Inside MQA

Jim Austin briefly discusses MQA in his review of the Explorer2 in this issue, but a more complete description of MQA can be found in an article posted on Stereophile's website at the end of 2014.

MQA involves two fundamental concepts, discussed in a paper presented to the Audio Engineering Society in October 2014 (footnote 1), the first responsible for a potential improvement in sound quality, the second responsible for a large reduction in the bandwidth required for storage and streaming of high-resolution files:

• MQA can compensate for the time-domain errors of both the original A/D converters used to make a recording and the D/A converter used to play it back. This results in the complete recording/playback chain having an impulse response equivalent to a few feet of air, and temporal resolution of the same form and order as that of the temporal sensitivity of the ear-brain.

• With files sampled at 2x, 4x, or 8x the baseband rate of 44.1 or 48kHz, the information in the one, two, or three ultrasonic octaves can be encoded and packed below the music's baseband noise floor in a 24-bit container. (Bob Stuart says the buried data, which are encoded to resemble noise and are uncorrelated with the music, lie beneath the analog signal's noisefloor.) This "audio origami" results in a much smaller file size than the hi-rez PCM equivalent, yet when the file is unfolded, the resolution and bandwidth of the original file are preserved. If an MQA file is played without MQA decoding or time-domain correction, 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 duplicate of their hi-rez master, with all the implications for piracy that that implies. Instead, they are selling something that might well sound identical to the master, but doesn't allow the master to be re-created.

The only way of testing MQA's time-domain correction is through listening, so I sent Bob Stuart the 24/88.2 masters of some of my recordings, for him to produce MQA versions. I will be reporting on the sound-quality differences I hear between the original and MQA files in a future issue. But in the meantime, I examined the files' properties.

The reduction in file size is significant. The 24/88.2 WAV master file for my recording of the Portland State Chamber Choir's 2014 performance of "Amazing Grace" is 169.5MB. The MQA-encoded FLAC version is just 51.5MB: 30% of the original size. For reference, the 24/88.2 Apple Lossless version is 90.7MB, or 53.3% of the original, and the version on the CD (footnote 2), limited to 16-bit/44.1kHz, is 55.7MB, or 33%. MQA, therefore, gives the greatest reduction in file size and thus necessary streaming bandwidth, yet potentially possesses the same bandwidth and resolution as the hi-rez original.

Does it preserve those hi-rez qualities? To answer that question, I played both the original file and the MQA file with Audirvana running on a Mac mini and with the Explorer2 selected as the playback device. (Although Audirvana reported the MQA file as a 24-bit FLAC sampled at 44.1kHz, the Explorer2's second white LED illuminated to indicate 88.2kHz data, and its first LED turned green to indicate MQA encoding; with actual artist/producer authentication, it would turn blue.) I recorded the analog line output of the Explorer2 with a 24-bit Ayre Acoustics QA-9 A/D converter so I could compare the frequency-domain properties of the decoded files with those of the originals using Adobe Audition's FFT engine. I also recorded the analog output of the undecoded MQA files played back with our review sample of the original Explorer.

Fig.1 shows a spectral analysis of the entire "Amazing Grace" 24/88.2 master file with the levels plotted in white (highest) through yellow, red, magenta, and blue. It is difficult to see at the scale this graph is printed, but the spectrum does reveal content extending to 27kHz or so at the musical climaxes. This content would, of course, be eliminated when the file was converted to 44.1kHz for release on CD.

Fig.1 "Amazing Grace," Portland State Chamber Choir, 24/88.2 WAV file, spectral analysis of complete track (left top, right bottom).

Fig.2 shows a more conventional spectral plot for the track as a whole. The spectra of the original WAV file (blue and magenta traces) are exactly overlaid by the spectra of the MQA FLAC file decoded by the Meridian Explorer2 (green and gray traces), indicating that the MQA encode/decode process does accurately preserve the frequency-domain characteristics of the original hi-rez file up to the Nyquist frequency of 44.1kHz, despite the baseband lossless-encoded file being less than one-third its size.

Fig.2. "Amazing Grace," complete track, spectrum of original WAV file played on Meridian Explorer2 (left channel blue, right magenta); MQA FLAC file decoded by the Explorer2 (left green, right gray); MQA FLAC file played without decoding by the original Explorer (left cyan, right, red) (10dB/vertical div.).

The cyan and red traces in fig.2 were taken with the original Meridian Explorer being fed with the MQA FLAC file. It can't unpack the MQA data, of course, so plays the file as a 24/44.1 file. The traces overlay the WAV and MQA spectra up to 12kHz, above which the noise floor starts to rise, peaking between 20 and 22kHz before being rolled off by the Explorer's reconstruction filter. There is another small peak evident just below 44.1kHz. Fig.3 compares a spectral analysis of the hi-rez WAV file's noise floor (footnote 3) played on the Explorer2 with that of the undecoded MQA file played on the original Explorer. Again, the undecoded noise floor starts to rise in the treble, this time covering a broader bandwidth than in fig.2, but again peaking between 20 and 22kHz.

Fig.3 "Amazing Grace," noise floor, spectrum of original WAV file played by the Explorer2 (left channel blue, right red) and of MQA FLAC file played without decoding by the original Explorer (left green, right, gray) (10dB/vertical div.).

This behavior with undecoded MQA data is very similar to what happens when a 24-bit file is decimated to 16 bits using the POW-R narrowband Nyquist dithering scheme. Provided that MQA's encoding of the super-baseband information results in data that resemble random noise components that are not correlated with the musical information, therefore, playback of a 24-bit MQA-encoded file on a non-MQA D/A converter should be benign.

This is very much a preliminary report, but it suggests that the "audio origami" aspect of MQA does work as advertised, folding super-baseband content under the recording's analog noise floor to achieve a dramatic reduction in file size/streaming bandwidth. There is no such thing as a free lunch, of course, and the tradeoff is that a decoded MQA recording is basically limited to an effective resolution of less than 24 bits. But D/A converters offering 19 bits' worth of resolution or more are rare, in my experience, and what might be at most a loss of theoretical resolution may well be outweighed by MQA's time-domain correction. And that brings us back to listening.—John Atkinson



Footnote 1: See Stuart, J. Robert, and Craven, Peter, "A Hierarchical Approach to Archiving and Distribution." See also the free download "Journal of the Audio Engineering Society.

Footnote 2: Into Unknown Worlds.

Footnote 3: I used "Amazing Grace" for this report as it is one of the quietest recordings I have engineered. Its noise floor has basically a red spectrum, dominated by environmental noise at low frequencies and tilted down at high frequencies.

COMMENTS
latinaudio's picture

Ok. Here is a fine article by Mr. Atkinson regarding several aspects of the MQA. Now, I'd like all the defenders of the open-source model to resume their arguments on why they believe MQA is a licensing model, but with the same depth of knowledge shown in this article!!!

Archimago's picture

Can you not see the inadequacies of the tests?

Do you actually think there has been depth added that we didn't know before?

eriks's picture

I'm not sure that's even the argument. I would argue instead that if MQA is actually whorthwhile we should see open-source equivalents pop up soon. I mean, sounds very cool as a compression technique, but why?

The open source model has allowed many innovative companies to flourish, by the way. Now if it would only kill off Windows 10....

I don't see MQA as being nearly as relevant as say my 1984 pair of 3D glasses that came in TV Guide. In most developed countries with good broad band (i.e. outside of the US) we can easily stream 128k/24 audio or better if available. I listen to a Canadian jazz station that streams at 96/24. Disk and network bandwidth are cheap. The only benefit I can see is on the vendor side like HDCP for HDMI.

One thing though, this whole thing about needing to know the decoding chip (according to DAR) ...smells like burning snake oil to me. So long as I have my Phyl Colins CDs to fall back on after the Zombie Apocalypse though, you all go right ahead and MQA everything. ;-)

Best,

Erik

synth's picture

Correct me if I'm wrong but MQA decoding is happening in the hardware domain, i.e. the audio playback software such as iTunes, Fidelio etc don't do any of the decoding? Or perhaps both need to be MQA compatible? This is where MQA is confusing me as with most formats as long as the software will output data at a rate your DAC can decode (D/A) then you're golden, but MQA seems to be tied to specific hardware requirements.

Dr.Kamiya's picture

From how it was explained I think MQA can be decoded at either the software or the hardware level.

If the player can't decode MQA, all it has to do is output a bit-perfect 16/44.1 of the encoded file, and then the MQA-capable DAC can decode the stream.

If the player can decode MQA, then it can output as analog, or decode MQA and expand the stream to 24/88.2.

synth's picture

So essentially the hardware (DAC) has to be able to decode the MQA data regardless of the player software.

Dr.Kamiya's picture

I don't think there are any MQA capable software players right now, but if there are then the hardware (DAC) shouldn't have to.

MQA is basically about encoding high res (ie. 24/88.2) into 16/44.1. If a software player can decode MQA, then it can simply output an already decoded 24/88.2 stream and any ordinary DAC can take that stream for conversion to analog. There wouldn't be any need to decode MQA again because the stream has already been expanded/restored to high-res.

John Atkinson's picture
Dr.Kamiya wrote:
MQA is basically about encoding high res (ie. 24/88.2) into 16/44.1

Small correction: 24/44.1 or 24/48.

John Atkinson
Editor, Stereophile

Dr.Kamiya's picture

bad reply

StCugat's picture

From my understanding of this article I am concerned that, unless you have an MQA decoder, you are getting a noisier file. It may be smaller, but it is noisier in the high frequencies. And it is only marginally smaller than the (cleaner) CD equivalent you can get without MQA.

My understanding seems at odds with the noise being completely below the analog signal noisefloor - as per your report of BS in your brackets above. I thought you simply re-recorded the analog output of the decoded and non decoded versions.

Also the point about protecting the original master seems easy to circumvent by simply using an MQA decoder, which should leave you with the master? Also, what are the implications for the authentication detection that is added.

I will be very interested to hear your impressions of the time smearing improvements. I presume an all analog recording and playback system has all of those benefits.

dce22's picture

At 44.1khz sampling rate
Time-domain error is 0%
Temporal resolution is infinite

only difference between MQA and non MQA playback is frequency response aka bandwidth size, if you think you can hear above 20k than go for it.

This timing problems are invented to sell you licensed distribution of music and fill there pockets nothing new.

Listen the MQA file on a DAC with good oversampling DSP like Chord, Mola Mola or MSB and MQA decoded file (when software decoding is avaiable) they will sound identical.

HammerSandwich's picture
Quote:

Provided that MQA's encoding of the super-baseband information results in data that resemble random noise components that are not correlated with the musical information, therefore, playback of a 24-bit MQA-encoded file on a non-MQA D/A converter should be benign.

JA, please explore this a bit more in your next article. The frequency-response variations between red & green in Figure 2 appear likely audible. I find it hard to believe that the different DACs caused these 8-10dB differences. Is there some other mechanism I'm missing?

dce22's picture

Incorrect oversampling inside Meridian DAC high sample rate decoded file moves that peak above the music.

John Atkinson's picture
HammerSandwich wrote:
The frequency-response variations between red & green in Figure 2 appear likely audible.

I don't think so, given the very low level and the fact that the difference lies above 18kHz, where the ear is extremely insensitive, even when the listener has good HF extension. As I wrote, the spectrum of the undecoded MQA file resembles that of a +16-bit recording reduced to the CD's 16 bits with a Nyquist dither scheme such as POW-R, and no-one has reported ever hearing such narrowband, high-frequency dither.

John Atkinson
Editor, Stereophile

HammerSandwich's picture

In Figure 2, look at the 1-2kHz region as well as the smaller differences across the 100-200Hz octave. (I realize you were talking about the HF noise, so apologies if I quoted out of context.) I also note that green & red are different channels. Oops! Still, red & gray show significant differences.

What caused these changes?

dce22's picture

It's not common during this kind of tests to have two DAC's playing in unison probably John had just one dac to test and that is a snapshot in different time in the music track.

John Atkinson's picture
HammerSandwich wrote:
In Figure 2, look at the 1-2kHz region as well as the smaller differences across the 100-200Hz octave. . . What caused these changes?

These lower-frequency differences are almost certainly due to small differences in the peak levels of the files - remember that the graphs were taken with the analog outputs of two different DACs, the outputs of which were not synchronous with respect to either the re-digitization or the FFT windowing.

The primary point I was making was that the spectra of the original WAV file and the decoded MQA version, both examined at the output of one DAC (the Meridian Explorer2) overlay exactly up to the 44.1kHz Nyquist frequency. The 2Fs data that are embedded in the baseband lie well below the recording's analog noise floor.

John Atkinson
Editor, Stereophile

HammerSandwich's picture
Quote:

...the outputs of which were not synchronous with respect to either the re-digitization or the FFT windowing...

Excellent point that points to measurement artifacts. Thank you for the explanation.

PeterMrozik's picture

I get a bit lost on this point:

"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 duplicate of their hi-rez master, with all the implications for piracy that that implies. Instead, they are selling something that might well sound identical to the master, but doesn't allow the master to be re-created."

So, if I own a MQA encoded file/disc/whatever and the appropriate MQA compliant DAC etc., every time I play the recording it will play back in a way that sounds "identical to the master". If someone (surely not me) should elect to "share" the file on Pirate Bay and someone with the appropriate mqa-enabled equipment downloads the file, they can listen to that recording in the same "identical to the master" mode that I just did.

So how does this deter piracy in any way?

BTW, I'm very much looking forward to hearing some of my favorite recordings in this new approach, just looking for some better understanding of the nuance. Thank you!

synth's picture

Essentially the hardware itself is the copy protection, at least thats the way I see it.

My understanding from the above is that unless you have an MQA compatible DAC all you will hear from the file is a standard CD quality recording, not the master that MQA playback can reproduce. And this is where I get confused, essentially there is no way to buy an MQA recording and get the sonic benefit without purchasing a new DAC that has MQA compatibility. Hence the pirating of the file being pointless to the user without the hardware.

PeterMrozik's picture

Thanks - I modified my question to clarify that I am assuming the downloader ("pirate") of the media would have the appropriate MQA-enabled equipment in order to get the benefit of the MQA process. They're only $299 for a Meridian Explorer2, and that's just the first quasi-mass market unit available, not a very high bar for the equipment requirement.

returnstackerror's picture

In the old days of LP's, you couldnt separate the music from the media.

Then we got compact cassette which did allow this, but at reduced quality

Then we got CD, which after the PC revolution allowed us to rip them to files, thus removing the music from the media and making it replayable at the same resolution (or better via DSP) on any device we liked.

This lead to music being delivered as LPCM (native or in flac) files or more recently DSD, divorced from a physical media, which like ripping allowed us to tinker and tailor our sound as much as we wanted.

MQA, if universally adopted by the music companies/streaming services, takes us all the way back to the LP days where the only data stream we could tweak were post playback (ie analogue outputs) and the hobbyist playback tweaking that LPCM/DSD files gave us are gone.

Thats my issue. We can all bitch about the MQA sound quality improvements or not , but we are still bitching about LP v CD or LPCM v DSD etc, so that isnt my concern.

My concern is CD's/ripping/file downloads opened up digital tweaking for those that wanted to and MQA stop thats dead in its tracks

michaelavorgna's picture

...anything in hi-fi that raises more questions as more answers are given. Just an observation, not a judgement.

returnstackerror's picture

Being an old bugger, I was around when CD came along (which was the last big change)... but the differences between that and MQA are:

1- CD was a big industry play (Sony/Phillips)
2- the rollout was relatively quick and was open (ie anyone could read the standard)
3- The motives were honorable (putting aside commercial reasons) being better sound, more compact packaging, more durable media
4- being such a sea change, there was nothing to compare it to
5- no DRM
6- an open standard

MQA has been
- rolling out for years,
- details drip feed over that time
- partnerships drip feed,
- real technical details in some peoples opinions shrouded in fluffy marketing speak
- the potential use for DRM
- not an open standard
- existing format that deliver the same “stuff”

Hence the angst

I have no issue with MQA if it is placed at the same level as PCM/DSD.. which is I as a consumer can pick what flavour I want.

If it becomes the end-to-end solution and/or PCM/DSD are "forks" at some point in the mastering , based on a MQA feed OR Tidal for example MQA all there stuff so everything streamed is MQA, then thats a problem as my choice is gone.

michaelavorgna's picture

Are you suggesting that "Perfect Sound Forever" is somehow less fluffy? As far as "honorable" goes, I'd beg to differ. Profit was the main motivation. CDs have proven to be terribly inefficient and environmentally unfriendly storage media.

When CD rolled out, there was no Internet. Thus the lack of, um, discussion related to the false claims.

The "angst" related to early CD and CD players was the way they sounded.

returnstackerror's picture

I am not wanting to go down any rat holes on this... my points are not if CD was good or bad, its just my view on Mikes (actually your) comment and the differences between the product rollouts.

Also crappy CD sound wasnt the specific fault of 44.1/16... we all know the early recording chain sucked.

But what you bring up is valid at a higher level.

Last time around we didnt really have an option (the industry decided) and it was relatively quick and painless (in the sense that LPs stopped being pressed and CD was all there was)

Now we do have options... we already have digital formats that people love and many of us audiophiles have a reasonable knowledge of the digital music reproduction chain.

Now our cheese is potentially being moved again..without our input.

michaelavorgna's picture

I would suggest the digital filters in early DACs sucked. 44.1/16 can sound simply marvelous.

The only reason for any MQA-related strife is in thinking that you are missing out on something that sounds *better*. Sure, we can dream about a world where MQA-encoded files are *the only option*, but reality intrudes upon such scenarios.

I have concerns about MQA, one of them being the fact that the more answers we get, the more questions we have.

One more note on CDs: we did have a comparison--the LP. And CD playback, in comparison, sucked ;-)

returnstackerror's picture

My issue is not on the quality side... its the worst case of a complete hijacking of the recording chain by one company (all be it, at this point, a well respected one)

Their plan could be world domination and it could happen but consumers have gotten used to choice (due to open source and I include PCM in that).

Audiophiles are also a bit jaundice by the usual cast of industry media pundits jumping on the MQA bandwagon (with the usual hyperbole that they use for other products of potentially useless efficacy) and assisting with the MQA push.

MQA is clearly a commercial play and the worry is how far that will go and how locked in/blocked out us consumers will be.

michaelavorgna's picture

And I'm enjoying our exchange.

As a "media pundit", although I don't think of myself as such, the first question regarding MQA that needs to be answered is, does it sound better? A thorny question indeed since any review of MQA is necessarily limited to the DAC being used and the MQA-encoded music provided for review. At best, under these constraints, the most we can speak to is, hypothetically, "Wow these tracks sound great through this DAC."

returnstackerror's picture

The most we can speak to is, hypothetically, "Wow these tracks sound great through this DAC.

Which is why I am not going down that line of discussion because there are too many variables and some people have postulated (after being present at such events) that potentially there is something "smelly" about the official demo's.

We can play "wait and see" on the quality side and as we know technology gets better over time but the commercial aspects are rolling ahead without the quality debate having being settled (almost a fait accompli).

Everybody gotta make money and if MQA is legit then it can stand alongside vinyl, 320k MP3, CD, Hires PCM, DSD and streaming as options we have.

Meridian (yeah I know MQA is now spun off) have some history in becoming embedded in an industry rollout of a new format (DVD/Bluray/MLP) so they have the contacts to do this again

Therefore we dont want MQA subsuming the mastering / delivery (either in part or in whole) of CD, Hires PCM, DSD and streaming… all I want (and I think others do) is a level playing field.

michaelavorgna's picture

...right?

returnstackerror's picture

thats why interweb discussion are a pain... stuff like humour and sarcasm get lost in translation.

crenca's picture

...that it was in jest and is typical blog speak. I find it strange Michael even questioned it given that he manages a blog.

I am with you returnstackerror, the "first question" is not "what it sounds like" but rather the implications it has as a digital format. By the way, MQA is already DRM in it's current incarnation (v 1.0) as it "manages" what the end user hears based on rather his equipment is "authenticated" (i.e. is MQA hardware). What is in store in version 1.1 (required "upgrade") or later is anyone's guess, but we are all quite aware of the desire on the part of "the industry" (i.e. artists, labels, etc.) for any DRM, even "weak" (i.e. no copy restrictions, MQA phoning home, etc.) DRM that MQA 1.0 offers...

michaelavorgna's picture

...writing for a blog, you'd realize that comments like "evil web" are not uncommon and are not always meant as a joke. I was simply checking.

Based on some of your comments, Crenca, I'd think you'd understand this perfectly well ;-)

music or sound's picture

The file size of your "Amazing Grace" recording is what MQA claims: slightly smaller than 16/44.1 and I could see that as a viable size for streaming via internet. But looking at http://www.2l.no/hires/ the MQA decoded files are approximately 2.5x larger than 16/44.1, independent of the original source resolution. Why this discrepancy?

philr's picture

This was answered here along with many other questions about MQA.

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

David Harper's picture

Don't laugh at me now, but would MQA eliminate dynamic range compression?
Also, would MQA have the potential to be superior to the best analog recording/playback?

philr's picture

Sadly I doubt it. MQA will not change the reasons that dynamic range compression is used, which is largely to make things sound louder and thus "better" over low quality equipment :-)

philr's picture

I think so. The theory behind MQA would indicate that it should be at least as good if not better. Certainly I think MQA should be able to approach the best analog systems at a much lower price and with greater convenience. Time will tell.

dce22's picture

MQA will sound the same as any digital format.
In high end dacs it will make no difference in low end dac's it will move oversampling distortion above 30khz in theory it can improve cheap dac's,

In theory it will sound better than CD's but worse than high-res files

It will sound nothing like tape or vinyl.

PeterV's picture

Hi John,

I am listening to MQA now for 2 weeks on a NAD C390DD - Bluesound MDC module combo and I am positively impressed by the performance. At the moment I can only enjoy the recent 2L albums and other albums which can be purchased at www.highreasaudio.com (currently about 89 albums available)

Since the MQA algorithm is capable to compensate for pre- and post- ringing effect of the Analog to Digital chain of both recent as well as older (digital) recordings and masters and also does the same correction for the DAC which is MQA 'certified' I was wondering the following:

Will a MQA 'certified' (upgraded) DAC also benefit a LOT when playing normal Non-MQA coded music files..? It would be a logical choice and huge benefit for MQA certified DAC owners!

It is clear to me that a Master Authenticated MQA file will sound even much better, but if the MQA-DAC will work in some kind of a 'neutral' position where standard .flac files and .wav files will be processed in the usual way, but WITH intrinsic advantage of the MQA algorithm compensating for the DAC ringing errors, it will be already a large benefit.

Since you are frequently performing Impulse Response measurements with DAC's during your reviews, the influence of MQA must be measurable and visible as well.. So I am much forward to such a test and review!

Thanks in advance - Peter

X