More on MQA

An economy of information transmitted . . . what was encoded was only what was needed, nothing more. (footnote 1)

As I wrote in the January issue's "As We See It," Master Quality Authenticated (MQA), the encoding/decoding system developed by J. Robert (Bob) Stuart and Peter Craven, has been widely criticized, despite reports in this magazine and others that MQA-encoded files tend to sound better than the PCM originals from which they were derived. Also in last month's issue, Jim Austin investigated the time-domain performance of the MQA reconstruction filter and I examined some of the more general aspects, ending with: "Other criticisms of MQA involve its implications for the recording industry, for manufacturers of audio products, and for consumers. I will examine those in next month's 'As We See It.'"

In this issue (p.125), in the second part of his ongoing series on MQA, Jim Austin looks at the subject in terms of information theory. In Claude Shannon's coining of this term, information, bandwidth, and time could be considered as "three precise, swappable quantities [that] could show which ideas for sending messages were 'within the realm of physical possibility.'" (footnote 2) MQA appears to use this equivalence to achieve the goal expressed at the start of this essay: what is encoded is only what is needed.

But despite what I consider the theoretical elegance of MQA, others are unconvinced (footnote 3), or even unwilling to go on the record for or against it. At the 2017 AXPONA show, for example, Jana Dagdagan spoke to industry folk and audiophiles about MQA, but when she asked if they would be willing to share their thoughts on camera, more than half gave variations of these answers: "I don't want to publicly state my opinion of MQA because the company I work with hasn't decided yet." "I don't want people to know I'm not a fan of MQA because we may eventually adopt it." "I like MQA, but the company I work for doesn't support it right now." "I haven't blind-tested MQA yet, and I don't want to share my opinion without any backing."

One criticism that is repeatedly raised is that MQA implements Digital Rights Management (DRM). Yet there is nothing to prevent people from copying, e-mailing, and sharing MQA files (though selling them would presumably run afoul of the Digital Millennium Copyright Act). I have copies of the same MQA-encoded files on all the various servers I use for my listening, and they all decode perfectly well. So when people refer to "DRM," they are actually talking about the fact that you can't obtain the full hi-rez resolution from an MQA file unless you have an MQA-enabled digital processor.

This is an exact parallel with HDCD, from the 1990s. If you had an HDCD-capable DAC, you would get 1–2 bits greater resolution from what appeared to be a regular CD. If you didn't want to pay for an HDCD decoder, you could still play the CD but at lower effective resolution. In that sense, the same "DRM" argument could be made about HDCD: It restricted the quality available to people who lacked an HDCD decoder.

If an MQA file is played without MQA decoding, the sound quality will be that of the baseband file—ie, the same as a CD—meaning that the record company need stock only a single inventory. As well as the bandwidth benefit for streaming, there is another commercial benefit for the record industry with MQA that is not true of lossless-packing schemes such as FLAC: The record company will no longer be selling a clone of their hi-rez master. Instead, they are selling something that might well sound identical to the master, or even better than the master, but doesn't allow the master to be re-created.

This is presumably why, according to MQA Ltd.'s corporate filing in the UK, all three major record companies—Sony, Universal, and Warners—are stockholders. But as long as audiophiles can download or stream original, non-MQA, hi-rez PCM files, why should MQA be an issue? However, what if they no longer have such access to the originals, but only to their MQA versions? This, I suspect, is why so many are against the idea of MQA. As Ayre Acoustics' Charley Hansen pointed out last November on the Audio Asylum web forum: When Jim Austin interviewed MQA's Spencer Chrislu for Stereophile in September 2016, and asked if he hoped or expected that all digital music will someday be MQA-encoded, Chrislu's response was blunt: "Well, that's the goal!"

"This is why we are concerned, why we care, why we must ask questions and why the questions need to be answered," wrote Hansen, who passed away last November. "We want to get this one right. The stakes are too high to get it wrong!"

We live in an era of corporate consolidation—see how Amazon, Google, Facebook, and YouTube now dominate our lives—hence potentially reduced choice. And as I write these words, the FCC has announced the end of Net Neutrality, which I regard as the bedrock of the communication of information. Regardless of MQA's technical elegance and promised increase in sound quality, the removal of consumer choice in recorded music is indeed a relevant issue.—John Atkinson



Footnote 1: From a discussion of Homer Dudley's 1939 invention of the Vocoder, from Jimmy Soni and Rob Goodman, A Mind at Play: How Claude Shannon Invented the Information Age, p.100 (Simon & Schuster: 2017).

Footnote 2: Shannon adopted Ralph Hartley's term information for his landmark 1948 paper, "A Mathematical Theory of Communication" (A Mind at Play, p.135).

Footnote 3: See the lengthy analysis by Stephan Hotto of XiVero, a German company I understand is used by HDtracks and Highresaudio to check high-resolution files before offering them for download.

COMMENTS
tjf's picture

this is the way forward...no data can be considered "free" anymore, and tech like blockchain, digital currency, AI (machine learning, if you'd prefer), etc. & the "cloud"-- will ensure that every use of data can be accounted for, certainly in terms of "ownership"....just not the end-user/consumer's ownership...
End-user physical storage & end-user ownership rights to the data on that storage medium, thy days are numbered...

dalethorn's picture

I don't think you'll ever see audiophile users forced to stream music that they've purchased if they prefer to play it from local storage. There are numerous reasons, the first being that the "Web" will never be reliable enough to replace local storage, and the second being the cost of maintaining the account and the Web connection just for music playback.

https://en.m.wikipedia.org/wiki/Possession_is_nine-tenths_of_the_law

drblank's picture

you are just buying a COPY and the rights to own a copy, but not the owner of the original files and all of the copyrights that goes along with them. You also can't sell your digital download copy to another person. Also, no one really "owns" anything since when you die, you can't take it with you, so in essence, you are merely paying for the right to listen to the file and listen to without having to pay for a monthly streamed version.

Now, here's the dilemma that a lot of people are in.

First off, If you calculate the average person's library collected within about a 50 year span. It's not that difficult to have a rather large collection of content for which one pays tens of thousands of dollars, even hundreds of thousands of dollars.

it costs a LOT less money to gain access to MQA files (catalog is growing) for a 50 year period and you can listen to ANYTHING in the catalog as long as you have an internet connection to your music system.

More and more people do want better quality audio, but simply can't afford to spend $100+ a month to buy content to build their collection. But they can afford maybe $15 to $20 a month to stream anything.

So for less money they can listen to anything, rather than paying tens or hundreds of thousands of dollars to only listen to a small portion of the entire catalog.

For relatively little money, someone can get a MQA DAC/AMP and a really nice pair of headphones, along with Roon Software if they wish to add to their computer and get high quality audio. It's now VERY affordable. MQA is great because they can actually stream HiRes beyond even 24/192 Lossless and still with not big bandwidth.

within the next year or so 5G is going to be rolling out in the US and other markets. I can actually see someday that the Video formats like 4K eventually having 5.1 MQA files giving us better than the AC3 Dolby Digital, which is where it's been for quite sometime.

I welcome MQA as it makes Hi Res more affordable to a much wider audience that's already here or soon to be here always waiting for Hi Res to be for those on a more limited budget. Not everyone can spend $10 or more on a stereo and thousands a year to buy content.

Do some math, add up how much you have paid for in your current catalog of every format you still listen to. 50years x 12 Months x $20 a month is only $12,000 for the average lifespan of someone that starts to listen to music in their teens all the way up to their 60's. Not that much money is it? I paid more than that for my CD collection and I only have 1,000 CDs, which isn't that many CDs to own for someone that's an avid collector.

dalethorn's picture

Umm, no. Your theories don't hold water in the real world. First, there's no relevancy to "when I die", even though my estate has full rights to my purchased data. And even if they didn't, somebody will get the stuff and toss most of it anyway. And furthermore, you dodged the reality of the Internet in six different ways here.
1. Paying to stream again and again for the same music? Absurd!
2. The stream will *never* be guaranteed, just "assured".
3. The quality of the stream will *never* be guaranteed, just assured.
4. The presence of my playlists and immediate availability of every track will *never* be guaranteed.
And I can add more...

dalethorn's picture

BTW, even though photos are not music, there's a great analogy to music for when I leave this mortal coil. My heirs will undoubted save a few photos and toss 98 percent of them (digital), and they'll be forever lost unless someone either pays the renewal fee for the Website (unlikely) or archives them where they can be accessed for free as now (also unlikely).

To me, the most relevant concern is the hundreds to thousands of hours I've spent building a very unique or esoteric collection from likely irreplaceable sources -- music by friends, one-off recordings, tracks gleaned from unknown sources during the original Napster days (1999-2001), masterings that are no longer available, and so on. Any future playlists in the "cloud" will not be able to access half of my tracks, because those would not be "approved" to transfer to a site that controls digital rights. And I have the rights, which can be demonstrated credibly, but lacking certificates they wouldn't pass.

tjf's picture

All your points are certainly valid, all wonderful reasons to shun a streaming model of music delivery...but, the horse is out of the barn/train has left station, etc...you get the point. The Pono debacle should be kept in mind here...
Within a few years we'll see the download sites evaporate against the might of Warner/Universal/Sony, their streaming partners..and MQA gives them a revenue-friendly format, in a number of ways..
Wonder if we'll get a blockbuster CES Apple Music announcement??

dalethorn's picture

I don't think I suggested shunning, did I? I know of ways to capture a perfect audio stream and save it into a file, no matter how many ways the streamers try to block it. As I said, the idea is to possess everything that I might want to listen to repeatedly, and streaming is just as valid a source as sneaking the Nagra into a live session. So what if the download sites diminish their activity? What I want millions of others want, and we will get around the anti-possession nonsense as surely as anything.

tonykaz's picture

With the Global ( 2.1 Billion and growing ) population of Smart Phone owners, we need an agreed upon "Standard" for our Music.

I hope for "RedBook" or even 24/44.1 to be our Gift to the Next-Generations to base their Music Collections on.

MQA seems a useful method, well-able to serve everyone's purposes !

"Everyone's purposes except we "hair-shirted audiophiles" who will continue to shower praise on the lovely $50 Vinyls we buy ( invest-in ) from Chad Kassam and the Audiophile owners of Exotic DAC design implementations ( FPGA, Ladder DAC and even that Super-Borito thing from the Shit Co., etc...).

I hope the Audio Industry learns it's lesson from all the variations of the Analog Era :

78 with it's varying speeds, 1920 to 1957

7" 45s to replace the 78s 1957 to 1967

33 Mono long play Albums 1957ish to 1965

33 Stereo long play Albums 1965 to 1985

33 180gram Stereo Long play Albums ( adjust the VTA )

33 Quad Channel Long Play Albums

45 Stereo Long Play Albums

45 180 grams Long Play Albums

Compact Cassettes ( coming back now-a-days, I'm told )

El-Cassette for the Radio Stations

8-Track Cassettes

Reel-to-Reel in all it's variations

Phew,

Audiophiles have been thru a lot of Formats

I fell in love with the concept of my entire music collection on a SD memory card the size of a finger nail.

Now, it looks like I won't even need the Memory Card.

Bring on the 21 Century,

I'm all Ears !

22st Century Tony in Michigan

drblank's picture

coming to 4K Video since they could get 24/192 files that True HD or HD Master uses, but at a considerably smaller file size so they can stream 4k video with Hi Res audio.

I think Apple recently added FLAC support since FLAC is what some of these streaming audio sites are using. Apple also added FLAC playback to iOS and macOS, so I honestly think Apple might be gearing up to support MQA playback sometime in the fairly near future. If Apple got behind MQA for both 2 channel, but multi-channel audio for TV and Movies, I think MQA will gain traction VERY quickly. Then it would be up to the rest of the stream video guys to do the same thing.

tonykaz's picture

LG of South Korea just announced some sort of committed business relationship with MQA.

I'm presuming it has to do with their leadership position in Screen technology ( OLED ) and the future of higher resolution Audio Formats being the "Standard" for all Video i.e. 24/96. ( Geez, thats a realized 120db of dynamic range ) . KER~POW!!!!!

It's starting to look like MQA is going to be around for a long time to come.

Tony in Michigan

foxhall's picture

I'll be very surprised if Apple ever adopts MQA unless they bought the company and IP; Apple tends to avoid technologies they don't wholly control or are open source. Dolby Digital in the video content they sell and deliver on iOS and TvOS is one exception and I'm certain there are more but I think you get what I mean.

spacehound's picture

One is WAV, a joint development from IBM and Microsoft, which is not a 'codec' at all but raw PCM in an 'envelope' which allows you to add tagging, pictures, etc. IBM/Microsoft put it in the public domain at no charge a very long time ago. It was designed for 'sound' of which music is just a subset. Which is why WAV tagging is so 'open' - a rocket exhaust doesn't have a 'composer' for example but you can put one in if you want and if you use 'ID' tagging everything will understand it.
Everything can play WAV.

The other is FLAC, an efficient lossless codec (more efficient that the FLAC compressed lossy MQA) that has always been free and in the public domain.
Everything except iTunes could play it but from a couple of months ago iTunes can play it too.

You don't have to buy a new DAC for either. Nor do record companies or DAC manufacturers have to pay fees.

So who needs MQA?

Why do people use MQA?
It's just another label on the box for DAC manufacturers, record labels, and streaming outfits. They hope it will be a selling point, like the eleven labels on the front of my Yamaha AV receiver, most of which labels are long dead.

Qobus, bigger than Tidal, have gone straight from 44.1 to high-res FLAC, missing the lossy MQA out completely.

DH's picture

DRM isn’t just about the ability to copy files. It’s about the ability to control how files are presented to the user.
MQA is designed with this in mind, which I’d guess it is one reason the labels like it. The single inventory (no non MQA available) is one, but all sorts of scenarios are possible.

One I fear is that DRM could be used to punish consumers who aren’t on board. MQA could be used to actually deliver a degraded file/sound to the non MQA user-something not just of “baseline” or CD quality, but altered somehow and worse.

Along those lines, it could also be used to deliver various levels of playback quality, depending on how much you pay for the file, or even give you varying playback quality (including degraded), depending on how much you pay for your streaming subscription.

That’s all in addition to the possibility that only MQA versions of files could be made available. And as you implied, what happens to consumers who don’t prefer the sound of MQA (and not everyone agrees it is an across the board SQ improvement)? Will they be left with access only to what the feel are inferior versions?

deckeda's picture

... and it makes no difference to consumers.

Higher-res files cost about double what lossless 16/44 files cost, despite economies of scale for storage and bandwidth for distributors that says it’ll never cost THEM “twice as much” to offer the higher-res version. If anything, lower-res versions should cost consumers more, if we’re looking at production costs. We pay for the privilege and benefit, of course. But hey, we’re also talking about an industry that pays digital royalties based partly on song length.

Ford offered extra horsepower on the Mustang a few years ago if you bought a special ignition key. If the car was started with it, the computer unlocked the power. That’s functionally the same as the hot rodder who bought headers in the old days, to unlock some power.

As tinkerers we want everything the manufacturer “hides” inside (because they’ve already provided it, by merely making it ...) but will spend money on 3rd party companies (Hedman Hedders, Holley carb ... or audio DACs) because someone other than the original provider is offering a workaround.

MQA fits in with all of this just fine. The most fascinating aspect, if realized, is that it could sound better than the original it’s made from.

Conceptually, I understand how and why that could be the case ... right up until the point where you have to ask, if MQA knows how to make a digital file sound better, their model of encode-decode is ultimately self serving, not consumer serving.

Imagine if no Mustang could ever be it’s best until you went straight to the auto parts store from the dealer. This is the fundamental “fuck you” about MQA, a sort of permanent drug from which addiction help is impossible.

In the end, we don’t care if it’s lossy. We don’t even care if it sounds better. There are bedrock freedoms being challenged here.

Yes, MQA deserves their time in the sun for a better mousetrap. But they’d do well to take a page from Dolby Labs, who over the years has managed to deftly swerve in and out of pro audio, consumer audio, and today more or less behind-the-scenes as another new logo on the surround receiver. We don’t blanche at paying for it, even though we do with every movie ticket, every rental, every surround receiver. There’s now way that scenario is not MQA’s goal with audio for songs.

Their challenge is swimming in shallow waters. Dolby learned years ago that audio-for-movies is where the real digital compression game is, business-wise. So much so, that when CDs killed Dolby NR revenue, Pro-Logic and its successors was already the clear path ahead.

tonykaz's picture

Low and behold,

We have an actual Audiophile Movie ( documentary ) about a "Record Collector" who's Lifetime Quest to collect all of America's vast Music History has been documented by CUBE Media in 2006.

"Desperate Man Blues" On DVD

Beginning in the 1950s, Joe Bussard begins his Adult Life collecting Records abandoned by his previous Generation of Record collectors.

It now seems that we record ( music ) collectors have a history going back to the 1920s. Hmm. These people even had Record Cleaners!, who could've guessed ?

Geez,
I'm a 3rd Generation Record Collector !

21st Century Tony in Michigan

kjackson's picture

MQA may have tiny audible improvement. I can't hear it , but maybe it does.
I hear very, very slight differences between 96K downloads and the same recording streamed 44 K on Tidal. Even audiophiles are sick of the imaginary resolution improvements in sound.
Real improvement would be better engineered recordings.
I think audiophiles would get excited about that if they had the choice

Axiom05's picture

"Real improvement would be better engineered recordings.
I think audiophiles would get excited about that if they had the choice"

Yes, absolutely! Garbage in, garbage out, nothing can fix that after the fact. We need to start with better engineered recordings.

foxhall's picture

We all have heard 16/44 content and LPs that blow our minds in terms of sound quality. I agree completely that better recording practices and exceptional mastering solve so many problems and elevate the art.

FransZappa's picture

well THAT would be something, wouldn't it? If MQA could do away with of all those brickwalled mastering jobs that render recordings unlistenable.

spacehound's picture

From Stereophile.

Epitomised by the image, with WAV (which is raw PCM in a 'package') and FLAC crossed out. So this time the propaganda has gone totally 'over the top'.

John, MQA is lossy.
'Lossy' means the original cannot be fully recovered, and the original will be PCM or DSD.
WAV isn't a codec at all, it's pure PCM, and FLAC recovers the original complete unadulterated pure PCM one hundred percent.

And we all know that Stuart tried to re-refine 'lossy'. And we also all know that the Stereophile group went along with his efforts to re-define 'lossy' in at least one earlier article.

And you say "MQA-encoded files tend to sound better than the PCM originals from which they are derived".
So you just admitted that the MQA files aren't "the original direct from the studio" as MQA claims, didn't you :):) :)

MQA cannot be "better" as it discards data, and also reduces the bit depth to about 17 rather the original 24 (in 48K and up files), and also introduces possibly audible aliases in the 20Hz-20KHz band.
MQA may sound different, but it cannot sound "better".

Why do you persist in saying that MQA is "better" when you can only mean that you personally prefer the sound of it's losses, bit depth reductions, and distortions?

Taking all the above, which is entirely factual, into account, I can only conclude that you and Stereophile are taking us for fools. And you must have a motive for doing that.

Jon Iverson's picture
FWIW, the image at top is meant to illustrate the statement in the story from MQA's Spencer Chrislu ("...asked if he hoped or expected that all digital music will someday be MQA-encoded, Chrislu's response was blunt: 'Well, that's the goal!'") It is certainly not Stereophile's wish that other formats be eliminated! In fact many of us wish just the opposite.
crenca's picture

Is that not just a little (I hope I don't offend but no other word fits) naive? Putting aside consumer choice, what about MQA makes you think it just wants to be one choice among many? Is this not something that you should have focused on from the beginning - these "More" aspects of MQA like its end to end aspirations, or its black box intellectual property/DRM/closed nature?

Claiming "we at Stereophile are innocent because we *wish* MQA's intent/goal does not happen and we get our sound quality tweak and consumer choice as well" has an almost child like quality to it...no offense intended. I mean, I *wish* to win the lottery, and for peace to break out in palestine, and...at this point I kind of want to pat you on the head and say "isn't that nice" :)

spacehound's picture

But I assume from your "many of us" that you are on the Stereophile staff, so "You would say that, wouldn't you?" (taking an extremely well-known quote from a UK court case that has become part of the English language here in the UK).

First, the link between the picture and Chrislu is not at all obvious, as it's YOUR picture and HIS quote, and also they at as far apart as Stereophile could make them in one article.

More importantly, I see Jim Austin's ludicrously non-critical and very uninformed 'reports' on MQA have so far got 'Atkinson introductions' where few other products get such from the official Guru.

It is not just this Atkinson piece, it is the constant referring to pro-MQA stuff as "reports" and anti-MQA stuff as "claims", the constant referring to MQA as "better", when its known and measured distortions MUST result in inferior sound quality even though a few might personally prefer it, and the refusal to ever refer to such distortions in a critical manner, if at all.

And then their is the total and utter acceptance of Stuart's farcical attempt to re-define 'lossy' when he was caught out and Stereophile's parrot-like repetition of it as if it was true.

We ALL know what's going on - Stereophile is in the 'leisure' business and it's master, 'Ten' or whatever it's called, depends on pushing new stuff and the advertising revenue gained from such pushing for it's very existence.

Like it or not, while Stereophile might be reasonable credible on other stuff, it has lost all credibility on MQA. It's MQA observations, together with those of it's junior associate, Audiostream, have become a joke.

michaelavorgna's picture

I'm not sure what I've done to deserve such high praise from you, but I thank you nonetheless.

Michael Lavorgna
Editor, AudioStream.com

spacehound's picture

At least you've got a sense of humour.

It's a hobby. The magazines and websites are therefore hobby magazines/websites.

And your site is every bit as much about music as all this nerdy 'gear' to play it on. Which is quite refreshing.

We are not killing babies here :):)

metalmorphosis's picture

The improvement you may hear comes from the time-domain correction, not the compression, they're two separate things. Some have been doing a similar correction process on non-MQA remasters and it works beautifully.

spacehound's picture

I will start at the beginning:

1) A computer file doesn't contain timing data and a 'music' file is just a computer file, the same as a print file, an image, or whatever. All a music file contains different is a part at the beginning that says "Play me at xxx bits per second". That just looks like any other data, only the player software knows what it means, the computer hardware itself doesn't.

2) So as there is no timing the timing can't be 'wrong', so this 'jitter' stuff so often worried about is nonsense at that point.

3) It eventually reaches the DAC, and if it is a USB connection it is in NRZI format so there aren't any gaps between the pulses, so the DAC cant tell where any particular pulse ends and the next one begins. The "play me at xxx..." stuff is at the beginning of course or the DAC would not know what clock speed to set.

4) It goes into the DAC buffer and is read out by the DAC clock. Which is the same clock that goes 'slice slice slice slice....' to see if the value is 'J' or 'K' (in NRZI format). It does that where it expect the middle of the pulse to be and so that it doesn't get too far out NRZI forces a value change from J to K every few pulses if there hasn't been a change already, which re-synchronises the clock with the pulse stream so the clock doesn't stray too far from the middle of each pulse.

5) And of course a modern 'asynchronous' DAC 'pulls' the USB NRZI stream from the computer with it's clock. The computer's clock doesn't play any part in this stuff.

So as everything depends on the DAC's clock I don't see how this 'time domain' correction is supposed to work.

Axiom05's picture

I believe that the time-domain correction is not jitter-related but rather due to errors from either the digital filter/DAC in playback or the digital filter/ADC used in the recording process (de-blurring).

John Atkinson's picture
Axiom05 wrote:
I believe that the time-domain correction is not jitter-related but rather due to errors from either the digital filter/DAC in playback or the digital filter/ADC used in the recording process (de-blurring).

This is correct. See my examination of the issues at www.stereophile.com/content/wadia-850-cd-player-page-2.

John Atkinson
Editor, Stereophile

Fokus's picture

.. make that 'perceived errors', as contrary to popular belief, there is no evidence that these 'errors' are audible at all in well-produced music. MQA's papers and publications carry an impressive list of references, but when you dig deeper you'll find these references hardly relevant.

AJ's picture

Please cite your ITU standards blind tests (as JA advocates) to support this claim, thanks.

Please also cite your evidence that ADCs over the last 2 decades need time "correction". Thanks again.

music or sound's picture

"If an MQA file is played without MQA decoding, the sound quality will be that of the baseband file—ie, the same as a CD—meaning that the record company need stock only a single inventory."

That was not at all my experience it sounded much worse than red book more like mp3. Actually I prefer mp3 at 320 to what I heard from MQA files not decoded.

Axiom05's picture

Yes, the same arguments can be made for HDCD. I am now stuck with a bunch of HDCD-encoded CD's that I am unable to play with decoding. This pisses me off as undecoded playback results in obvious inferior sound to that with decoding. People who supported HDCD with their dollars got "had." This is one reason why people do not trust MQA. However, HDCD was never going to be incorporated into every recording available for purchase. This is huge if you are talking about taking away a person's ability to choose between MQA and Non-MQA.

Fokus's picture

And there are free software decoders for HDCD, so you are not really stuck. When will we see a free (or paid) MQA decoder, allowing the purchaser of a file to expand it fully. You may want to do this for making it compatible with a specific system (such as one built around a digital amplifier, or digital-in active speakers), or for inspection (is it really hi-res? Do not trust the Blue LED!)

jimtavegia's picture

if we don't start with something better than how most albums are made now then we will always be getting less than the original performance. To me this is just another layer of "control over the music" that is less destructive than most other forms of "encoding".

I love tracking at 2496 as it sounds so smooth to me and more "real", but my old ears know that 24192 is better, but I just can hear it anymore...99% of the time I try. The math doesn't lie, but the converters must be pristine to hear it.

It is sad that SACD just really never caught on as the the masses could really care less about quality, so this will be fine for them and many others. I had always hoped for an affordable, portable SACD player, but it was just not to be. I have 4 home SACD players, but not enough new titles of what I would love to hear to make that matter anymore. I have to give BlueCoast credit for keeping the DSD dream alive. I do support HD Tracks, Linn Music, and eClassical often downloading 2496 files and burn then DVD-Rs.

David Harper's picture

a well recorded well mastered uncompressed redbook CD sounds much better than a typical hi res playback of a typical shitty distorted compressed poorly mastered recording. I have SACD,pure audio blu ray,and DVD A recordings and my best redbook CD's sound better than all but a few of them. The difference in SQ between 2496 and 24192 does not exist to the human ear. Anyone who says it does is experiencing the placebo effect. But then, that's what all of this is really about, the placebo effect. All we need is good recordings. No amount of playback technology can make a bad recording sound good, and in the case of a good recording we don't need any encoding/decoding at all.

jimtavegia's picture

Pop recordings are not compressed and poorly recorded? I can in my inexpensive home studios record at 24/96 and 24/192 and it sounds better than most redbook recordings that come out today, mainly because I don't try and "fix" them in the mixdown or mastering process. Of course if you NEED pitch correction what is one to do? I feel sorry for anyone who cannot hear the difference from redbook to 24/96 or better. It is most definitely better.

The engineer's job is to take the performance and let us hear what he is hearing live, but of course that never happens as too many don't know THAT is the goal, or for the artists, they may not want that and just want loud, a full sea of red and white, top to bottom in their DAW. Hardly great engineering.

Tape has its own inherent compression, high noise floor, and higher distortion as you approach 0 db, which most do to get higher than the noise floor. Once you start manipulating or "fixing" it, all bets are off.

I love Billy Joel, but his pianos are recorded so badly that it is hard to listen to. He is a great story teller and performer. I am not sure where it all when wrong, but even his stranger SACD is lacking and it was done by one of the best.

One of the real tests of high bit rate recording is on choral work. If you can't understand the lyrics sung by mass voices it could be a combination of poor continuity of the singing, enunciation, but on the high rez recordings I have I am able to pick out the words much better.
Choral work must be well tracked at redbook to make this happen and you need low self-noise mics and a quiet venue.

David Harper's picture

I said A WELL MASTERED WELL RECORDED CD. I didn't say anything about poorly mastered poorly recorded CD's.

crenca's picture

Well done. I could quibble (e.g. whether unencoded MQA through what Mr. Stuart calls our "legacy DACs" is CD quality, if MQA is the DRM equivalent of HDCD, MQA's bandwidth savings, etc.) but consumers need to give credit to Mr. Atkinson's explicit recognition that MQA is not just another audio product.

Your mention of Net Neutrality is especially apt. However in some ways if MQA (or anything like it) were to ever become the standard format in audio, that would be even more of a limit to consumer choice than our now tiered and throttled (based on application/service) internet. At least in theory, we can always choose another ISP, another Telecom, etc.

foxhall's picture

Appreciate the balanced tone of this piece.

I've always wondered if MQA is working on technology that would deliver 16/44 quality at the same file/stream sizes seen in lossy applications. It seems like the mainstream streaming industry would welcome a way to deliver conventional lossless (maybe with the timing fixes) in the very small sizes of mp3 and aac.

spacehound's picture

And the basis of ALL their original marketing.

And as they eventually realised it would fail as few MP3 etc. users have any interest whatsoever in sound quality they moved the goal posts TOTALLY, introduced meaningless gibberish such as "deblurring", pretended it was 'lossless' until caught out, etc., in the hope they could sell it in the Hifi arena.

tonykaz's picture

Philip Bloom is a Professional Cinamatographer ( and an Englishman ) who reviews Pro level Cameras.

Jan. 6th, Philip Bloom published, on YouTube, his lengthy Review of the new EV Panasonic Camera.

This man reviews Cameras with the same level of competence that Alex Dykes reviews Cars.

I might suggest that our Tyll comes the closest to these highest levels of reviewing professionalism and competency, up there the air gets pretty thin.

Cameras have always been part of my Professional Life.

Tony in Michigan

ok's picture

..give me a decent digital equalizer and I can make any recording sound “better” than the original master; I could even do it just by looking at a spectrum analyzer without ever having to actually listen; now let me also alter the "original" and you'll never know what hit you in the first place while you keep fighting another man's little "format war".

dalethorn's picture

Ain't that the truth. Thinking of one example - the Loudness Wars, and how changes in the available music can be made 100 different ways. Scary. I don't think there's any substitute for a real music lover for doing the hard work to build a collection of the best available - the best available at each moment in time when collected and possessed. I do occasionally replace an existing track or set of tracks, but I'm extremely careful with that, since a new version of any particular album may have better-sounding and worse-sounding tracks, or tracks that have various other edits I wouldn't want.

ok's picture

..there’s another way of doing things, when one finally realizes that inherently flawed recordings (and that’s a huge part of the existing inventory) tend to sound “better” when played through a less accurate, but easily configurable audio system; above a certain point of fidelity most hi-end systems tend to pick up more artifacts than music from lesser recordings. I personally have come to the conclusion that a second, more down-to-earth software-based audio rig, serves me better (and cheaper) than buying the same stuff again and again or subscribing from the outset to some "unlimited" extra-charge HD/MQA streaming service which may or may not be available next monday morning for me and my listening buddies. By the way nothing guarantees that the most recent remasters of legacy recordings are more faithful to the "original", since analog tape frequently degenerates at quicker pace than remastering technology progresses thereof. As for digital masters, there's practically nothing a "remastering" can fix; ADC clock jitter, the main problem in early digital recordings, stays with us from there to eternity.

Mike-48's picture

+1 to the earlier comment that if subjective "improvement" is the only goal, it can be achieved by any talented mastering engineer with digital EQ. I'd rather have the original release available, without any fiddling, please.

spacehound's picture

Downloads.

I buy them. So they are mine. I can therefor do anything I want with them.

Sell them, give them away, copy them and then sell the original, anything I want.

On copying, the local state run library has a copier. I can copy pages from books, sections of books, and even entire books. If I don't know how to work the copier a member of the library staff will show me.

They also have CD you can borrow, just like books. So on exactly the same basis I can therefore take a CD home and rip it, then return it.

And in the UK your contract is with the seller, nobody else.

So if I buy a download from, say, HD tracks, only HD tracks might be able to sue me if I copy it and then sell it. Not the 'manufacturer' (which is presumably the record label) as I have no contract with them.

And in UK law it is a civil case, not a criminal one. And the only people, from the above UK law, who might try to sue me are HD tracks. But they won't succeed as I own the download, not them.

As for any 'terms and conditions' I might have agreed to before they transmitted the downloads, in the UK we can't sign our rights away even if we want to, so such agreements have no legal validity.

DH's picture

At many of the download sites, I think they only have one version of an album, the highest res one. When you order a lower res version, it converts on the fly to the format you ordered.

Mike-48's picture

A big issue with MQA is that it totally screws up DSP as done by MiniDSP, Holm Acoustics, Classe, and other manufacturers. MQA can't be output to full-resolution digital, and DSP needs that to do its work best. Though audiophiles have been slow to adopt DSP, it seems to me that it is the future of good sound in many cases. MQA puts a knife into its back.

spacehound's picture

It's because DSP messes up the sound.

We hifi enthusiasts like and are willing to pay for accuracy, which is the very definition of high fidelity. And high quality. Who wants
an 'inaccurate' Mercedes?

And the expensive equipment we use mostly has a
reasonably flat frequency response, and that includes the speakers, so it complies with the definition of hifi and high quality.

"Good" sound. No, it isn't "good", you've just used DSP to mess with it so it is to your personal preference, which is different from everyone else's. What is more, your messing with it will impose the same inaccurate 'tone' on every piece of music you listen to.

It's the same nonsense with DSP used as 'room correction'. If the orchestra, band, or singer was playing in your room you wouldn't be able to 'correct' their playing so you shouldn't do it with a recording either. If you do you will be completely wasting the money you spent on buying your source, amp, and speakers which you personally chose because you liked their accurate sound.

If you don't like what you hear just play a different recording of the same music. You can do that easily with streaming, where MQA has its main application

dalethorn's picture

"If the orchestra, band, or singer was playing in your room...."

Yeah, IF your recording and your playback gear are "just like" that live sound in your room, then fine. But it ain't that way.

spacehound's picture

However.

We all know that live and recorded music never sound the same. And live music on a good FM tuner using a good FM broadcast (BBC Radio 3 is a fine example) don't sound the same as recorded music either (though on Radio 3 you can instantly identify 'live', and I don't mean by crowd noise, also with competent classical orchestras, not useless pop groups who require a lot of work in the studio to be 'sellable' at all).

And as I said, we can't ever say that DSP is 'better' because it alters the original. We merely mean "I prefer it". And not being in the studio when it was recorded we don't have any reference for "As the studio or label intended".

So we cannot detect goodness or badness (sound quality). But what we CAN do is buy 'flat frequency response' equipment and know we are getting as good as it gets within whatever price range we are prepared to buy.

Why spoil that with DSP?

And as for 'live' I used to live in central London within a few minutes walk of Hyde Park and the Albert Hall. (Curzon Street if you know central London.)
There are often brass bands on bandstands in Hyde Park and occasionally even full classical orchestras playing in the open. The sound in the park is totally different from the same people in the Albert Hall. And they would sound different in my room.

So we don't have a reference for ANY of it. But messing about with the sound using DSP can only make it less real that it was or is ("is" being a Radio 3 live FM broadcast).

dalethorn's picture

"Why spoil it with DSP?"

Because, if needed (if) and done right (if) it can make a good system with very minor flaws better - the flaws become less noticeable or not noticeable at all.

There's a very simple rule for this that's as old as music itself - try it - if it isn't better just drop it or try something else.

BTW - this is not a defense of MQA - I welcome the efforts to give it a good long trial, but like a few others, I worry about recoding hundreds of masters with the stuff and then cutting off access to the originals. I think there's *lots* of reason to not only keep the originals 100 percent safe, but to make sure we have continued access to those originals. Streaming? Don't know, don't care.

spacehound's picture

Unless it's actually broken most 'good' equipment doesn't have any flaws. It's just that they sound different from each other.

Which ideally they shouldn't of course as there is only one 'perfect' measurement of 'reality' (and music reproduction is, in effect, a measurement).

And as I said, as we weren't in the studio when the music was laid down we don't know what the reality was.

So trying to 'correct' it with DSP is not only impossible, it's idiotic. Trying to 'correct' something when you don't know what it should be is the action of a madman.

I also see you have fallen into the "better" trap. You don't mean that, you mean 'more to your preference'.

MQA? In many magazines and commercial websites it's a good long 'positively commenting on the existence of MQA as often as possible' and not a trial at all, though disguised as one. 'Hobb'y magazines such as this one largely survive by reporting on bandwagons. Nobody would buy them and so read the adverts (which is what the magazines are for) otherwise.

It has been proven over and over again by technical people with far more knowledge than any writers in commercial magazines that its output is very different from its input.
Some of the input is removed, so it's lossy. if the input sample rate is greater than 96K it reduces it to 96K and replaces it with a 'fake' that isn't real but will show up on your DAC as a false 192, 384, or whatever in the hope of convincing suckers. It splatters added noise all across the main audible range.

I don't think you need worry about the loss of the original. Nobody is actually recording in MQA as the MQA people falsely claimed would happen.They are just inputting the master though an MQA process that degrades it, they are not destroying the original. And you can still buy the originals as downloads in most cases. This will continue. Not all of it is 'high-res' of course as the originals weren't in the first place.

MQA won't survive, no more than SACD did, which exists only as a small niche within a small niche. That was introduced with a great fanfare, boxes to play it were made, and so on. Nobody except a few cranks cares about SACD now.
99% of music buyers/listeners don't care about quality and outside specialist 'hobby' magazines like this one nobody has even heard of MQA.

dalethorn's picture

It's idiotic to correct obvious resonance effects and other flaws in hi-fi gear? That's the most arrogant elitist bullshit I've ever read on this site.

spacehound's picture

That most things posted here are just opinions, except 'facts' such as "today is Thursday"

And opinions, not being facts, are not subject to 'right or wrong' tests, so they are allowed to differ.

dalethorn's picture

"I've noticed.......that you don't like being disagreed with "

Amazing how that survived your edit, eh? So you don't like the message so you attack the messenger. Bottom line: *You* don't have any right to tell people on a budget that they cannot improve their sound by various customizations - instead they have to spend big. That is elitist crap.

Edit: I prefer "I agree or disagree with what you said" rather than "I agree or disagree with you".

Fokus's picture

DSP does not (have to) screw up the sound. The one in my system doesn't. And DSP used in digital-active speakers as crossover and as driver linearisation also does not screw up the sound, it is part of the design. And yet, not compatible with MQA.
The example with the musicians in the room is invalid on so many counts that I can only advise you to read a few good books on acoustics, speaker design, and auditory perception.

Mike-48's picture

Thanks for sharing your opinion, with which I disagree. You can read numerous reviews of DSP systems in which respected professional reviewers report startling improvements. In my experience, improvements far outweigh any issues in most domestic rooms. If we all had ideal rooms, DSP would be far less valuable, though I think digital crossovers are the way to go.

But more to the point, the type of DSP I'm talking about is under control of the user, who can turn it off if he wants! The opposite is true of MQA, and a criticism that has been raised is that its DSP (yes, it is a form of DSP) does impose the same sound signature on everything.

spacehound's picture

On what? They weren't in the studio when the original was laid down so their so-called "improvement" means "to my personal preference" and nothing more. They don't have any reference for their "improvement".
And always remember magazines and commercial websites depend on 'keeping the pot boiling' with 'new things' for their very existence, and if they report negatively on them they won't be given free access to much new stuff.

"under control of the user". You mean you are going to switch it on or off or alter it for every piece of music you play? Very few people will do that which is precisely why it has not been widely taken up. We listen to music, we don't fiddle with switches and knobs, not even software ones.

I use the JRiver player software, which has very powerful DSP options, including Sox, and I have tried all of them. I can make some files sound more to my personal liking but I am always aware that I am deliberately selecting a 'falseness' so I have given up on them.

However, I agree on your MQA 'imposing' its signature on everything. I have been listening to it a lot lately and am not impressed. The differences are small but always a degradation over a 'straight' high-res file. MQA files can have a false 'livelyness' probably due to the 'aliasing' that MQA splatters over everything it touches.

Mike-48's picture

(Sorry, I can't figure out how to quote in this forum. This is in response to spacehound.)

The main thing people use DSP for in audio is smoothing bass frequency response in typical domestic rooms. This in turn reduces boom and improves articulation. It is not done with SOX or any other resampler, but most commonly with equalization.

To do it right, you need to take measurements, iteratively adjust the DSP with them in mind, and at the end tune it by ear. There are automatic systems; they also need tuning by ear, or they can make things worse.

You don't need to have been at the recording venues to judge when a sample of recordings sounds better on average. It's just like buying a new piece of gear. You look for more natural sound and the thing you hope to achieve is a reduction of any sonic signature (in this case, the one imposed by the room). In other words, what you accuse DSP of doing is precisely what it can be used to fight -- imposition of some constant sonic signature that makes everything sound the same. Of course, physical bass traps and such also can help, as can careful positioning of speakers and listener.

To date, there is no perfect DSP system, but to my ears, reducing bass boom and reverb time usually give a net improvement in musical enjoyment. I believe that, just as other audio gear has improved, the ability to do transparent DSP will improve over time -- unless the adoption of some DRM system like MQA prevents it.

AJ's picture
Quote:

If an MQA file is played without MQA decoding, the sound quality will be that of the baseband file—ie, the same as a CD

John, can you cite the Telecommunications Union document ITU-R BS1116-3 standard listening tests, as you mentioned here https://www.stereophile.com/content/simple-everything-appears-simple
for the above claim, thanks.
If I recall, when Bruno Putzeys questioned Bob Stuart about such listening tests for MQA, he stated there were none.

AJ's picture
Quote:

MQA is backward-compatible with every DAC out there, and that MQA file, even undecoded, will sound better than CD due to the de-blurring we do.

https://www.stereophile.com/content/spencer-chrislu-master-quality-authe...

Mr Chrislu/MQA comment appears to contradict yours. He not only claims they don't sound the same, but rather the undecoded unauthentic aliasing file will sound better! I missed that article previously, thanks for linking it!

spacehound's picture

1) "the same as the baseband file-ie, the same as a CD"

While that is 'true' as far as it goes, it omits that it is only true for a CD.

A CD is 16 bit. MQA is "about 17" - Bob Stuart.

Files above CD resolution are usually 24 bit. MQA truncates everything to "about 17". So it is lossy over the entire audio band (20Hz-20KHz) for all 24-bit files and everything above 20KHZ as well.

Currently we mostly listen to MQA via streaming, where 44.1/16 is usually the 'hifi' standard, so under those restricted circumstances MQA is not lossy up to a nominal 22.05 KHz if you exclude the impact of the filters.
But as soon as MQA 'expands' it to 48K or higher it becomes lossy. Which renders MQA downloads totally pointless. (And I doubt 44.1/16 will be the hifi streaming standard 'forever' so MQA will become pointless, and lossy, from then on.)

It is lossy in other ways too but I won't go into that.

The quote doesn't mention bit rate but I will.
MQA claims up to 384 bit rate.
I don't have any 384 files but I do have 192 ones and my MQA-capable DAC shows "192". A non-MQA DAC should show it too if you tell the Tidal (and some others) software to do the 'unfold'.

But it isn't real. Everything above 96K is merely upsampled.

2) "Due to the deblurring we do".
They say it corrects "temporal blur".
Unfortunately for them no such term exists in the relevant scientific and engineering worlds. So the MQA people made it up.

Maybe they mean 'dispersion'. This term does exist.
Again unfortunately for them, dispersion CANNOT be corrected under any circumstances whatsoever.

Fokus's picture

MQA's 'temporal blur' is the ringing seen before and after the main lobe of a filter's impulse response. The ringing occupies a small frequency band at the filter's transition frequency. It does not touch any other part of the signal's spectrum. Such ringing is audible when the filter transition is in the audible band (try it with 3kHz) and when the filter is steep enough to have the ringing extend beyond the time constant of the cochlea in the same band. This ringing is not audible, cannot be audible, when the transition frequency is high. Even for CD rate this is the case for most adults. The ringing also does not even exist when there is little or no spectral content in the signal at the filter's transition.

The MQA reasoning is entirely based on looking at impulse plots, conflated with irrelevant and/or misunderstood auditory research.

spacehound's picture

And ringing doesn't occur at all if you limit the analog input so it obeys Nysquist/Shannon and put your filters outside the audible band, as you in effect say with your "spectral band" observation.

Three things to note:
JA initially commented that impulse testing was 'illegal'.
When he realised that it would mean that Jim Austin's testing with impulses would be rendered invalid by this comment he 'softened' his comment.

About a year ago Stuart called his process "de-ringing". He later changed it to "deblurring", presumably because he thought it sounded more 'technical'.

And Stuart refers to "the second unfold" when it's just upsampling to make your MQA DAC falsely display "192", "384", whatever.

In the end it won't matter, it will vanish just like Stuart's DVD-Audio protocol did. The magazines were all fanboys of that when it appeared.

-Rudy-'s picture

Another month, another opinion piece shoving mqa down our throats. This is starting to sound like Trump's Twitter account around here. How predictable.

Don't know when to quit, do you?

Yeah, we get it. Your opinion is that it sounds good...and apparently, woe to those who don't agree. Opinions aren't worth the pixels they are typed on.

Meanwhile, consumers find two things suspicious:

1) Why is the audiophile press cramming this down our throats, so hard and so often?

2) Why does bob stuart or anyone from mqa go cowering any time their feet are held to the fire? (*cough* RMAF *cough*)

Those of us in consumerland smell a rat...well, a few of them. Maybe if the high-end press would back off for a change instead of being so obsessed and annoying about it, consumers might actually want to see what it is all about.

David Harper's picture

My system at home has been all digital for twenty years, but the best sound I ever heard in my life on a home system was a direct-to-disc all analog vinyl called "Dave Grusin discovered again". It was new about 35 years ago. The only time in my life that I thought the musicians were playing in my listening room, for real. You could hear the physicality of the wood that the piano was built with. I no longer have the LP. Don't know what happened to it, I guess I just lost it along the way somewhere in there with two divorces and everything else. Maybe pure perfect analog really can sound superior to digital. All I know is I've never heard anything like it since.

labjr's picture
Quote:

"The record company will no longer be selling a clone of their hi-rez master. Instead, they are selling something that might well sound identical to the master, or even better than the master, but doesn't allow the master to be re-created."

You don't actually believe this, do you??

David Harper's picture

wow, so MQA is a time machine that can go back to the original recording session and improve THE ORIGINAL MASTERING OF THE RECORDING!!!!!!!

X