world while you look at other things. Right now, I am typing in a browser, using open protocols (such as HTTP and TCP/IP) as well as closed/proprietary/IP protected software (such as my OS, which is Microsoft Windows) to communicate with you. We use these things without thinking about them and their importance, and they make up the "digital ecosystem" in which we operate (in this case, purchase, store, and listen to music).

MQA could be (if it actually impacts the market and music digital ecosystem in any significant degree) a fundamental change to this world. Right now, we operate in a mostly "free" and "open" digital musical world, where our music is either IP/software patent free or (such as PCM or DSD) or is closed in ways that are mostly benign to the music lover (mp3 and other popular IP protected lossy encodings). In a world where MQA is the primary (or only) encoded format, everything changes. Instead of ownership, you "lease" your music with a software license that explicitly allows the real owner of the file to do all sorts of things at his leisure, such as "authenticate" (DRM is such a dirty word ;) ) your music and hardware preventing your to play "unauthenticated" files, requiring updates to your hardware that prevents "unauthenticated" play back, or copy limits, etc. While MQA in it's current form (v 1.0) does not do this, you explicitly sign over the rights such that in v 1.1 such things could be implemented. Even in it's current form (v 1.0) MQA is a "rights management" implementation because it "manages" what the end user hears based on whether he has licensed MQA hardware or not.

Compare the current state of digital music with digital video for example. Have you tried to *legally* rip, backup, or otherwise store your favorite BlueRay disc today? Have you tried to play it in a different "region"? Have you tried to purchase Blueray playback hardware that helps you get around these DRM limitations? If so, was what you did *legal* and within the terms of your license?

MQA is a *legal entity* in a way that all our digital music (from the very beginning of CD's up until today) has never been, including even SACD. It fundamentally changes everything (if it takes hold), because it is the very ground up which the entire digital music chain walks on. I am still a bit perplexed as to why this rather obvious set of facts is so thoroughly ignored (or just as often, outright denied) by the "audiophile press".

Also, take the first paragraph Jason writes:

"MQA's game-changing breakthrough has arrived. On 9am UK time on May 6, 2016, Warner Music Group (WMG), whose vast catalog includes everything from the Beatles to Maria Callas, announced a long-term licensing deal with MQA (Master Quality Authenticated). The agreement makes it possible to digitize the entire WMG catalogue in the superior MQA-encoded format of various resolutions, and disseminate the files via download and streaming services in a far more efficient and user-friendly manner."

This is not reporting. It's not even promotion, it is over the top propaganda that would make Joseph Goebbels smile. I have to ask; do you guys have a golden calf with an "MQA" brand on the rump that you worship when you walk in the office every morning? Truly, make some small effort to step back and try to understand the bigger picture...

John Atkinson's picture
crenca wrote:
What the "Audiophile press" does not understand about the digital world . . . do you guys have a golden calf with an "MQA" brand on the rump that you worship when you walk in the office every morning? Truly, make some small effort to step back and try to understand the bigger picture...

As you seem more interested in beating your chest than having a civilized dialog, I see no point in responding.

John Atkinson
Editor, Stereophile

Jason Victor Serinus's picture

And with anger. The day before Munich High End began, my husband and I spent three hours at Dachau. I saw the boxcars where my relatives died, and the gas chambers where my homosexual brothers died.
The bigger picture is that you are offensive beyond belief.

crenca's picture

I studied Joseph Goebbels in my marketing 101 class for goodness sake, and understanding his place in the history of propaganda should be required reading for any educated person. Perhaps you should take the time and look him up...

ChrisS's picture

I'm reading this!

eriks's picture

I'm pretty sympathetic to the general idea of brands and licenses taking value out of the music / entertainment industry or making products with cheaper parts because they have to have so many brand labels on them. I would rather see going to writers, musicians, and those who actually build physical things but still I would wish for an anxious posting like that to be more specific and on point. Kind of with the Editor here. Can't really find a way to reply to that.

On the bright side, if MQA actually has consumer facing benefit (unlike say HDCP) they have described it in such clear and open fashion that open source versions will soon follow IF it's actually something we need. That value will have less to do with audibility than cost of data over a network, and storage thereof.

MQA may in the end of the day die not because it's not useful, or brilliant bit-juggling but because it's decades too late in solving a problem. The speed of home networks, and cost of storage has dropped orders of magnitude over the last 20 years. Being able to compress 24/96kHz audio 8 times may no longer have a value worth paying for. I already stream 24/96kHz Jazz from listener supported FM 91 in Toronto just fine thank you.

I will now go read the Stereophile articles on MQA and see if I can learn more. I'm rather ignorant of the final claims and Meridian's licensing direction. I hope Stereophile will keep informing us of them.

Best,

Erik

crenca's picture

What Jason does, marking the date and time of a contract as if it is comparable to some great event, perhaps even approaching a religious one. Sure, it rises above a base level of discourse in that it is a "strong" response but it is subdued compared to the literal propaganda on display in Jason's post. Sleep on it, and think about MQA in terms other than the "second coming", and perhaps you will see what I mean...

RichT's picture

Crenca,

what you say only makes sense if you believe that MQA is a trojan horse for DRM. It can't do DRM now, and there's no evidence that anyone intends to add that in the future, indeed there are statements from Bob Stuart that say the opposite. If there's a balance to be struck between getting access to higher quality audio and an extremely remote chance of someone adding DRM to MQA then I know which side I'm on.

crenca's picture

To paraphrase Forest Gump. DRM stands for "Digital Rights Management", and thus does not point to any particular technical implementation of the "management" of the end user "rights". What you have done is bought into the self serving and narrow definition of DRM that Bob and others serve up to then (correctly) claim that MQA does not implement these specific features. All true, but quite irrelevant because DRM is in fact something much wider and MQA in its current form is already DRM, in that it "manages" what the end user hears based upon what hardware and or software (assuming there will be a software decoding for MQA) the end user has licensed from Meridian.

What this means is that the vast majority of current DACs, what Bob calls "legacy" DACs (very revealing choice of words on his part) are not MQA licenced, and thus they are "managed" by MQA such that they can only decode (and the end user can only hear) the 16/44 part of an MQA encoded file. If you as an end user try to decode the high res content of an MQA file you will in fact be breaking the law...yep, MQA is DRM all day every day.

Besides, even using a definition of DRM that includes encryption, copy management, MQA-phoning-home sort of stuff which it does not currently have, by purchasing MQA hardware and software (aka music encoded in MQA format) you are agreeing to such things if the real owner of the file decides in the future that is what they want you to have (perhaps in v 2.0, a mandatory update).

You should read Robert Harley's stuff over at the other major audiophile publication - he writes about the real reasons for MQA (i.e. the "broken business model" of digital music, etc.) and while he incoherently then regurgitates the Meridian/Bob line about MQA currently not being a DRM entity, he reveals the real reasons why MQA (or something like it) is coveted by the music industry (hint: they want DRM in the strongest form possible, perhaps it even being an non-negotiable necessity for their future survival).

No reason to "choose sides", however one should not be naive about the real motivations behind the industry and MQA. As Robert understands, it is not about sound quality/fidelity as that is just the "hook" to get us to buy into the paradigm...

RichT's picture

Crenca, I understand where you're coming from but we'll have to agree to disagree, I think. I can't see any elements of DRM in only allowing you access to the MQA stream with a decoder. You can still copy the file, send it to someone else etc and treat it as a standard LPCM file without a decoder.

As for the future, we will have to wait and see. I don't think it's likely that MQA or the content owners will retrospectively add DRM to the format as it would result in a massive consumer backlash and loss of trust.

Best wishes

Rich

Audiophools's picture

Crenca, I'm sure you mean well but you are talking complete boll*cks!

As long as MQA stand by their assertion that there is no DRM at play then lets go with that and be done with it.

The important bit, does it sound any good is all anyone is really concerned with and luckily the answer is yes.

I pay for a streaming service (Tidal) so don't own that portion of music. If I buy CD's encoded using MQA then I own a physical copy and im happy with that also as I will be able to listen to it on my DAC whenever I see fit and no one will be turning up at my door to take it away.

Just concern yourself with the facts and the technical practicalities about what you can do with your media and enjoy the music.

crenca's picture

"As long as MQA stand by their assertion that there is no DRM at play then lets go with that and be done with it."

As long as you go in with me on this beautiful beachfront property I have my eye on in Arizona...in fact, why don't you just front me the money and I will pay you back latter...really, I mean that - just "go with it"... ;)

Audiophools's picture

Crenca, You really are just spouting nonsense for the sake of it. Either contribute in a meaningful way or take your tin foil hat elsewhere.

LS35A's picture

When I first read about MQA it appeared to be merely a way to store hi-rez files with less space.

So you could have a 24/96 file, for example, and not have it take up any more hard drive space than a 16/44 file.

Since I have been frankly underwhelmed with the difference higher resolution files have on SQ (when using the IDENTICAL master) this did not interest me much, if at all.

But now I'm reading about some kind of 'filtering' or other audio processing in the MQA chain that flat out improves the sound quality, not just makes a higher res file take up less space.

Is that right? If so what's the skinny on this 'audio filtering' that MQA brings to the table?

dce22's picture

If someone knows more please chime in,
so far the speculations are that MQA is 24/44.1 file that has the last 8 bits replaced with the upper part from 88.2 sample rate (from 22khz-44khz freq response) that the processor glues back together as 16/88.2 stream so you get
16bit from 2-22khz and
8 bit from 22-44khz
if you listen the file in non MQA DAC you get 16bit resolution with last 8 bit chattering in the noisefloor

It was common knowledge that more bits (dynamics) is better then samples (frequency range) i guess someone believes that ultrasonics are more important than the thing that we can hear.

Real 24bit/44.1khz will always be better than MQA 24bit/44.1khz that is 16bit+8bit/88.2khz.

If im wrong please elaborate and expand my mind.

eriks's picture

I'm not quite sure that's the argument. Rather, that with a little nuance you can put your bytes where they matter the most. That I can support in theory. Another argument being made, that I can support less is that it's the timing rather than the bandwidth and that is why we need 384kHz music. Consider a low frequency signal at 20 Hz. At 44.1kHz it gets 2,205 samples per cycle, and needs a lot of bits for it's dynamic range. A 10kHz triangle strike however only gets 4.4 samples and needs far fewer bits just because it's not as loud. MQA is trying to take advantage of this discrepancy. It is in a way right-sizing the use of the bits to the need of the music notes. Now, have they done so inaudibly, and will they cause a new set of licensing/DRM headaches for the listeners? I don't know.

dce22's picture

"Consider a low frequency signal at 20 Hz. At 44.1kHz it gets 2,205 samples per cycle, and needs a lot of bits for it's dynamic range. A 10kHz triangle strike however only gets 4.4 samples and needs far fewer bits just because it's not as loud."

You are completly wrong in you deduction, you are looking the encoded information (Pulse code modulation) when you demodulate the PCM you dont have 4.4 samples you have infinte samples at 10khz the way it works is in 4.4 + 5 infront and 5 behind the 4.4 samples you stated the original signal is encoded in time and in level and the perfect curvature that is recorded between every single sample in perfect timing and level can be decoded (run PCM thru decoding mechanics sinc function) and you get exact detail at 10khz as the original analog signal.

In the decoding algorithm the more FIR taps you have the more correct detail between samples you will get so you dont get 10khz from 4.4 samples but from 1024 samples also you get 20hz from 1024 samples in high end dsp oversamplers in regular dacs you get 64 taps so it use 64 samples for 20khz and 64 for 20hz.

More taps more precision of the curve between samples less distortion on the waveform
64 taps is around -124db Noise and Distotion.

You can use alot memory and store alot samples and demodulate PCM at insane level of presicion that is imposible to even do in analog.

If you study Sampling Theorem you will see that there is zero timing error and no phase shift at all in 44.1khz sampling below 20khz and 96khz sampling below 45 khz and so forth.

Your 10khz 4.4 samples is completely off the mark for more easy to understand info read this.

http://lavryengineering.com/pdfs/lavry-sampling-theory.pdf

You are missinformed read the paper.

eriks's picture

It's disappointing to me that you should have spend to much effort replying to statements that were never made. Did you mean to reply to someone else?

First, I do not attempt to claim any theorem's truth or falsehood. I was just restating what I understand to be MQA's raison-de-etre for the original commenter. Second, I make no deduction. Third, I never actually talk about samples in replaying.

I think if it's important to you, please apply your understanding of Nyquist's theorem to Stereophile's article on MQA and critique MQA directly.

Now, I certainly could have misread that article, and if I did please let me know.

Thanks,

Erik

dce22's picture

"Consider a low frequency signal at 20 Hz. At 44.1kHz it gets 2,205 samples per cycle, and needs a lot of bits for it's dynamic range. A 10kHz triangle strike however only gets 4.4 samples and needs far fewer bits just because it's not as loud"

That statement is susceptible to incorrect interpretation and is misleading triangle sound does not need less bits, bits in PCM only represent where the noise floor is gonna be so MQA uses 24bit lossless format and trim the lowest 8 bits to put 8bit info to create ultrasonics to "improve temporal mumbo jumbo" is wrong direction it is better to have 140db audio dynamics storage capability in the audio band 2-20000hz then to slash it to 94db dynamic range just so you can have 20000-40000 khz range with 47db dynamic range (8bit) into frequency that people cant hear.

MQA can go one step beyond slashing 24bit/96khz that is superb into 16bit/192khz turning high res into cd quality

"MQA is trying to take advantage of this discrepancy. It is in a way right-sizing the use of the bits to the need of the music notes"
It is not right-sizing it is very wrong to sacrifice dynamic range for freqency range that is beyond audible.

Problem with 44.1khz sampling is only poor anti image filters that can be fixed, and even with this poor filters it is better to have 24bits at 44.1khz than to have 16bit at 88.2khz.

At 96khz sampling rate all the filter problems are in inaudible ultrasonics, and faster rates like 192khz and above only mangle and distort the music more but people are missinformed that more samples equal more dots better representation that is incorect assumption that looking at the pulse code data (PCM) might look that way but that is just the code when you demodulate you get infinite dots at all freqency until half of the nyquist rate

"A 10kHz triangle strike however only gets 4.4 samples"

is incorect perpetuation that more dots equal more resolution
No it dont only more bits equal more resolution.

MQA is willing to monetise the fact that almost every audiophile thinks that more dots equal more resolution and the informed guys like John Atkinson knows but perpetuates this lie so he can collect advert funds from these false audio prophets like Bob Stuard that is willing to reduce High Resolution audio to CD Quality to extend into ultrasonics that are 8 bits less the FM quality and promote it as a better sound quality.

Even analog guys like Michael Fremer who does not give a crap about digital commented a few years back that to him 24bit 44.1khz sounded better than 16bit 96khz i did test it with Weiss Saracon converter took the original 24/96 dithered to 16/96 against original converted to 24/44.1,

Guess what 24/44.1 is better than 16/96.

Edit:
I need to correct myself apprently i was misinformed about how MQA works the principle is the same using lowest bits to store ultrasonics it still reduces the music to around 16 bits but it has lower noise into ultrasonics more than 8 bits but it is lossy not lossless like i clamed above, it is a Meridian version of Spectral band replication you mix data compressed ultrasonics into the audioband and decode them into ultrasonc data when playing back while the audio band stay distorted the more the original recording has ultrasonics the more distortion gonna be.

My opinion still stands that you should not sacrifice audio band for ultrasonics.

eriks's picture

You keep taking my writing in a direction I'm simply not. If you wish to argue against the original theorem, which you seem to be a bit haphazardly, you are free to do, but as a reply I can't derive any meaning from except your dissatisfaction.

Best,

Erik

dce22's picture

if i came out strange i apologize i just noted some things that are universaly missunderstanded like high frequency loses precision and needing-using less bits at high frequency, and 44.1 is not enough for 18khz music.

I said that i dont agree with sacrificing bit depth for incorrect theory that its better to spend it on high sample rate that you have claimed and offered technical documents why it is so.

"If you wish to argue against the original theorem" no im not there is infinite samples at 10khz when decoded plug in analog osciloscope into a dac and see for your self.

John Atkinson's picture
dce22 wrote:
MQA is willing to monetise the fact that almost every audiophile thinks that more dots equal more resolution and the informed guys like John Atkinson knows but perpetuates this lie so he can collect advert funds from these false audio prophets like Bob Stuart...

As MQA doesn't advertise in Stereophile, your statement is BS. Actually it is BS regardless. If you wish to keep posting to this site, please stick to what you actually know.

John Atkinson
Editor, Stereophile

dce22's picture

I never said MQA has adverts in Stereophile i said Bob Stuard aka Meridian Audio and my statement is not BS, [flame deleted]

John Atkinson's picture
dce22 wrote:
I never said MQA has adverts in Stereophile i said Bob Stuart aka Meridian Audio and my statement is not BS...

Meridian doesn't advertise in Stereophile either, so you are doubling down on your BS. Please note that advertising has no influence on what Stereophile's writers say or what Stereophile chooses to publish. If you continue with these baseless accusations and flames, I will have no choice but to block you from posting. (Note that I deleted more inflammatory text from the posting to which I am responding.)

You are a guest on this site. Please behave like one.

John Atkinson
Editor, Stereophile

dce22's picture

I apologize i was wrong about Meridian adverts i must have seen speaker ads in some other publication.

May i say it as a suggestion to write article about how digital signals are recorded and reproduced how oversampling works and why its needed and so forth you can interview people like Dan Lavry, Daniel Weiss who are on the front line in recording and mastering gear, just to trim the fat a little bit.

John Atkinson's picture
dce22 wrote:
I apologize i was wrong about Meridian adverts i must have seen speaker ads in some other publication.

Thank you. Apology accepted

dce22 wrote:
May i say it as a suggestion to write article about how digital signals are recorded and reproduced how oversampling works and why its needed and so forth you can interview people like Dan Lavry, Daniel Weiss...

A good suggestion.

John Atkinson Editor, Stereophile

JimAustin's picture

Read the June Stereophile, which will be out in a few days. I give my (entirely subjective) take--but more important, John Atkinson describes the technology in a way that addresses some of its oft-misunderstood aspects.

John Atkinson's picture
dce22 wrote:
If someone knows more please chime in, so far the speculations are that MQA is 24/44.1 file that has the last 8 bits replaced with the upper part from 88.2 sample rate (from 22khz-44khz freq response) that the processor glues back together as 16/88.2 stream so you get 16bit from 2-22khz, 8 bit from 22-44khz...

The situation is much more complicated than this. See the explanatory video at https://www.youtube.com/watch?v=T5o6XHVK2HA. HT to Ayre's Charley Hansen for the link.

John Atkinson
Editor, Stereophile

Charles Hansen's picture

The link is to the YouTube channel of a Dutch reviewer, Hans Beekhuyzen. It is Part 2 of a two-part series. The first part explains why there is a need for high-resolution files at all - improved timing precision to which the human ear/brain is exquisitely sensitive.

Unfortunately Part 2 (linked above) repeats much of the misinformation that MQA has used when describing their technology. If one starts with an original high-res 192/24 digital audio file and runs it through the MQA encode/decode process, many things happen. Let us look at the case where the listener has an MQA decoder and the "audio origami" is fully unfolded/decoded.

After "folding", the original file is stored in a 48/24 FLAC container for further lossless compression and reduction is bandwidth required for streaming and/or downloading.

1) The baseband information (as sampled with a 48 kHz sample rate is reduced in resolution, apparently from a full 24 bits to somewhere between 16 and 18 bits of resolution. All of the graphs created by MQA on this topic are somewhat misleading, as they show a 16-bit file as having a noise floor of either -144 dBFS (as seen in Figure 8 of the original JAES paper published in October 2014) or -120 dB as seen in the current literature MQA is releasing and which can be seen at 17:00 in the linked YouTube video.

The fine print that allows for this increase of either 24 or 48 dB in S/N ratio is that the noise is given in "noise per-root Hz", something I've never seen used before in any other presentation of digital audio, and not explained nor justified anywhere in any of the MQA literature I've read.

2) The audio data from the quad-rate information is compressed using lossy compression techniques, which MQA refers to as "encapsulation". This is confirmed in the linked YouTube video.

3) The audio data from the double-rate information is compressed using lossless compression techniques. This is confirmed in both the linked YouTube video and in other information released by MQA.

4) The combination of the double- and quad-rate information is buried in the bottom bits of the 48/24 FLAC container. Apparently the amount required by the double- and quad-rate audio data varies according by the program material. Specifically, musical content with low levels of energy above 20 kHz can presumably be compressed further than can musical content with high levels of energy above 20 kHz.

The current MQA graphs (as seen at time 17:00 in the linked video) show the double- and quad-rate information buried in the "unsused" bits below the baseband audio data. However this graph shows the lowest levels starting at -168 dB.

Normally a noise level of -168 dB would require 28 bits of resolution to capture (the generally accepted approximation is that one achieves 6 dB of S/N ratio per bit, so a 24-bit file would have 144 dB of S/N ratio if the analog circuitry were completely noiseless). Remember that the FLAC container has only 24 bits of depth. Again "noise per-root Hz" is used to show that 24 bits of audio can achieve a noise floor of -168 dB, and again with no explanation or justification by MQA. In the graph shown at 17:00 in the linked video, the double- and quad-rate "folded" audio data fills up the room from -168 dBFS to -120 dBFS.

The example used by MQA is of a Ravel string quartet. It is unknown which one, but it is known to have sections of pizzicato playing, as this is stated elsewhere in MQA's promotional literature. The reason this is important is that strings have the lowest levels of high-frequency content of any common musical instrument, and pizzicato strings (plucked rather than bowed) have even lower levels of high frequency content than do bowed strings.

Please refer to Caltech professor James Boyk's paper which specifically details the amount of musical energy above 20 kHz with a variety of common musical instruments, available online at:

http://www.cco.caltech.edu/~boyk/spectra/spectra.htm

In it we can learn that a bowed violin only has 0.04% of its musical energy above 20 kHz and a pizzicato violin has even less, with only 0.02% of its musical energy above 20 kHz. In contrast much music (symphonic, popular, jazz, et cetera) has large amounts of cymbals as a basic element of the rhythm section. Prof. Boy measured cymbals to have a whopping 40% of their musical energy above 20 kHz, and their energy extends past 102 kHz (the measurement limit of Prof. Boyk's equipment).

Presumably when music other than a string quartet is played, the audio information in the double- and quad-rate sections will require more than the 8 bits shown on the MQA graph at time 17:00 in the linked video. It is unknown at this time how low the resolution of the baseband audio will be reduced in the case of high levels of musical energy above 20 kHz.

What is known is that the decoded MQA file does not give a lossless reconstruction of the original 192/24 file. While the information extends out to the frequency limit of the 192 kHz master, information in the top octave is stored with lossy compression techniques ("encapsulation"). The information in the next octave is stored losslessly, but all of this is done at the expense of reducing the resolution in the baseband audio (0 to 24 kHz, the Nyquist limit of the 48 kHz single-rate sampled material).

Again it is unknown exactly how much resolution is lost, but based on the MQA supplied graphs shown at 17:00 on the linked YouTube video, it appears that the baseband resolution is no more than 16 bits when music with very low levels of high frequency content is processed by MQA. It is possible that when music with greater amounts of high frequency content (particularly cymbals), that the resolution of the baseband audio data is reduced below 16 bits. Again, this is all masked by MQA's unconventional use of showing the noise floor in terms of "noise per-root Hz" and lack of any other musical data other than a string quartet, shown by Prof. Boyk as having the lowest amounts of energy above 20 kHz of any common musical instrments.

I hope this helps to clarify some of the issues surrounding MQA and leads to further discourse, as it seems that currently there are more questions that should be answered before adoption of this technology, in my opinion.

crenca's picture

It is heartening to see an important "industry insider" openly questioning what has so far been a one sided discussion (that is understated of course - it has not been a discussion but a full frontal promotional attack) around MQA. It is good to see, because it is becoming apparent to me that music lovers (i.e the "consumer") have more to lose with MQA than gain. This can not be good for the livelihood of many in the industry as well...

RichT's picture

Thanks for this Charles.

I saw that in their AES paper, Bob Stuart and Peter Craven report an analysis of high frequency content across a sample of material (sample size not given). They say:

'This picture of the content occupying a ‘triangular’ space is common in all recordings we have analyzed and the converging point is usually below 48 kHz, with the highest so far being at 60 kHz.'

So the Ravel quartet does not seem to be too atypical.

Best wishes,

Rich

Charles Hansen's picture

Did you get a chance to check out the linked paper by Prof. James Boyk? I thought it odd that MQA chose the program material with the least amount of ultrasonic content possible. As you note, MQA don't give details of the sample size used for their spectral analysis.

You note that the "convergence point" can be as high as 60 kHz, but MQA does not state how that affects the bit allocation. From what we know about the MQA system, encoding the dual- and quad-rate information in the "lower bits" through the "audio origami" can only come at the expense of resolution of material in the baseband audio data. If a string quartet only yields 16 to 18 bits of resolution in the baseband, how much resolution is available in he baseband of music with much higher levels of energy beyond 20 kHz?

One thing that is obvious from looking at the data presented by MQA in the JAES paper is that they used a mix of 192/24 and 96/24 files for their spectral analysis. Look at Figure 7 and you will note a -24 dB drop-off in high frequency signal levels at 48 kHz. (This is due to the brickwall filters used by standard 96 kHz A/D converters.)

This clearly indicates that a significant portion of the material used for their spectral analysis was sourced from 96/24 files. Why focus on anything less than the highest quality reproduction available? If I were making a new format that was supposed to redefine what is possible, I would not be designing it around double-rate source files. That's just my approach, YMMV.

dce22's picture

Thanks to John for correcting me that is not lossless doubling 16+8bits and thanks to Hansen for shining some light.

Its looks like somekind of dynamic storage into the last bits that are tagged by a header that is read as data streams from the file and decoded into the upperband damaging things that we can hear for things that we cant pure BS.

The video has no info at all it parrots the press release about impulse response and temporal BS and minimum phase filters,
People need to know to create the best audio from PCM you need to decode all the info that is recorded you do that by using sinc(x)=sin(x)/x and this function looks like this

https://upload.wikimedia.org/wikipedia/commons/5/59/Si_sinc.svg

Anything that has Impulse response that does not look like it is wrong.

Minimum phase filters exist and are designed for monitoring purposes so the football interviewer that is on the field can talk to the mic thru wireless and broadcast console and wireless headphone without stuttering because the latency buildup is too high there is no pre ringing no post ringing (there is pre/post echo nothing to do with ringing)
Before you say that im wrong i will post the link i have posted again

http://lavryengineering.com/pdfs/lavry-sampling-theory.pdf

Hansen need to disable this weird filter (low latency) options in his dac's that are damaging the sound quality.
But people like it , well people are basically morons it is different so it must be better, no its not, not in mathematical sense not in performance sense not in subjective sense if you have heard the original music signal before adc-dac conversion and toggle the filter you will be convinced too.

Hbeekhuyzen's picture

Hello Charles,

You questioned the technology used for MQA or at least the way it was presented in the AES paper and as a result in my youtube videos. I write this one answer and will publish it on my Youtube channel (https://youtu.be/T5o6XHVK2HA) and the Stereophile discussion (http://tinyurl.com/jzaj2eh) mentioning links to both. Sorry for taking so long but I wanted you to give you the best answer I could. Let me begin by stating that I am a technology journalist and not a researcher.

Then I would like to make clear that MQA sounds like a very good technology, ie. my ears tell me that the MQA version of a given recording produces a more natural sound than the same recording, digitized at the same Fs and bit depth without MQA. And John Atkinson (Stereophile) and others seem to agree. The most interesting group of people that agreed were the people from the recording industry that were extremely enthusiastic about the Pono player. A number of the must have heard 192 kHz recordings before so why then so extremely enthusiastic. Well it took some digging but it appears the initial version of the Pono player contained MQA. They even liked it so much they wanted the sole right of use. After that, you were asked to design the Popo player and you did a good job but couldn’t use MQA. Nevertheless, the Pono commercials showed the reactions on the MQA versions. So MQA does sound better. The article “Inside MQA” by John Atkinson (may 2016) shows that the folding works.

Now let’s take a look at your remarks:

Re: Volt per square root hertz
I don’t know why Bob Stuart used Volt per square root hertz, but his might be an explanation:
If you want to measure a power spectrum and vary the spectrum width, you’ll have to take into account that the wider the spectrum, the more energy it contains. You therefore have to take the bandwidth into the equation. Since we sample voltages  and since P=U²/R, you’ll end up with volt per square root hertz.

As you know the dB expresses a ratio and on it’s own has no absolute value. If MQA had just used the ‘dB’ as it is an absolute value without specifying the reference, he perhaps might be accused of misleading. But here MQA specified clearly the reference. This reference might by unusual and unknown to us, that doesn’t automatically makes it suspect. To explain an innovative and radically different system might ask for different ways of measurement.

Re: Spectra.
It is clear that a string quartet is limited in bandwidth and that music containing metal or brass instruments have a far greater bandwidth. They, however adhere to the 1/f rule for energy and will loose that energy fast over distance due to air resistance. The study you mentioned (There’s life above 20 kHz) measured a trumpet with Harmon mute at 4 feet distance. And even then the content above 20 kHz is over 60 dB down in this measurement. In classical music the microphones will be at greater distance and thus will pick up even less energy above 20 kHz relative to midrange. For popular music this is not the case since that is all recorded using close miking. But here the dynamic range is very limited to extremely limited. 

As you know there are no recordings exceeding 20 bit resolution since analogue electronics doesn’t offer this dynamic range. In practice classical recordings might be in the 16 to 18 bit resolution region when done properly. For popular music that often is considerable less. MQA offers a flexible way of locating bits there where they are needed. If there is lots of energy in the twoband and relative little dynamic range in the baseband, more bits can be located to the twoband and vice versa. MQA is working hard on plug-ins to give these tools to sound engineers.

Allow me a kind remark: I love to have a lively discussion on my and other places on the web. It is a very good way of stimulating and learning from each other. But to achieve this, it doesn’t help to accuse others of misleading unless you can prove it. And even then is’t more elegant to inform after their reason for doing so. I have deep respect for the work of both you and Bob and like you two, I am a long standing AES member. Let’s think and work in the spirit of the founders of the AES.

Hans Beekhuyzen

Hbeekhuyzen's picture

removed double entry

Archimago's picture

From LS35A:
"When I first read about MQA it appeared to be merely a way to store hi-rez files with less space.

So you could have a 24/96 file, for example, and not have it take up any more hard drive space than a 16/44 file."

Actually, it's more like squeezing the 24/96 file into either 24/44 or 24/48. At least the implementations we see so far, these are the file sizes that end up produced.

"Since I have been frankly underwhelmed with the difference higher resolution files have on SQ (when using the IDENTICAL master) this did not interest me much, if at all."

Yes. Much of the current 24/96+ "hi-res" files are just the same masters with potentially the same issues like "loudness wars" dynamic compression. 24-bits don't matter in any of these releases so it would be most unusual for anyone to claim they hear much if any difference. As for >44kHz sampling rate... Well, not exactly much evidence in the research literature to suggest significant difference for perhaps all but the best trained Golden Ear :-).

"But now I'm reading about some kind of 'filtering' or other audio processing in the MQA chain that flat out improves the sound quality, not just makes a higher res file take up less space."

Yup. Supposed to "de-blur". Obviously some kind of DSP effect.

"Is that right? If so what's the skinny on this 'audio filtering' that MQA brings to the table?"

Ahhh... That's the most important and key question and hopefully the investigative journalists of the audiophile world should explore. If there is anything of benefit MQA brings to the table it is potentially this DSP process. But there are issues with this claim given that we are still working in a PCM foundation. Furthermore, there are limits to the sophistication that they can extract out of those lower 8 bits or so. Low power DACs are supposed to be able to decode so again the algorithm can't be too computationally intensive especially when it's being expanded back to 24/192 or even 24/384 for DACs that support the higher samplerate.

From my perspective, the DSP "de-blurring" can be dissociated from the data compression piece. If the de-blurring does actually do anything of value, it can be applied to the original 24/96 file and we can easily download this "remastered time-domain improved" version just as well.

In terms of streaming hi-res - what's wrong with just flat 24/48? Looking at the MQA compression scheme with the samples out there from 2L, you'd likely get better overall compression with just flat 24/48 FLAC files than an MQA 24/44 FLAC lossless.

Unless the de-blurring truly sounds *that* good, this MQA "format" IMO just doesn't make sense. It's too convoluted and inconvenient, requiring a hardware upgrade for questionable technical benefits. IMO, the MQA Q&A posted on Computer Audiophile adds nothing new of tangible, technical value.

master3d's picture

Hello World,

a rather longish, but beautifully comprehensive and honest (sounding) Q&A regarding MQA can be found at the URL below, debunking many technical and world domination (i.e. conspiracy) theories:

http://www.computeraudiophile.com/content/694-comprehensive-q-mqa-s-bob-stuart/

masteR3D

monetschemist's picture

That is a nice article / interview.

However the comments related to "walled gardens" are, in my opinion, misleading.

We can think of MQA as two formats in one. The open part of the format gives us a lower-quality CD-ish data stream that we can play back on any DAC, and a higher-quality data stream that we can only play back on a DAC that has licensed MQA. And if someone takes it in their head to develop "MQA improved", what then? Does the patent prohibit that?

Conversely, if we have a music file in FLAC format, no one can interfere with our use of that data stream as long as we don't violate copyright, which does not pertain to the means of decoding that stream. If someone wants to invent a better way to decode that information, nothing is stopping them.

To me that is an important and fundamental difference. Because the MQA includes a patented decoding mechanism, it really is digital rights management. I would rather not buy my music in a format that requires someone else's permission to be decoded. I already have a bunch of HDCDs that are like that.

RichT's picture

@monetschemist.

Bob Stuart has said that all future versions of MQA will play on version 1 decoders. So I think the chances of future versions being able to implement DRM are extremely low, because the version 1 decoders will still work and they have no DRM enforcement capability. You can choose not to believe him of course, but personally I think that's what will happen.

Licensed technology is used in many places in the music and audio industries, for example Dolby True HD as used in DVDs. This doesn't have a DRM component either, afaik.

John Atkinson's picture
I wrote back in December 2014 that while MQA doesn't include DRM, there is still a commercial benefit for the record industry with MQA that is not true about FLAC etc: The record company will no longer be selling a duplicate of their master, with all the implications for piracy that that implies, but something that in theory will sound identical to the master but doesn't allow the master to be recreated. In other words, it is like DRM but without all the hoopla that traditional DRM involves, which, as well as the single-file inventory, may well be why record industry people appear to be supportive of MQA.

Regarding the open-source vs licensing models, I think that taking a moral stance favoring the former is misplaced. People deserve to be compensated for creating intellectual property and both models are valid.

John Atkinson
Editor, Stereophile

crenca's picture

One has to grant a narrow definition of DRM to say that MQA in it's current form is "like" DRM but not "traditional" DRM. DRM is and always has been more about IP and it's enforcement in courts than any technical implementation of encryption, etc. DRM's "power" is the fact that IP makes software (in this case a digital music file) a *legal entity* and issue (in a way that does not apply to PCM, DSD, etc.) and thus MQA makes file ownership and "rights" something that puts all the real power to the licensor of the software.

One of the coups of the MQA promotion machine has been the amount of buy in by the average reader/music lover to the lines that "MQA is just PCM" and "MQA is not DRM".

If MQA in fact allows the end user to "hear" the master but can not "recreate" the master, then the reason is DRM - the end user is being "managed" by the force of law (aka "Intellectual Property", software patents, etc.).

Finally, the "moral" argument can not be boiled down to a simple dialectic. Closed proprietary formats have a significant impact on digital ecosystems and end users that extend way beyond the "moral right" (something I do not disagree with) for an artist/label to be compensated, and "free" formats such as PCM are not in-of-themselves responsible for piracy and the other problem areas of the current music industry, and thus DRM will not "fix" the industry (though this I believe is what they think is necessary step in the resolution of their current troubles)...

mcgilroy's picture

MQA is being marketed to Audiophiles as an advance in sound quality but it really is about controlling assets. Streaming audio is growing; high-resolution audio a lucrative subset of it and the music industry does not want to experience the Spotify/Amazon/Apple fiasco once again.

They desperately need a tool to control the monetarization of the high-rez versions of their catalogs. A tool which is not controlled by the big Cloud companies (or Stacks as Bruce Sterling has called them).

A little company from England offering DRM in disguise is just what is called for.

Since Audio-journalists apparently cannot look beyond a press-statement and simply read up what DRM is and does - for example on Wikipedia - let me quote the definition given there:

"DRM technologies try to control the use, modification, and distribution of copyrighted works (such as software and multimedia content), as well as systems within devices that enforce these policies"

I'd be very surprised if anybody is able to explain how MQA is not doing exactly that. MQA controls certain uses of content via a clever combination of software-techniques both on the content and on the device-side of the audio-chain.

Sure MQA allows to access the standard-resolution part on any device, but it's an access control technique for the high-rez part of the asset. Half-free is only half-free no matter how much one try's to explain the DRM-part away.

There might be another important twist: the Hifi-industry not the music-customers will bear the brunt of the DRM-scheme deployed here. Hifi companies will have to license with MQA and MQA might exert control over the capabilities of the audio-device and the data generated by it. Apple had shown that the money with digital music is made with the device - MQA aims to copy that via a it's licensing mechanism for which the industry pays.

There maybe even are some sonic-advantages of MQA. But so far no sound technical explanations have been offered. Rather the opposite and it seems like psychoacoustic tricks are employed to achieve a different sound-impression. Which would be ok with me - if just explained in all honestly by Bob-the-surival-of-the-human-race-depends-on-this-Stuart.

We do know it's proprietary, we do know it uses lossy compression for frequencies above 48khz and we do know it's a clever take on DRM.

In essence MQA is a proprietary, lossy DRM-Trojan solving primarily a problem for the music industry and (maybe) secondarily yielding a "better" sound.

RichT's picture

Hi McGilroy,

I think it could be helpful to think about two types of asset control.

- systems that attempt to control use and distribution of works on an individual basis, e.g. attempting to restrict use to the original purchaser and to prevent copying.

- systems that require a licensed decoder to operate, but don't limit the copying and distribution of individual works.

I agree that the first category counts as DRM. It sounds like you think the second is too, but if so, it's not what DRM has traditionally been about. There is no control of individual assets.

Are you aware that the MQA spec allows for a number of 'transport' rates, and all encoders and decoders have to support 1x, 2x and 4x rates? The 48kHz threshold for lossy compression only applies to the 1x rate, which is what all the discussion has been about so far. For the other transport rates, I believe the frequency thresholds will be proportionately higher.

How important do you think reproduction of signals above 48kHz is? Let me explain why I ask that. From my own experience, 192kHz sampling does sound better than 96kHz, and 96kHz better than 48kHz, but I don't believe that's down to extended frequency response, rather down to better impulse response. As I understand it, though I am not clear about how, MQA captures the impulse response differently, and at better resolution than 192kHz sampling or even 384kHz. So when we talk about lossy capture of signals above 48Khz with MQA we are not talking about limitations in impulse response but purely reproduction of frequencies above that frequency. Do you think that would be audible?

best wishes,

Rich

dce22's picture

"extended frequency response, rather down to better impulse response"

Same thing
Frequency Response=Impulse Response

monetschemist's picture

Mr. Atkinson,

If the record company is able to guard against anyone using the master-quality recording except in the case of playback through a licensed decoder, I fail to see the difference between that and DRM. I respect that they are somehow different from your perspective but to me the duck test applies - if MQA limits my playback options to equipment containing licensed proprietary software, then it is managing digital rights.

And if the whole recording industry gets on the MQA bandwagon and only releases their high-quality data in MQA format, that sounds like collusion and monopoly to me and not just Meridian being compensated for their investment in intellectual property.

I get your point about open vs. proprietary but I suggest there are other models that permit receiving value for intellectual property that could perhaps be applied in this case.

For instance, in theory at least, the MQA encoder could be proprietary and sold for-profit, and the bitstream specifications and decoder could be open. Of course then it would not be nearly as attractive to the record company since who gets to listen to what is no longer controlled by the MQA license.

John Atkinson's picture
monetschemist wrote:
If the record company is able to guard against anyone using the master-quality recording except in the case of playback through a licensed decoder, I fail to see the difference between that and DRM.

A serious question: if you purchase a recording, what legal right do you have to gain full access to the data?

monetschemist wrote:
if MQA limits my playback options to equipment containing licensed proprietary software, then it is managing digital rights.

Note that an MQA file can be copied, shared, just with any other FLAC file. But without an MQA-enabled decoder, it will only play as a baseband 16-bit file. If you don't want to upgrade your DAC, then you still have access to the music in some form. Your choice, as it is with the LP: if you don't want to spend a lot of money on an LP player, you can still access the music, albeit in degraded form.

monetschemist wrote:
if the whole recording industry gets on the MQA bandwagon and only releases their high-quality data in MQA format, that sounds like collusion and monopoly to me and not just [MQA] being compensated for their investment in intellectual property.

It was exactly the same situation regarding IP with DVD and Blu-ray. Take a look at the number of licenses involved in releasing a disc or bring a disc player to market. Why is this an objectionable scenario when it applies to music but not so with movies?

Incidentally, MQA is a separate company from Meridian. Though Bob Stuart remains chairman of Meridian, he no longer has any day-today involvement in that company.

John Atkinson
Editor, Stereophile

crenca's picture

Now this is the kind of questioning, reporting, and discussion around MQA, IP, proprietary vs closed formats, and copyright law and "fair use" that needs to occur before our music digital ecosystem begins to look like our video digital ecosystem.

I won't answer any question you ask specifically (others hopefully will chime in) except to say that for the vast majority of music lovers (i.e the "consumer") the recreation of our digital music worlds into one like our video world would be quite objectionable indeed. Granted, most won't ever take the time to actually understand what this would mean, but an INFORMED consumer would reject such a scenario 9 out of 10 times, based simply on self interest alone (putting aside any principled objection). The DMCA is an act of congress, written for the benefit of a few and for the use and abuse of the many (aka the "consumer") enforced by the power of the gun (the only power the government has). Our digital music lives have been free of the worse aspects of this law and the video world precisely because we stand on "open" formats (PCM/DSD). Sure, this has largely been by accident but now MQA is presenting us with a choice - let us understand that choice.

I suppose publications like yours are going to have to make a choice themselves. What is your position vis-a-vis the audiophile and music lover? Do you exist to inform them and in some sense support their interests, or do you exist to solely to promote the interests of "the industry" or some part there of? If you are going to openly promote MQA then you are of course doing the latter. If you are going to in some sense help the audiophile/music lover, or at least keep some sort of balance in reporting, then Clinton-esque "...that depends on what the definition of is, is..." speak around DRM and what MQA means to the digital music ecosystem will of course have to stop and a much more forthright and honest discussion will have to take place...

monetschemist's picture

Mr. Atkinson,

I really appreciate your careful and reasoned discussion of this matter. It feels quite important to me at least and it's nice not to just be blown off.

I don't feel very comfortable addressing my legal rights to obtain full access to something I purchase. I'd rather speak of my preference.

I have already been fooled twice in the past by DRM or what ends up behaving like DRM: first in the case of DRMed music purchased from Apple, and second in the case of HDCD-encoded music. Now going on 10+ years later, there are no Apple devices in our home and so that DRMed music is no longer usable without paying ransom to Apple to un-DRM it. And my Linn Genki is the only device I have on hand that reads my small collection of HDCDs, and it forces me to go back to the physical media. I appreciate that Ayre has made the effort to build HDCD capability into its new player but note that it had to wait for the HDCD patent to expire to be able to do so.

So my preference is not to get fooled a third time by MQA. I would rather continue to purchase music in open format (preferably FLAC) with no proprietary encoding, so that as long as there are FLAC decoders out there, I can listen to my music.

Therefore, as long as I have a choice, I won't buy MQA; and I really hope that music companies will not see MQA as part of a solution to music piracy and stop supporting open formats.

Anyway, as far as music piracy goes, I imagine there are two kinds of music pirates: those who don't care about the extra proprietary content in MQA and will be satisfied with stealing the low-res content; and those who do care, and will reverse-engineer the protocol so they can steal the hi-res content. If I am correct, all MQA is really going to do is restrict the legitimate enjoyment of hi-res music.

As to DVD and Blu-ray, those are absolutely objectionable scenarios as well. Buying a DVD of an interesting movie while on vacation in France and not being able to watch it back home in Vancouver? How is there any value in that to consumers? How does that not create a distribution monopoly that begs to be investigated by antitrust authorities? Oh sorry it's Hollywood forget I asked that.

As to MQA and Meridian being separate entities, I apologize to the readers here, to you and to both companies for not understanding that.

And thanks again for taking the time to discuss this in such a gentle and reasonable fashion.

crenca's picture

We really only have two "lossless" encodings, PCM and PDM (DSD), all the others being "lossy". MQA is in fact a lossy encoding as pointed out by Mr. Hanson and others, though how much it will matter in these ultrasonic frequencies it encodes with lossy compression remains to be seen (heard). Based on the glowing reports of MQA's SQ, I am not sure it does.

Flac is a open source lossless compression "container" for PCM encoded audio...

RichT's picture

Hi Monetschemist,

Sorry to hear about your painful experiences with DRM! I feel the same way about DVD region control - an unjust restriction on the consumer.

MQA, I think, has more in common with your HDCD experience than your Apple experience. It all depends, really, on whether MQA takes off in such a way that it becomes a format with universal support and a long lifetime. In my view it's too early to say.

I may take the risk and go for MQA before we know for sure. At least, in the streaming world, I won't end up with a lot of useless content, just a useless decoder!

best wishes

Rich

Charles Hansen's picture

Hello Monetschemist,

You will be pleased to know that the HDCD patents have expired. There are currently software solutions from both Foobar and dBpoweramp that allow the ripping of HDCD files with decoding of the HDCD compansion algorithms.

I've not looked at all of the ones in great detail but the ones I've seen have noticeable errors in their implementation. Ayre had carefully reverse-engineered the original algorithms and now that the patents have expired the new Ayre QX-5 Twenty Digital Hub will decode HDCD files perfectly.

All discs mastered with the Pacific Microsonics A/D converters will light up the "HDCD" light regardless if the mastering engineer engaged either of the two features that can be decoded - Peak Extend (PE) and Low-Level Expansion (LLE). If the QX-5 Twenty detects the presence of a file that actually requires decoding, it will illuminate a "d" in front of the "44" sample rate to show that it is decoding an HDCD disc that requires decoding.

Not all do - even Reference Recordings discontinued the use of PE and LLE in all their recordings made after 2009, presumably because of the lack of DACs that could decode the compansion schemes once Pacific Microsonics discontinued the chips that could decode them. There are a few DVD and Blu-ray players that used the DSP capabilities built into them to decode HDCD in software.

Not that it is a huge feature. LLE only engages when the peak signal level falls below -45 dBFS. This only happens in popular music during fadeouts, and in extremely quiet passages of classical music when only one or two instruments are playing. PE is arguably more important, but again only begins to take effect when the maximum signal level exceeds -9 dBFS.

If you want to look at this behavior, download Foobar and the HDCD plug-in and go into the "File - Preferences - Default User Interface" window. In the middle text box labeled "Status Bar", replace the default text with the following string:

%codec% | %samplerate% Hz | $info(bitspersample) bits | %channels% | %playback_time%[ / %length%] | %bitrate% kbps | HDCD = $if(%__hdcd%,'yes | ','no')$if(%__hdcd%,PE: %__hdcd_peak_extend% LLE: %__hdcd_gain% TF: %__hdcd_transient_filter%)

Then to monitor the signal levels, enable the "Peak Level Meters". I like to make them nice and large, filing up the area below the playlist. They are well implemented, with a minimum level of -60 dB, very fast acting, and with vertical bars that hold the peak values for a second or so. One of the more useful pieces of freeware out there (Windows only).

Hope this helps.

monetschemist's picture

Mr Hansen, thank you very much for this clear and insightful commentary on HDCD. Very kind.

I may well try your suggestion out though I first have to find a Windows machine to borrow (as I am a Linux guy). I have the faint hope of ripping my small collection of HDCDs to 44.1/24 so that I don't need to decode them in the future.

I find in your description of the current situation - features in the specification often unused, uncertainty in the minds of other implementors and erroneous results - two of the precise reasons why "the whole HDCD thing" makes me very uncomfortable with MQA, why I prefer an open standard for digital audio decoding and therefore why I hope to never need to buy MQA-encoded music.

Again, thank you for your generous and detailed explanation.

germay0653's picture

I'm just amazed at the closed mindedness of some of the comments. Keep an open mind, read and educate yourself about the process, listen and then make your own decision. As stated by Mr. Stuart, there is no "rights" management involved with this.

What piques my interest is the adjustment, for the better, of the time (temporal) smearing caused by the current digital recording/playback process. If you don't have a DAC that decodes MQA you still get 16/44.1 CD quality playback so how can this be viewed as requiring permission to be decoded.

You get to have your cake and eat it too. If you don't like cake, walk away and move on.

Archimago's picture

Mmmmm.... What's wrong with discussing concerns about a software technique and the likely limitations? It's just a partially encoded PCM file making various claims that can be objectively tested.

Nothing to do with open/close mindedness. If the technique has merit, it will succeed. If not, and presents too much burden in terms of cost (including need for hardware upgrade) for little benefit, it will fail.

germay0653's picture

Some have already vehemently condemned it without objective testing results to back up their contention and if that's not closed minded, I don't know what else to call it.

Time and testing, both objective and subjective, will determine if it has merit.

brenro's picture

I've been reading article after article about MQA and certainly want to experience it. The problem has been very little available music to make me want to purchase equipment that can decode it. (Still waiting on you Tidal). This news is definitely a step in the right direction.

fetuso's picture

I'm keeping an open mind on this MQA business, but I do have one gripe about Jason's post. In the first paragraph he called MQA the "superior...format." How do we know that at this point? So few people have actually heard it that I find that statement a bit like putting the cart before the horse.

Jason Victor Serinus's picture

The bottom line is this. I have reported on multiple demonstrations of MQA. So have others in the audiophile press, both from Stereophile and other magazines. The most convincing demonstration I've experienced was at CES 2016, when John Atkinson, Michael Fremer, Peter McGrath, myself, and others sat in a room and heard the differences that MQA made to Peter's recordings, as well as to others. (That was on Wilson loudspeakers, which Peter knows inside out, and had set up himself.) The differences in realism between before and after files were remarkable. This was reported honestly, and is available in the archives for anyone to read: http://www.stereophile.com/content/mqas-sound-convinces-hardened-showgoers .

No one has to believe any of us, of course. Then again, if you don't tend to trust any of us, why in the world are you bothering to read what we say? To indulge in "get the critic" or, in the words of the despicable "Crenca" who hides behind a handle because he's a coward, to call out "persecuted homosexuals"? Although I must congratulate him on finding the right buzzwords to actually get my goat - it will not happen again, no matter how vile the words or insinuations - it is time to move on.

When I hear a format that makes a major difference in sound quality, and it touches me, I am enthusiastic. In fact, I am joyous, because the communicative powers of music, and the truths about our human existence that great music conveys, are central to my life. I have yet to find a single reason on planet Earth why I should hide or even temper my joy and excitement because they disturb those who revel in negativity and cynicism. Hence, I shall continue to share my joy, enthusiasm, and support for a format that, to these ears (and, I might add, to the ears of people in an increasing number of software and hardware and music producing companies), makes a major difference to sound quality.

jason victor serinus

fetuso's picture

Jason, I appreciate the response, but why so defensive? There was nothing in my comment that was a put down or an attempt at a war of words. The concerns that some people have about MQA are legitimate. There is a certain sense of "here we go again" involved here and some might just be fatigued by it all. My concerns are these; what about my equipment and all the music I already own? Or does MQA only matter if you're a music streamer? Anyway, I meant no harm and I hope you're enjoying the show.

RichT's picture

Hi Jason, please do continue to share your enthusiasm. I'm really looking forward to hearing MQA.

crenca's picture

...and would otherwise welcome a significant technical advance in high fidelity recording. Indeed, if I have not made it clear allow me to do so now: I believe you! I believe you when you say that MQA is a real and significant advance in SQ compared to PCM/DSD (caveat: I reserve the right to question just how much it is over well recorded high-sample rate PCM/DSD - obviously such a statement would need to be expanded a great deal).

However, at what price? Digital formats are sort of the "ground" that the rest of the digital world (and on top of that, our physical playback chains like speakers, amps, cables, etc.) rest on. If I was to invent a new version of TCP/IP, patent it, and was able to get the Internet to run on it, well that would be a "revolution" in many ways, quite apart from any narrow technical advantages my new and improved TCP/IP would bring.

It is this larger picture, these aspects of MQA quite apart from its real SQ advantages that I point to. This is not "cynicism" because the change(s) that MQA (or something like it) would bring to the music world and our digital ecosystems is real and much larger than mere fidelity.

We tend to think of technological advancement as a good thing, and it can be. However, when most of us think of the invention of the thermonuclear bomb we begin to see the "larger picture" and how technology is not a simple moral equation. That is of course extreme, and MQA is of course not to be compared to nuclear weapons, but it has the potential (if it in fact gains significant market share or becomes the de facto digital music standard) to revolutionize our digital music worlds in ways that simply are not being discussed nearly enough. Again, this is not "cynicism"...

RichT's picture

Cenca,

I agree with what you say here - it could well have very wide effects and we should think about what they are.

best wishes

Rich

Jason Victor Serinus's picture

Some companies are announcing MQA firmware upgrades to existing products at no cost. But due to system architecture, that isn't always possible. Regardless, those without MQA-equipped servers, DACs, and players will still be able to listen to MQA-encoded files, albeit without reaping all the sonic benefits.

sasami's picture

All MQA is to "zip" an audio in a form of stream. Unlike "ZIP" we used compress files and folders, MQA compress stream coming to you so we can use less computer "bandwidth" or in computer terms less line speed to get the file transfer in real time. So MQA enable devices is just devices able to decompress MQA "zip" code. All Meridian's demo try to do is mix us up with audio term with computer term and make sure no one understand them instead of making it simple.

I have no idea what the fuzz about. MQA might be useful 10 years ago. Today with line speed we have and massive hard drive. What's the use of MQA? The only party is going to benefit from MQA is the companies doing streaming. Because that mean lees total bandwidth and less hard drive. Will they lower the prices to pass onto consumers, doubted that would ever happen.

At the end of the day this is nothing to do with end users, MQA is more to do with big companies. If they all go MQA like they did with HDCP and HDMI. We will have to adopted by buying MQA compatible DAC or steamers. For me MQA means bad new for consumers just like HDCP and HDMI.

RichT's picture

@sasami,

no it's much more sophisticated than this. Have a look at
http://www.computeraudiophile.com/content/694-comprehensive-q-mqa-s-bob-stuart/

sasami's picture

And that the reason why I said Meridian is try to confuse consumers. After I read a few on those Meridian's MQA report, I suddenly remember I saw the same origami graph they shown somewhere before on one of my IEEE spectrum. Surely I found the German white paper that was explain how that works. I am not saying it is not sophisticated frankly to compress stream as it goes is very difficult math, frankly even zip we use these day contain so much math is not funny. What I am saying is Meridian just won't explained it in a way that everyone can understand. I guess that is how they sell stuff, keep the Voodoo.

Anon2's picture

As I've written before, this is all thoughtful innovation and worthy effort. But I am waiting for some stability and standardization in all of this. Expenditures on the latest and greatest from 4 years ago would now be expenditures on disfavored, if not out-and-out obsolete technology.

I'll just sit on the sidelines, with my CDs, SACDs and excellent, free (but lower-rez) streaming options (RNE, BBC, Swiss Radio) until this all gets sorted out. This is all great work and great engineering, but it's as if you had to change the fuel intake system in your car each year due to a new type of fuel to burn. When we get to Regular Unleaded, Mid Octane, and Supreme, then I'll get interested.

steve21's picture

So much written about a format that has been heard by so few (so far). And perhaps that's why the opinions seem so strong and comments so personal. MQA seems very complex, understanding is not for the layman. But really, who cares, if it makes steaming music better, bring it on and let the consumer decide. The way I see it the pendulum went all the way to MP3 and now it's swinging back. My question would be why would you not expect new discoveries in music technologies going forward, there will always be new discoveries.

doak's picture

Just WHY is it that we,supposedly, NEED MQA????

Nellomilanese's picture

we don't need it. Jason heard a Meridian MQA demo that sounded better so now he's convinced he's seen the light LOL little he knows that when the magician controls the stage and the gear he can make everything sound "better".
WE have F***IN PERFECT formats...they're called Vinyl and Sacd (when it's real dsd not just converted cds).
Since the PureAudio Bluray colossal FLOP they need a new format to milk us, YET ANOTHER Amy Winehouse "Back to Black" or Pink Floyd "DSOTM" release LOL
Warner representative: "Hey dude I promise this time IT WILL SOUND BETTER" hahaha

RichT's picture

Because, if what people say is true, it sounds better. If you're happy with the sound you have, then you don't need it!

doak's picture

And be compatible with my digital signal chain - sounds SO good I don't want to change it.

RichT's picture

Yes, understood.

jimtavegia's picture

I have never done a recording for anyone in my home studio where I track at at least 2496, and some at 24192 and then use Sony Sound Forge to convert it to redbook and not hear how much better the hirezs files sound.

I have even taken the recording feed and recorded at redbook with no need for conversion and not say that the converted files sound at least as good as the native redbook did. That was just to allow them to decide if Sony Sound Forge did something bad in the conversion. HIrez wins every time and when burned on a DVD-r can be played back in most DVD players.

I have read that everyone who has heard MQA loved what it did for improving the sound, except for one here. I feel sorry for those who don't understand the math and why hirez IS better. If MQA lives up to its promise then how could anyone be against it? Makes no sense to me.

Better is just better even outside of some conspiracy theory. JA and many on his staff have forgotten more than I will ever know, and I know they are not easily swayed.

volvic's picture

A voice of reason, well said man.

crenca's picture

Respectfully, I am not sure how to respond to such statements such as:

"If MQA lives up to its promise then how could anyone be against it?...Better is just better even outside of some conspiracy theory."

except to state the obvious - no, "better is not better" unless one judges something (anything) by just one single criteria. Who does that? Nowhere, not in any single area of life, do you narrow something down to a single quality and ignore all others and if you did you literally would not survive the day as you would do things like deciding that you were hungry, and thus getting to the restaurant/store the fastest was the best way to satisfy your hunger and so driving at the highest rate your vehicle is capable of "is better" and you would of course crash and kill yourself and possibly others.

Humans are complex animals, who live in a complex environment and have to make decisions on multiple criteria and facts. We can not afford to look at things in a simplistic, "SQ = good, thus MQA = good" way.

You mentioned you record and/or create music. Perhaps this is where the divide lies, between those in "the industry" (i.e artists, producers, recording engineers, labels, manufactures, the "audiophile press", etc. etc.) and the actual flesh and blood "consumer" of music. We have to purchase and store and otherwise enjoy our music, so we have to think about our digital ecosystems. "The industry" on the other hand has the luxury of largely ignoring this fact, and can thus afford to focus on other things such as sound quality (or DRM, or whatever else is in their interest) and ignore the consumer's interest. Indeed, they in fact BLAME the consumer (e.g. when they tell us music has no intrinsic value anymore) for their current predicament (which is serious) and feel like they can demand DRM and other objectionable things from the consumer - and if the consumer actually notices or complains, well then it is a "conspiracy theory".

But hey, perhaps this sort of anti-consumer thinking has been working for "the industry" for a long time and it is just a matter of course. I would be interested in thoughts on this...

jimtavegia's picture

People often do make decisions based on just "one criteria". Often it is price, isn't it? I have heard it a thousand times, "I would never buy a $1,000 CD player". "Have you heard one," I asked? "No, but that is just too much money," they replied. No other criteria..."Not shinny enough, too big, my wife won't like it," never came up. Just the money issue.

If you don't believe that an "improvement" is not made then don't buy the product. I did not choose to record at 2496 or higher just because I could...there must be a sonic reason to do so. If my clients can hear it and want it then there is no harm and no foul. Pretty simple stuff, actually. When they can still hear the improvement of 2496 in their cheap DVD players over the redbook version, why not do it?

I can tell you that a better and higher sample rate is no different than better mastering. My wife has a favorite artist and her first redbook release I had a hard time understanding some of her lyrics she sang. Kind of sad I thought. I bought her second album and it sounded so much better I was surprised. Recorded and produced by the same people. The only difference I could tell is that the 2nd was mastered by Ted Jensen at Sterling Sound. Is he the equivalent of MQA? Better is better. Did she use a different mic or mic pre? I have no way of knowing, except that she and her team thought the first album was good enough to put out, but to me, the second was much better "sounding" by far.

I also ran into the same thing with another artist and my wife kept asking me why I played one of the 4 albums from this artist more than the others and I said it just sounded better. Further inspection of the booklets told the story. Sure enough it was the only one mastered by Ted Jensen and recorded at the old Ocean Way Studios. No surprise there to me. Those were not subtle differences either, but my wife never gave it a thought.

There are just some things that make music sound better...different microphones and mic preamps, where they are put and then how is it tracked (analog or 24/48 which is most popular) or by someone who really cared and did it at 2496 or better (DSD?) ) It all matters, but ultimately it only matters if you can hear it and feel like it is an improvement. There is no other reason to do it differently unless it sound better to a number of people in the decision making chain. Can you market it? If enough people can discern an improvement on a very revealing system(s) then who is to argue against MQA if you haven't heard it and been a part of the testing.

Does someone really want to sing a solo through an Shure SM-58 when the Neumann KMS 105 or a Sennheiser e965 is around and you can afford it? If the SM-58 is the sound you want, then that IS what you should do. There is the reason more of them have been sold than any other vocal mic. It is a great $99 mic. It also doesn't mean it is the best.

I know that the idea of "better" is lost on many and it is all an opinion, but if enough trained listeners think it so and have heard it on a remarkable high end system that reveals "fatal flaws", how should anyone doubt it? It is the same reason that many will spend thousands on a particular mic preamp looking for that once in a life-time sound that no $300 mic pre can offer.

The sad part is that the masses don't really care about this sort of thing unless it is free and maybe at some point in their gear buying life it will dawn on them that the MQA discs or downloads do sound better, but they will have to make listening and judging something better than a secondary life event. Sit down and just listen, please. Plus, if the record labels are not going to start tracking everything at 24/96 or 24/192 then this is an even bigger deal in MQA making our music sound better in the way older music was tracked. I for one can certainly not be against that. The fact that analog guru, Michael Fremer, heard an improvement "in digital", that is something! I have certainly enjoyed his "experiments in providing us sound files of various phono cartridges and phono preamps to test our hearing acumen. Better is always in the ears of the beholder. I like being challenged to hear what many think is "better".

If the sound quality of the music you listen to is not important, then why be a part of this discussion at all? If your listening is just background filler so be it. This may not be for you. Audiophile ideas and interests are not for everyone. I get that. This is not a conversion experience that I worry about at all.

crenca's picture

It is often THE criteria is it not? However, the functionality (of whatever widget is being considered) is sort of taken for granted. This is where the hype around MQA is accurate (or could be if it becomes the standard) is that is a change in functionality in a way that most are not aware of - the place of formats in a digital ecosystem.

I thought I made it clear but I will say it again - I believe the reports about the sound quality! I don't doubt that MQA can do wonderful things with certain recordings. The only part of it I question SQ wise is whether it is really an improvement over well recorded high sample rate stuff when played back with filters that do the least amount of pre/post ringing. For the vast majority of recordings however (which as you point out are sub-quality for a host of reasons) I do not doubt that what some call its "DSP tricks" actually work and the sound is just wonderful.

However, I am not willing to do ANYTHING for a sound quality tweak. Audiophiles do worry me sometimes. I wonder just how many puppy's (or worse, humans) some would be willing to dash against the rocks for a sound quality tweak.

It is apparent to me you don't really understand the foundation(s) of our digital music ecosystem, and that's OK. Just be warned, MQA is a potential "revolution" in ways far removed from the SQ aspect, and as you are no doubt aware "revolutions" have their pros and cons (many of them quite bad). These aspects of MQA are things well informed consumers/audiophiles are interested and should be. I won't say something like "digital computer audiophile ideas may not be for your" because if you are in fact purchasing digital music of any kind, you really ought to be well informed...

John Atkinson's picture
jimtavegia wrote:
If you don't believe that an "improvement" is not made then don't buy the product.

An excellent point, Jim, in what might well be the most salient posting in this thread.

John Atkinson
Editor, Stereophile

Anton's picture

I have sat by quietly as the magazine has extolled the virtues of MQA as heard at shows/demonstrations. Fine.

I have sat by as the staff has scolded me for asking about how something sounds at a show or demonstration, being told to keep it down because how can y'all be expected to report on how something sounds based only on demonstrations?

Consumer advocates, corporate cheerleaders, or what? Is there an official ethos of the magazine?

We have been through CD's perfect sound forever, SACD, Blu Ray, etc...so why such a vehement defense of a compression algorithm, of all things? I don't get that part. The affection of Stereophile for MQA feels a little disproportionate.

Does Stereophile have a concern about MQA insinuating itself in a negative way into a hobby that is based on seeking the highest quality possible, as others have mentioned on this thread?

I would like to see some advocacy, as a consumer, from the flagship publication of my hobby!

I don't mean any of that in a mean way, I am just getting mixed signals from you folks on this 'development.'

Is it an industry development, or an audiophile development?

John Atkinson's picture
Anton wrote:
I would like to see some advocacy, as a consumer, from the flagship publication of my hobby! I don't mean any of that in a mean way, I am just getting mixed signals from you folks on this 'development.'

I have to admit that I don't understand your point. We have reported on the technology behind MQA in-depth and have pointed out the non-sound-quality benefits to both consumer (the reduction in streaming bandwidth) and the record industry (single inventory, no longer making their hi-res masters available in bit-true form). The latter is, I believe, what has prevented the record industry from making more than a small fraction of their catalog available in hi-rez format.

we have also reported, enthusiastically I admit, on what we have found to be a significant improvement in sound quality offered by MQA-encoded files in direct comparison with the hi-rez originals. This magazine was founded on the idea that better sound quality is what is important. And MQA does appear to offer better sound quality, regardless of all the other social and marketing issues that have been raised in this thread.

Anton wrote:
Is it an industry development, or an audiophile development?

MQA is both, we feel.

John Atkinson
Editor, Stereophile

crenca's picture

Interesting how you put the consumer's interest as a "social and marketing" issue - like the obvious disadvantages of MQA (such as the fact that it is DRM all day every day) are problems to be solved by the industry in a marketing sense - it just has not been SOLD to us better... ;)

Buyer beware...and no Anton, your perspective is just not on the radar - it is about changing your perspective (i.e. selling you something), not trying to see it from your perspective and inform you and help you from your very legitimate position. At least you know where you stand as a consumer! It is for this reason that much derided "forums" exist. Despite their drawbacks (which are many and obvious), they are the only part of "the industry" that you are going to see your position (i.e. as a consumer and music lover and audiophile) represented and discussed, or even acknowledged as legitimate instead of being dismissed entirely as "conspiracy theory" and the like.

janszenLabs's picture

Hi John,

I know I'm quite late to this party, and it seems the discussion of shades of DRM and of aspects of public disclosures about the algorithm have been pretty well exhausted here and elsewhere. Still, I must say I'm intrigued by your statement that you've found ". . . a significant improvement in sound quality offered by MQA-encoded files in direct comparison with the hi-rez originals."

One would not expect this result from an algorithm that's lossless or only lossy in the ultrasonic spectrum. On equipment that is optimized for hi-rez file playback (where I presume hi-rez refers to both high data rate and high resolution), vs. the same or other equipment that is level matched and equally well optimized for MQA playback, the original and the MQA would [hopefully] be indistinguishable from one another.

Unless the equipment is better suited to MQA playback than to the hi-rez originals, your statement implies that MQA adds something beyond what is present in the original audio band data. It might be thus called gainful compression. At any rate, any alteration of the audio band data makes it not what it was, so the scheme would be thus lossy in a strict sense in the audio band as well as more conventionally lossy above it.

I have seen some mention of how MQA can improve time domain aspects of playback, but a hi-rez file on proper equipment shouldn't suffer from time domain issues. Even if it did, I don't see where one would find the reference(s) within the data for impulse response corrections, for instance. Finally, assuming something is being done to improve the sound in MQA, I'd expect it should be possible without the compression baggage, which would make for a nice licensing opportunity in its own right.

Anyway, I haven't found a definitive explanation of what would account for an improvement over the original, and would be grateful if you pointed me in the right direction, or offered your own explanation.

DJ

synth's picture

For the weary consumer facing the prospect of yet another format requiring yet another hardware purchase I suspect the introduction of MQA will be met with indifference. For those of us with expensive DAC's that will never see an upgrade path such as a firmware/OS update to incorporate MQA we're only mildly interested at best in MQA or any other proprietory format. There's a wide gulf between reviewers who get to play with the latest tech basically free of charge and the consumer expected to keep coughing up to keep up.

crenca's picture

John Doe walks into the Verizon store because his contract is up and it is time to get a new phone. He notices among the list of features (#138 in the list) something called "MQA". He asks the sales person "what's that?". Clueless, the salesperson regurgitates his training "you need it for music". They move on.

This is how the DRMing of our digital music world will happen IMO, and of course streaming's role in this is obvious...

Jason Victor Serinus's picture

Am I the only one who's feeling that this conversation is getting more than a bit apocalyptic, as though MQA is the final straw that will take us all down, earbuds or no?

As for the Verizon store scenario, the penultimate time I walked into a Verizon store, in Alameda, CA, I was followed maybe 10 minutes later by three ski-masked gunmen intent on robbing the place. I went into surrender mode and did just fine, but many of the folks around me began to hyperventilate and grow hysterical. Thank goodness the store didn't have a panic button, because summoning the police in the middle of the heist would have no doubt created a hostage situation. As it turned out, the robbers left my iPhone on the counter. It remained in my possession until my coat was stolen in SEATAC maybe a month ago, with my iPhone in its pocket. Hence my most recent trip to the Verizon store.

Talk of a potential hostage situation inexorably leads me back to this thread. Surrender mode may be the best approach.

crenca's picture

Reminds me of a story my grandmother told about her being in a similar situation.

It is true, I have countered what I have called the "MQA promotion machine" with strong rhetoric that approaches (but does not equal) the terms with which MQA is being described (such as your first paragraph). If I am being "apocalyptic" it is a natural re-action to the "second coming" rhetoric of the MQA promotion machine. To Anton's excellent inquiry, Mr. Atkinson could not get himself to admit of a single downside of MQA at all. If true, it is a singular invention in history, littered as it is with mans technical innovations which are always problematic in some form or other - there is ALWAYS a downside.

If (or when) the transformation of our digital music ecosystem to a proprietary/DRM format occurs (whether it is MQA or something else like it), a kind of surrender will have occurred by the music lover (aka "consumer") obviously. This is not to say that it will be all "downside". Like I said, I believe you when you explicate the real SQ advantages of MQA. Also, "the industry" will likely $benefit$ in some way but for reasons I and others have said, I don't think the DRMing of our music will substantially change the industries predicament or on the whole benefit the consumer.

My Verizon scenerio has a serious problem - unlike many of the "audiophile" editors and posters here, the 40 and under crowd has a much stronger sense of what I have called the "digital ecosystem", formats, and what all that means. They (as a whole) are not so easily duped into DRM, proprietary formats, subscription/toll based software, etc. So we got that going for us... ;)

kenkirk's picture

This has brought back memories of Rec.Audio.high end or whatever that cesspool of flame throwing mess was called! Yes Mr Atkinson I think Arny has come back from the hive and has found his way here if in spirit only. Haha. :)

Ken

Jeffreylee's picture

After wading through all of this, I'll stick with used records. Jesus.

jimtavegia's picture

I don't have any issue other then when responders get in personal attacks and think there is something dishonest going on with no evidence in sight. This is probably the most opinionated hobby around I would think.

Yet, it all comes back to how does it sound and do YOU think it is better on YOUR gear? I will never come close to owning anything JA listens on, but when I can hear improvements on my middle of the road gear I can hop on that A Train all day long, and with the new, great headphones and amps coming out, joining the party is even more affordable.

If there was something out there that could get the younger folks to reverse a trend of ITunes downloads and at least try something else and see if they like it is a great thing. And even improving streaming quality is not a small thing either.

I am just interested in hearing it for the first time and get my impressions of what it can do. Now that my teaching school year being over is 7 days away I will have some time to truly spend some time to focus on MQA and see what it can do.

Now would be a good time for JA and his folks to let us in on how and where do we get started to begin to audition MQA for ourselves in the best way we can.

Jason Victor Serinus's picture

Gosh. I'm hardly an acolyte. But I can answer this question.

When MQA events are announced - in fact, when all store and showroom events are announced - Stereophile runs blurbs sharing the news. If we receive a press release about an MQA demonstration at an audio show, and it's sent maybe 10 days before show's opening, I attempt to include it in my show previews. If nothing else, I post it in the Comments section. And if we are informed about a demo at either a store or audio society meeting well in advance, the announcement may appear in the print and digital download editions of the magazine.

However, we are not marketers. What we cannot do, and have no intention of doing, is either scheduling MQA events, or controlling the sound quality at those events. I, for example, attended an MQA demo over a year ago at, I think, AXPONA that was conducted on equipment so poorly set-up that the differences I've heard in other demos were hard to distinguish. I walked out of the room and then reported my disappointment in my show blog. I've also criticized MQA demos on more than one occasion that only played MQA-encoded tracks, and did not offer the opportunity to compare before and after.

wozwoz's picture

1. I followed the link to the article on MQA: from the diagrams, MQA is clearly a lossy format, which is going backwards from DSD/SACD hi-res, not forwards.

2. Mass market consumers don't like DRM, so I suspect this is dead on arrival as a mass-market format, and especially if fresh hardware is required. The simple evidence of the last 15 years is that consumers don't care about music quality, preferring convenience (small downloads etc - even to the extent of listening on their phones). For the audiophile market, which is tiny, well, they are already well served by SACD for those who like physical formats, and by FLAC and various hi-res download formats for those prefer downloads. So what is new? DRM?

3. I haven't heard Warner's MQA format yet, so have no comment on that. But I would comment, briefly, that, in my experience, Warner's conversion of analog recordings to SACD have been highly suboptimal, adopting an approach that --- instead of simply taking an analog master and converting it directly to DSD -- instead adopts some convoluted de-greasing de-hissing protocol which involves several unnecessary D/A and A/D conversions. Aside from being technically suboptimal, the proof of the pudding is in the listening, and I have been most impressed with almost all the EMI / Warner SACD releases that I have heard, to the extent that I no longer purchase them. And what worries me is ... if they can't get something simple like converting an analog recording to SACD right, what hope does a compressed format have, which will probably involve the same de-greasing of the masters along the way. Purity of approach starts as a philosophy all the way through.

X