MQA thoughts

February 14, 2016
 by Paul McGowan

We’ve recently spent time in Music Room One auditioning MQA through a small test board supplied to us for testing by MQA. Our results have been unexpected.

MQA stands for Master Quality Authenticated and is the brainchild of Bob Stuart, founder of Meridian. I’ve sat on a number of panel with Bob and find him to be extremely kind and knowledgable. A real asset to high end audio.

You’ve read about MQA, and who hasn’t? I’ve attended two demonstrations of MQA, and both times was pleasantly impressed with what I heard. Blown away, no. Impressed, yes. But those demonstrations were closely scripted and performed on equipment I am unfamiliar with at a tradeshow. Other reporters have been gobsmacked by how good the differences are when attending these demonstrations. That’s not been my experience, but then I never attended the same demos they have reported upon.

Much has been written about MQA so I won’t waste your time repeating descriptions of it. What is most interesting to me is its claim that high resolution audio can be streamed with little more bandwidth than a CD – thus making services like TIDAL capable of upping their game to send us high resolution audio without choking the pipes coming into our homes.

The small sample board we were sent has a cheesy little DAC on it with a pair of RCA connectors to feed a preamp. Input is through USB. Not my first choice to audition a new technology, but it’s what they supplied. In later experiments we will try getting an actual digital signal out and into a DirectStream. This experiment was a simple one: digital in through USB, audio out and into a preamp.

We downloaded a track of music from the 2L website, which always has excellent selections from which to choose from. The selection was Mozart and a nice recording at that. 2L supplied both an unencoded 192/24 WAV file and the same, but encoded with the MQA process. Expectations were high since we’ve heard how much better the MQA file should sound, better than  even the original! I’ve seen pictures on the MQA website of people crying after listening – so much better the process is supposed to be. Imagine my surprise when it was worse than the original.

Worse is a bit harsh. Better, it was not; might be a nicer way to put it.

In fact, I was actually impressed how close the two were. After all, this is a pretty amazing process that allows streaming services to send high resolution audio without degradation – no small feat.

If all the hype and hoopla had merely stated the end results were indistinguishable from the original, I might be jumping up and down with how close they got it. Few systems have the resolving power of Music Room One and the fact they got close after folding the music into a smaller file size is quite an achievement.

But tears in my eyes?

Either the test board we were sent is flawed, or others that have openly wept with excitement heard something I have not, or this reporter is just confused and misguided. I do not claim to have the answer, I’m just reporting the facts as I hear them.

As Dragnet’s Detective Joe Friday might have said, “nothin’ but the facts, ma’am, just the facts.”

Subscribe to Paul's Posts

64 comments on “MQA thoughts”

  1. If I can crudely paraphrase myself from the last post.
    “It’s not the size of the file but what the engineer does with it”
    8^p.
    Some here must have rushed out and got one of Meridians cheaper compatable Dacs and tryed a test file…
    A small test board doesn’t seem the best way to promote uptake from HW houses. But then hype seems to makes more market demand than real world need. 196/24 v DXD….
    Good luck as always and roll on copper day..

  2. Paul, that’s again a nice non-representative anecdotical finding. You are biased in a multiple way by your expectations and your convictions about file format qualities. And your listening test wasn’t double blinded at all. Is the strange result of your finding that bit streams in the red book format with MQA coding are nearly indistinguishable from high res format’s bit streams?

    1. No, my finding was the a 192K/24 WAV file, encoded using MQA – having the streaming requirement and size of only 44K/16 – sounded very close to the original 192K/24 file. Thus, they’ve managed to reduce the filesize of high resolution audio and not damage the end result to a great extent.

      Their marketing claims are that they can do this, and when they do, the file actually sounds BETTER than the original file, a claim I did not find to be true. At least not in my setup, as described.

      Double blind listening tests are not valid and I’ve explained that any number of times. Perhaps it’s time again to explore this subject.

      1. Paul.. Lets pretend I drank the Kool-Aid…..a couple of questions…. ‘the cheesy DAC was supplied to us…. ‘ I assume supplied by Meridian? Secondly, the ‘2L Website’ … The 2L website… is that a Meridian website? So, where I am going with this is – if the cheesy DAC was supplied by Meridian – and the file was indeed encoded by Meridian with MQA – then there are no variables here…. If something or someone else was involved – then we obviously can still believe in the Unicorn!!! 🙂

        1. No, Meridian does not have much to do with MQA. MQA is a separate company Stuart started. The board came from them.

          2L is a Norwegian site with free downloads of excellent recordings and is unaffiliated with MQA except to say they have remastered some of their catalog with the process.

      2. Personally, I find the difference between 24/192 and 16/44.1 to be on the subtle side (for most recordings) when comparing like with like. Although the 2L recordings are excellent for proving to disbelievers that there is a plainly audible difference, I don’t find it remarkable that the smaller-sized file was “close” in sound quality.

        I’d be interested in having a third file including in the comparison: the 16/44.1 version from 2L. http://www.2l.no/hires/index.html?

        How does the MQA version measure up compared to plain ol’ downsampling? I assume the MQA is much closer to the 24/192, but I’d love someone to confirm that.

        The other issue is that MQA processing requires “knowledge” of the original recording or mastering chain (I believe) in order to remove errors that occurred when producing the final deliverable, whether CD or high-res files. Thus, an MQA file might be deemed subjectively “worse” even if the difference makes it closer in sound to the original.

      3. Hi Paul
        “Double blind tests are not valid and I’ve explained that any number of times. Perhaps it’s time again to explore this subject.”

        I don’t get it, why would you want to explain it again or why did you in the first place?

        Is it because you have to defend your position as a high-end equipment manufacturer or do you actually believe your above statement, and why only in this sector of science is it so?

        Any explanation is irrelevant since double blind testing is the only valid scientific process available to eliminate bias in a comparison, that’s if the objective is to find out the unbiased facts and not what you or I have a preference for, because it looks impressive or cost more than the competition.
        Reuben

        1. IAlmost all of the double blind tests that have been performed in this area are useless because they have been highly biased towards a null result.

          The latest ITU Rec, BS.1116-3 may actually be useful for determining audibility of small differences. Among other things, it requires sighted training before the test and more careful selection of test subjects than normal. Any tests not using the recommendations and having a null result must be considered invalid.
          https://www.itu.int/dms_pubrec/itu-r/rec/bs/R-REC-BS.1116-3-201502-I!!PDF-E.pdf

        2. It really depends on how you define it. I agree that testers cannot be told which is what. But here’s the difference. When emotions are involved, and when the complex structures of our auditory senses are involved, most of us need to be able to relax and not be on our guard. Once in a defensive position, we lose our ability to discern differences.

          The same is true for taste tests, or any emotional testing.

          Single blind works. Double blind works too if not done with a switch and if the test is constructed such that the tester controls when he’s ready for a change and confident that change is firmly with A or B.

          Maybe this type of test has another name.

          1. This reply is for you Paul and also “scottsol” whose comment puts the blame of the outcome, on the test.
            Belief – the reason humans aren’t able to be objective.
            Reason – because we were taught to believe from very young that life is a fantasy eg : tooth fairy, Santa, and most of our religious teachings have taken away our ability to be objective, as a result we are unable to make accurate judgements under circumstances where our beliefs come under scrutiny.
            Failure – the validity of the test is not in question, but the one that’s subjected to the test is.
            I see the discourse from my perspective as missing the forest….!
            Reuben.R

      4. Did you compare CD 44K/16 tracks with the corresponding MQA 44K/16 CD tracks? The same songs, but CD vs MQA CD? apples to apples? If so, exactly which tracks did you use? And exactly where and how did you get them? In any valid testing you need to document exactly what you did, and with what. This way the process can be reproduced by anyone to validate what you did. But this is just too vague to be taken seriously.
        Could you please document the exact tracks you used, and how you got them? Thanks!

  3. Paul,
    Is the small sample board you were sent that you refer to the original one that you thought might be defective, or a replacement board?
    If the MQA version of the hi-res file was close to the unencoded 192/24 WAV file, then that sounds like it would be an improvement over a 44.1/16 WAV file from CD, at little more bandwidth than the CD file. Isn’t that the intent of MQA, to improve the sound quality vs. CD quality for streaming with little increase in bandwidth? Shouldn’t the comparison be between CD quality and MQA quality without MQA decoding? If MQA wins in this test, then the average person who streams music online could get a gain in sound quality without buying a new DAC with MQA decoding.

    1. Not at all. MQA’s claim is that their end product will sound more like the mike feed than the master recording, regardless of resolution. From their point of view, sounding better than a CD is no more deserving of accolades than Kanye West being proclaimed better than Vanilla Ice.

    2. We did that test on the small board they sent and Red Book passed through the MQA decoder is worse sounding than without doing that. Which makes me suspicious something’s wrong with the sample board they sent us.

  4. As a lot of things in the audio world, the benefits of the new discovery are exaggerated.
    Manufacturers always tell us their new discovery is “so much better”.
    It’s what we all waited for, hallelujah !!
    PS Audio is no exception here.
    The BHK is undoubtedly a very fine amp.
    But why not resist the temptation to state the BHK is “one of the best amps” in the world, if not the best.
    Is that possible for a $7000 (stereo-)amp. Don’t think so, although price has not everything to do with it.
    Or PCM is much better than DSD… IMO mainly different, not so much better.
    Progress is slow, little steps instead of the giant steps manufacturers are telling us.
    So it often annoys me that every next “best thing” is so much more expensive !
    Probably has more to do with the high salary of the engineer than with the real technical progress.
    We often trade in to easily our already beautiful device for the “better” new one. I’ve made that mistake more than once.
    A bit off-topic here : what’s wrong with HDMI ?
    Yes that’s what I’m wondering too. I think it’s one of the best we have now. Isn’t I2s a kind of hdmi..?

  5. Eventually bandwidth won’t be an issue. A large number of us can already stream 1080p video so a 24/96 audio stream wouldn’t choke our connections especially if there’s caching. I do realize there are still many people and areas can’t access this type of service.

    The most important thing I’ve learned from Paul’s Posts is how much I was missing from all the 16/44 music I own and from the lossy music I sometimes stream due to a sub par DAC and music server. Now that I have a great DAC and a killer server, I hear things on recordings that take my breath away.

    I’ve not heard any MQA based music so I don’t have an opinion but I wonder about it’s necessity.

    1. A key reason for MQA (assuming it works as advertised) is the existence of music on physical media.

      On one CD you can store many different levels of quality, and which one you get to hear depends on the hardware you own. This vastly reduces stocking, manufacturing, etc logistics issues.

      If the future is ‘electronic package only’, then this reason goes away.

      But even then, some folk will keep their music in a music server. An 96/24 file for a track might be 224 MB. A 44/24 of the equivalent is 67MB. A 44/16 is 44MB. Five times as much storage, even when cheap, is a lot more storage.

      The 5x as much argument can be eased for commercial streaming servers, who still have to buy lotsa storage, but charge people (one way or another) for its use; and a given track may well be heard by many.

      It’s 5x bandwidth, too, and bandwidth is a naturally-limited resource.

      1. No thought of physical media was involved in MQA’S creation. The CD is becoming an increasingly niche product doomed to fade away. As production of CDS and CD players decline with no desire by any major player to reverse the trend, the idea of adding MQA to CDs makes no sense. While there are those who would certainly appreciate being able to get better sound from the CDs they own, most of them recognize that CD’s aren’t the way to add to their music collection.

        MQA does alow users to download or stream a single file and play it back on equipment of various quality levels without having to perform a conversion first.

      2. Okay, I was not aware of the multiple quality levels on one optical disc but I do think we are in the final years of optical.

        So in the name of efficiency, MQA is certainly prudent but storage might be so inexpensive in the next decade to null this. However, for streaming, it’s a good idea for economy. I can’t resist to think MQA would have been a bigger deal 5-10 years ago.

  6. Debunking this trash will take more time and space than usual even for me. So I’ll take it in chunks. Let’s see if I get through it in just one day. The first thing to know is what MQA, MQB, MQD, high definition, and high resolution audio are. No way to assess something if you don’t even know what it is or is supposed to be. First thing to know is something about frequency response (FR for short.) You’re getting the distillation of four years of hard fought hard won real knowledge from a real education. The investment in money was insignificant compared to the investment in effort. Do I remember ALL of it? Of course not. It was 50 years ago. What I got from it that I retained was understanding and how to solve problems. I like to think I could relearn whatever I forgot but let’s be realistic. My mind boggles when I look back at my textbooks. How did I ever get through so much math, so many thousands of equations. And that’s mostly what it is, math used to explain things. Lab work? Experiments to see how actual results correlate with the equations in the text books and classroom lectures.

    Among the first stereo systems I ever saw belonged to the father of a close friend of mine, Richard. The father was an accountant and apparently a successful one. He lived in a four level split that was twice as large as most other houses in the neighborhood. (no backyard though.) One day I went to visit him and there was this monstrous piece of furniture. I’d never seen anything like it and I’m not sure I’ve seen anything like it since. It had Bogen equipment in it, a name I later associated with PA systems more than hi fi systems. I was about 12 years old and of course at that age I wanted one too. I think that’s when my interest in hi fi was sparked. I never actually heard it. I used to go to Lafayette Radio and drool over all the shiny receivers, amplifiers, speakers, turntables. I was like a kid in a candy store with no money in his pocket and his nose pressed up against the glass display case. I went to trade shows, got catalogs (the Lafayette catalogs were great catalogs) read magazines, advertising literature. And I believed all of it at the time.

    I heard the siren song of the importance of FR extended beyond human hearing soon enough. It’s the same song that you hear over and over again periodically when the industry needs something new to sell. And at that time I believed it. My first amplifier that I bought was HK A500 with an FR extended to 70 khz. Not bad for tubes and transformers. The University Sphericon super tweeter extended to 40 khz (never had one.) Audio Fidelity Records boasted “The Frey Stereophonic Curtain of Sound” that extended to 25 khz. “You won’t believe your ears or so it said on the record jacket. But the all time champion was HK’s first solid state integrated amplifier, the A1000. Flat FR from 1hz to 1 mhz. And it actually lived up to its billing, it measured out to 1.5 mhz according to the magazine lab reports. DC to light. A straight wire with gain. What could be more perfect?

    At 17 I went away to engineering school. I did more math in 4 years than I’ve done in the intervening 46 years combined. And starting in the second year much of that math was targeted specifically at electrical engineering. EEs get math that applies specifically that discipline, math other specialized areas of science and engineering don’t get nearly the exposure to if they get it at all. FR, square wave response, transient response, impulse response, they’re all different ways of looking at the same thing. They give you one, you are required to give them back one or all of the others by calculating it. And then I knew. Look at the graphs. At the upper frequency limit of hearing, the threshold of hearing, the softest sound you can hear crosses the threshold of pain where sound is so intense it is experienced not as sound but as pain. FR above that frequency in a sound system is not only worthless, it can create problems that wouldn’t otherwise exist. The siren song was over. Those who still hear it, are slaves to it, must be tied to the mast like Odysseus was in Homer’s epic poem The Odyssey. They just don’t know, they don’t have adequate knowledge. And they are just plain wrong. That’s one of two elements in this latest marketing craze is about, extended FR beyond the range of human hearing. If they can’t sell you something new that’s useful, they hype up something to sell you that isn’t. They depend on the fact that you don’t know to excite you about what marvels you’ll hear and scare you about what you are missing. They have to be very careful in their advertising by suggesting something without saying it and lead you to infer what they want you to think. Advertising is an art that does just that. It stays just inside the law where seduction ends and fraud begins.

    1. If there is one thing that MQA is not about, it is extended frequency response. I will not go into details as the information is very easy to find if you actually want to.

      1. I have referenced two videos that give the rationale for the concept of MQA and the method of how it is executed. Anyone can watch those videos and decide for themselves but I have concluded that extended FR well beyond 20 khz is an inherent element of the concept. The way it is accomplished using a channel only slightly greater than RBCD is one of its characteristic “tricks.” If it works, it is clever but since no one can hear beyond 20 kHz even if there is content in the recording in the ultrasonic range, this capability is useless.

        1. MQA is much more about timing issues than amplitude response. The folding principle is, in fact, based on there often being little or no HF information. While short times are associated with high frequencies, it is dangerous to assume that the inability to hear high frequency tones is related to our timing sensitivity.

          Previous double blind tests that purported to show that we can’t hear the difference between tracks with and without ultra high frequencies or our lack of ability to detect small timing differences can not be considered valid unless they used a procedure similar to what is found in ITU REC B.S.1116-3

          The null tests that I am aware of did not follow these procedures and were inadequate for testing the audibility of small differences.

          There may or may not be audible differences, but the double blind test used to support your position are outdated.

  7. Paul,

    By your own standards, the test you described wasn’t a fair or reasonable test (yes, it was what you were given to play with…)

    Why wasn’t it reasonable? If it were, there’d be no need for DirectStream, DS Junior et al. You can do it all with a small board and a cheesy DAC.

    I’d consider harassing Mr Stuart to the point he provides you with a decoded bitstream from an MQA chunk o logic, and feeding that into a non-cheesy DAC (I hear y’all have several 🙂

  8. I posted this on Mark Waldrep’s web site. I’ve cut and pasted part of it here. The quotes are from Mark Waldrep’s talk himself. I’ve posted a link to it in the text. I like and respect Mark Waldrep. I don’t agree with much he says. He’s a recording engineer, I’m an EE. Our knowledge overlaps to a degree but each of us knows things the other doesn’t;

    What about High resolution audio? Here’s are some authoritative quotes; “The honest truth is that at the end of the day if I played a 44.1,16 bit recording in here and I played my 96K 24 bit originals none of you could tell the difference. My friends in the mastering community can’t tell the difference.” “Guitar Noir is one of our best selling albums….notice that the full scale of 96 24 goes up to 46 khz. …there are frequency components that actually exceed 20 khz. But Mark you said earlier we don’t hear that. I don’t care. My definition of fidelity means that if it’s coming out of…..I’d like to capture it ….maybe, just maybe there is something going on in our brains…..” “Folks, we’re deluding ourselves….compact discs, this is our reference folks, this is our standard definition audio. It’s not hi rez, it sounds wonderful and anybody sits there and says oh yea, they can instantly hear the difference between a CD and a hi rez file they’re crackers, it’s not easy to do. We did tests. They played them through headphones, they played my 96K 24….downconverted from 96.4 to 44.1. Well guess what, every recording that they used as a test…”

    https://www.youtube.com/watch?v=Z5S_DI99wd8

    I’d listen to this guy. He seems to know what he’s talking about. The difference between RBCD and hi rez is very subtle [if it actually exists at all and isn’t attributable to factors other than FR] and there are only a few thousand hi rez recordings available.

    I have no dog in this fight. I’m not in the biz. I’ve got about 3000 vinyls and 3000 cds, in very round numbers. I also have turntables that I like very much to play the vinyls on and lots of CD players. But I rarely listen to vinyl. For one thing it’s a PITA. And there’s no remote control. I also appreciate the lack of pops and clicks especially on soft music. CDs that I bought over 25 years ago sound like brand new. They never seem to wear out. Okay, I’m sold, RBCD is it for me. (it was from the first time I heard one but I had to wait until they got the steely sound out of violins before I bought one.)

    Here’s Dr. Waldrep’s response;

    “Mark, you’ve got way too much time on your hands to do all this research.”

    Here’s where my posting comes from.

    http://www.realhd-audio.com/?p=4103

  9. I have come to the conclusion over the past few months reading this blog, Mark Waldrep’s blog, the major magazine’s posts and others that, the whole fuss over cables and interconnects, file types, MQA, provenance and so on is pretty useless given how highly dynamically compressed most of the major recording label’s releases are. I bought the new Elton John CD the other day, and, imagine my surprise when Roon showed me the file map (at least that is what I will call it as I don’t know the proper term) and the file was so juiced up that it was simply a solid rectangle across most of the songs.

    No highs, no lows, just loud. I don’t care if you have an mp3 or flac or DSD or whatever the sampling rate is those tracks are going to sound as poorly on any system and any format. Sad, really, all this talk about the artist wanting to be heard as they recorded it but their own labels don’t even subscribe to that. I am not going to change the music I listen to however simply to accommodate a new file type or service and from my research as well none of the artists that I follow or want to listen to are doing anything at all in high-rez or anything beyond standard CD format. I’m kind of done trying to care about many of these matters, (at least that is my thought today) it’s just not worth it when the musical offerings of all my favorite artists and genres are so bastardized.

    Larry

    1. The difference between an Elton John and most other Johns is that an Elton John has a heated seat. Most useful on those cold winter days. Much more comfortable. 🙂

    2. Good points Larry.

      I’m currently listening to Kyle Eastwood’s latest album ‘Time Pieces’ (16/44) and the production value and mastering are stunningly good that some might think it’s high resolution.

      I think we will always return to the source and mastering argument as it trumps other arguments.

      1. Over the past few years I have gravitated more toward unplugged, acoustic or live albums by groups I follow, finding them less tweaked and compressed than the studio releases. I am not familiar with Kyle Eastwood will have to have a listen!!

  10. Yes Larry, I agree, you are right about recordings.
    The producer decides where every instrument is placed, how it sounds.
    He is the one who makes the choice concerning dynamics, sound, placement and other properties.
    Where we listen to is artificially manipulated. MQA does not change that. PCM or DSD does not change that.
    So, all the discussions about wrong or right, neutral or warm sounding,…It’s all relative.

  11. Paul’s comments strike me as entirely reasonable. I have never understood how a file encoded with the MQA (or any other) process could possibly sound better than the original. As Paul said, to come close to the original or even match it while providing a streaming-friendly smaller file is a considerable achievement.

    Yes, all audio companies engage in marketing hype. Is the BHK one of the world’s best amplifiers? I don’t know, because amps that cost $10,000 and up do not visit my home. But it could be, given the experience of the designer and PSA’s history of producing quality products. So that marketing doesn’t put me off; it’s just marketing and might even be true.

    Claiming that a processed file sounds better, though, strikes me as something of an insult to common sense. Of course, saying that the decoded file sounds as good as the original sounds unimpressive, even though the opposite is true. So I suppose MQA is in a bit of a bind in terms of publicizing what they can accomplish.

    1. I haven’t seen the propaganda that says MQA-encoded sounds better than the original.

      I have read that a complete end to end MQA encoding/decoding will sound better than a recording done to the same quantization size and rate (eg 24 bit 192KHz) because they claim to be able to correct for timing errors in the microphone/mike preamp etc chain.

      It is not explicit from Paul’s post that what he listened to was an end-to-end MQA recording, or simply an MQA ‘compression’ encoding (so without source ‘correction’) of an existing hi-rez file. My interpretation, however, that it is the latter – one of 2L’s files was “MQA’d”, giving the same information density as the 192/24 (or whatever) while being packaged in a space-saving format – which of itself is a pretty neat trick. But it won’t do the source correction.

      Whether or not this is true and effective, encoding an already recorded 192/24 image can only worsen it, whether or not the worsening is audible. The worsening, for audio which meets the decreasing-amplitude-with-frequency model of music on which MQA relies to fit the extra info into bits normally used to represent 16/44, should not be detectable by anything because it ‘fits’. Music which has too much energy above (say) 50KHz would pose a real problem for the system, but there aren’t many human musical instruments which work like that, so it’s a moot point.

      In short, it’s a pity MQA Inc sent a poor piece of kit to Paul – because he’s bound to report what he hears.

      1. Ah. Now I have seen propaganda which says that a straight MQA-encoded recording can sound better than the original digital recording – because the encoding captures appropriate characteristics of the specific DAC it will be played back through:

        “An MQA decoder will restore the original recording and ‘unfold’ it to optimally match its D/A converter. ”

        From http://www.2l.no/hires/documentation/2L-MQA_Comparisons.pdf

        Assuming we’re in general agreement that there are differences between DACs, and that these are common to ‘all’ the delivered DACs, and that you can characterize the differences, then this sounds pretty neat.

        1. The most important part of the MQA process is probably compensating for the encoding characteristics of the ADC used in the recording process rather than the playback DAC.

          It is this process that is the source of MQA’s claim of their files being superior to the original master.

  12. Here’s another of my postings on Dr. Waldrep’s site;

    THERE IS NO SUCH THING AS HIGH RESOLUTION AUDIO! It’s a made up marketing term that makes no sense. Look it up in Wikipedia, if it says it in Wikipedia it must be right. 🙂 But the truth is that it is not a scientific term that describes the products called Hi Rez Audio at all. Here’s why. Resolution as a scientific term refers specifically to optics and related images in photography that has to do with how closely two objects can be perceived as being separate before they become a blur to being seen as one. Therefore, lenses, telescope mirrors, film, even CCD imaging devices can be rated according to their resolution, usually in arc seconds. So if one telescope shows the observer two separate stars while another shows them as one, the first has higher resolution or greater resolving power. If one photographic film can photograph not just visible light but ultraviolet and infrared as well and can photograph images clearly in light ranging from a fraction of one candlepower to looking directly into the noonday sun but produces grainy fuzzy images all the time, its resolving power is poorer than ordinary film that produces sharp images but only in the visible light range and only in normal lighting conditions. So while what is called high resolution audio may have wider dynamic range and greater frequency range than ordinary CDs, it does not have greater resolution. It is not an extrapolation of this scientific definition related to light into the field of sound.

    What would increased resolving power mean in audio. In amplitude it would have to be able to increase loudness in smaller increments than RBCD that are audible. But over about a 95 db range RBCD gives you over 16,000 loudness levels already. Even if your hearing is as sharp as Atkinson claims his to be, able to hear in 0.1 db increments, over a 95 db range that would be less than 1000 loudness levels. What about frequency resolution? My own resolution is about 1/4 to 1/8 of a halftone (a halftone is about a 5% frequency shift.) So my hearing has a frequency resolution of around 1%. Some people’s are somewhat better. This is the ability to tell if a note is sharp or flat. RBCD beats that by a country mile so there’s no discernible difference there.

    http://www.realhd-audio.com/?p=5503

    Dr. Waldrep says he prefers to use the term “high definition.”

    Just about ready to discuss MQA, MQC, and MQD.

    One further note about Mark Waldrep’s contention. He says even if the technology is capable of greater bandwidth and dynamic range than what we can normally expect to hear, unless the source material contains that information, all you get from the supposedly superior technology is a lot of extra zeroes. And he’s right. Even though you pay more for them, so called hi rez or high def recordings made from originals that do not have content much beyond 20 khz are no different from ordinary definition recordings in that regard. If they sound different, it is due to other factors such as the specific equipment used or how it is used, not because the equipment had technologically wider bandwidth or other capabilities. This is why he can say that whether you believe in the value of high def or not, only a few thousand recording actually can be called hi def. And he proves it by showing a spectral analysis of so called hi def recordings that aren’t really having the same spectrum as ordinary definition recordings. To be hi def, the entire recording process from the microphone forward must have extended FR far beyond the audible range. By Dr. Waldrep’s definition, to at least 40 khz to 46 khz. Old recordings made in the analog mode and most digital recordings do NOT fall into this category. So with few exceptions, the whole thing’s a fraud designed to get more of your money. Buyer beware.

      1. Why not read his reply for yourself and make up your own mind. The term “resolution” has a technical definition that does not apply in this case. Therefore as used by the audio industry, it is merely a marketing slogan to impress tyros. If you want a further explanation of its scientific definition, look it up on Wikipedia. That’s what I did before I wrote this.

  13. MQA explained and debunked. The first video linked below explains the flawed rationale behind what is called hi rez or hi def audio and the second explains how the technology squeezes 24 bit 192 khz into a space not much larger than 16 bit 44 khz.

    https://www.youtube.com/watch?v=r_wxRGiBoJg

    Flaw #1, the auditory resolution of between five and ten millionths of a second has to do with the difference of arrival time between your two ears, not the period (inverse of the frequency) of the sound itself. That remains 20 khz or less.

    Flaw #2, you do not need to have a signal recorded at higher than 44.1 khz to avoid problems with filtering above 20 khz. Oversampling accomplishes the same thing with 44.1 khz being sampled over and over again such as 8 times 44.1khz or even at just 192 khz and you can apply the same filter as though the signal actually was recorded at 192 khz.

    Flaw #3, classical music has the widest dynamic range requirement. Other music no matter how loud it is played has a much more limited requirement. The Architectural Institute of America’s requirement for an empty dead quiet concert hall is 25 dba. The loudest an orchestra plays as heard in the audience is 105 dba. 120 dba is the threshold of pain and is dangerous for your hearing. Therefore 16 bit which gives a 96 db range is more than sufficient for any orchestral music which is only 80 db. In the real world, when a concert hall is filled, the background noise is probably around 40 or more db making the actual dynamic range of what you hear even less than 80 db. (You can verify this with a freeware SPL meter app in a smart phone at the next live concert you attend.)

    https://www.youtube.com/watch?v=T5o6XHVK2HA

    Assuming it works as claimed, it is a clever way to provide something that is IMO absolutely worthless.

    1. As to Flaw #1, my understanding was that shortfalls in the recording equipment, including microphones being physical objects, and each capsule bing different from others, meant that you couldn’t be at all sure that the (say) 5 microsecond delta-t observable in principle at the microphone position was actually detected/encoded by the actual microphone, but that since the microphone was an essentially linear system one could fix this in post-processing. Then, in a traditional digital recording chain, you need to encode this as higher data rate encoding, else it all gets smoothed back into non-existence because it all appears in one 44KHz sample bucket. (Or, I suppose, sometimes statistically flops around between two buckets with an even nastier result. En principe.

      And you (or rather ‘most people’) can hear into noise. It all depends on how you measure things what you can hear, but it looks as though folk can hear signals which use the bits below (as it were) the bits available in 16 bit encodings. That is, you need to encode the signals with a small number of bits on a 20 bit encoding. Been looking around to get documentation for this, but the best I can come up with (on St Valentine’s day, so there are other pressures…) is the discussion here:

      https://www.gearslutz.com/board/music-computers/553108-hearing-below-noise-floor.html

      1. First of all, a waveform with a period of 10 microseconds is 100 khz. No human can hear it. Even your dog can’t hear it. It’s the top range of what a bat can hear. At 5 microseconds it’s 200 khz, beyond the range of even a bat. This is the kind of mixed up nonsense they use to sell their junk. Test your own hearing with your hi fi system and a test disc, downloaded app or a signal generator. See what YOU can really hear. If you are like acuvox, take off your eyeglasses for the test if you wear them. He claimed he can hear above 20 khz through his eyeballs.

        Audio tape traveling at 30 ips can’t provide 30 khz response. The only magnetic tape systems that can are helical scan systems used for video tape. The tape head spins at high speed, for NTSC VHS at nearly 1800 rpm and has at least two magnetic gaps and over 180 degrees of wrap so that at least one gap is always in contact with the tape. This makes the effective tape speed against the tape head very high. At this speed the bandwidth is enough for a 3.4 mhz down converted NTSC TV signal even at the slowest forward tape speed although the slower the speed, the poorer the video resolution. We know this is so because if it were less you would not get the color burst signal on the tape which is 400 khz bandwidth all by itself. SVHS provided 5 mhz and could handle high fidelity sound channels. Early PCM adaptors allowed video tape recorders to be used for audio taping but editing was a problem unless you had an open reel machine. Only a few of that type were ever offered to the consumer. Professional machines of this type used in the early days of television cost $50,000 when that was a lot of money.

        1. I suspect that tape running at 30ips can provide 30KHz – the graphs on this page

          http://www.toneboosters.com/reelproject2/

          …only go to 15 ips, but it looks like the 30KHz mark ain’t far away.

          I also recall a discussion I had with the folk who manufactured tape heads in England, many many moons ago(Marriott). I’d had the idea that recording using a nice 100KHz centre frequency FM signal might be a fine thing to do to get nice, low-distortion tape recordings, and even mantled up a little demo circuit with a VCO and a feedback circuit including a tape head. They wrote back and said “don’t be an idiot – 100KHz at 7.5 ips for this sort of application is no problem at all. Just need the right gap, and we can do that”.

    2. hmmm…
      “THERE IS NO SUCH THING AS HIGH RESOLUTION AUDIO! ”

      Frequencies are composed of a fundamental frequency and its higher harmonics.
      If we want to hear a perfect square wave, for instance, we’ll need a great spectrum
      of high frequencies, the highest frequencies not needing great amplitude.
      If they are not there, you will get a, not so square, wave, with ringing artefacts,
      undershoot, overshoot and Gibbs phenomenon…

      The harmonics will be distorted and you will ‘hear’ that it is not the proper square wave
      you were aiming at.

      So, the higher the resolution, the cleaner your sound.

      1. Pure nonsense. First of all there is no such thing as a perfect square wave. The math makes it clear that to achieve a perfect square wave the harmonics would have to extend out to infinity hz. Square waves are just one more method of looking at the same concept as frequency response. That means that a square wave that is flawed because it does not have components above 20 kHz will sound identical to a human as one that does. Learn the math, then come back and we can discuss it when you are better informed.

  14. I must be one of the few in the audiophile world who did not know what MQA is, nor had I heard of it until I read Paul’s post today. I may have seen something in Stereophile or Absolute Sound about it, but I don’t remember. I likely would have blown right by it looking for analog news or music reviews. Then, I have almost no interest in digital recordings and zero interest in streaming from some cloud or internet site. I would rather enjoy the music.

    1. Hi Vinyl Lady,
      There is lots of intellectualism purely for its own sake here and very little mention of the single most important aspect of audio which is music. I listen to a very big and loud stereo but fell in love with reproduced music listening to a tiny little hand held transistor radio. It should be first and foremost about the music but a lot of guys like the machinery above all else. On the other hand I embraced cd’s immediately and gave up on vinyl decades ago.

        1. I first connected here after seeing Guru Paul’s IRS videos on youtube and specifically asking him what he would play to demonstrate what I expect would be their great authority and power. Not forthcoming I sent him some of what I would play and then had 3 cd’s sent to him. I was looking for recorded treasure.
          I mailed him Transiberian orchestra Lost Christmas Eve, and Nightcastle and a Cincinnati Pops fantasy album warning you not to blow up your stereo. Never got back to me. btw the new TSO album Letter from the Labrynth is best SOUNDING album I have ever heard but I say that TSO in particular is the sound I had in mind when piecing together my stereo.

  15. Enjoying the blog, Paul. I am also impressed by you honest listening impressions!

    Just wanted to let you know that your impressions are probably consistent with the objective testing results here using the Explorer 2:
    http://archimago.blogspot.com/2016/02/measurements-impressions-meridian.html

    looks like there is a loss of resolution in the MQA file even when decoded presumably properly with the new Explorer 2 firmware. Was the “cheesy little DAC” you were sent also the Explorer 2?

    1. Fascinating and thanks for this post. Turns out, in our case, we did not have the latest firmware installed on the DEV board we used (not the Explorer). I am heading off for some R and R next week but upon my return we’ll jump back into testing.

      1. Thank you Paul. Have a great week!

        Please do let us know when you get a chance to talk with the Meridian people if there are any plans for digital decoding so we can play these decoded files on excellent DACs we already have. If what you’re hearing and these guys measuring is the “final” form of MQA, there’s not much incentive to have to buy a new DAC. If they refuse to do this, I really don’t know what chance there is of success for this MQA format. Smells too much like HDCD all over again.

  16. This comparison misses the point. It’s like a Mythbusters episode where they did it wrong. The real point of MQA is redbook vs MQA-encoded redbook. The higher-res you go, the less difference you will notice. Why didn’t you go with redbook comparisons? Answer: perhaps because the forte of your products is higher-res and you are hoping people will pass up MQA and buy your gear based on what you said – or perhaps not. But most everybody will be streaming lower-res smaller files. I believe I read that on 2L.com there are redbook-level recordings and the same corresponding MQA/redbook recordings. If not on 2L.com, then somewhere else. Download one and the other, and then do the comparisons. The point of MQA is to upgrade lower-res / smaller-file audio, and make streaming of really great sound possible – or so they say. I personally have not heard this, and to be fair it annoys me hearing someone say something like, “I can’t see how it would be that much better”, or even “Wow this is great. Let’s go for it!”, without hearing it themselves. I “believe” in it, but only so far, so far.

    But at least be fair in comparisons.

    And judging by a little board? Doesn’t seem like a good marketing plan by MQA, but come on – let’s sample with a few DACs with MQA, and not a little piece of schmutz, even if your own gear doesn’t support it yet.

    Maybe MQA has come out too much too soon and too strong, since the MQA music sources aren’t really supported yet. But a lot of reviewers with reputations are putting those reputations on the line saying what they are about MQA, and I bet that MQA isn’t paying them off to give glowing reports.

    Let’s give it chance and hope that they can deal with licensing issues and actually start getting MQA stuff out there that the common person and use, so we all can hear a difference – or not.

    Gary

    1. Thanks Gary but perhaps I wasn’t clear. The first tests we did – and the majority of them – were Red Book comparisons with and without MQA. My current thinking is that the improvements they’re hoping for, with their filter changes – might help some DACs but not all. But, as I said in the post, I don’t know.

      We’re continuing to work with them in the hopes of getting to the bottom of this. I’d like it to work because as a streaming technology it’s something we can support – as long as it doesn’t turn into another means of DRM that restricts us from high resolution downloads.

        1. We did 192/24 WAV files – MQA encoded and the original files. These were all taken from the 2L website and they were apples to apples.

          On the CD versions there’s no MQA encoded CDs I know of. We simply compared a number of CDs straight into the DAC vs. through the MQA decoder.

  17. Another issue is, how do I do it right now???

    Before jumping in whole hog I went to the MQA website and looked at what companies they said support MQA. I typically don’t blindly believe what I read, even if I want it to be true. I went to some (not all) of those companies’ web sites and emailed them directly, asking if they actually do support the new format. A few did not reply, but of the ones that did, not one of them are on board with MQA yet: Auralic (several email responses), Aurender (email response), Tidal (no response), Oppo (several email responses. Oppo is not on MQA’s list but I emailed them just because I have an Oppo 103. According to Oppo, their current products will not be firmware-upgraded to do MQA (not enough DSP horsepower), which is really a shame, but perhaps by the end of the year a new product might support MQA), etc. This response is typical.

    From what I can tell, it boils down to licensing that is not in place yet. Why not? Companies seem to be playing a waiting game, and the main reason I was told is the lack of “software” – MQA-encoded files to download or stream.

    It’s the chicken or the egg – which should come first? Music of course, but why create MQA download / streaming files if nobody can decode it? Why buy an MQA DAC if there is little MQA streaming easily available? Not yet ready to pull the trigger on a promise of a new audio wonderland. Only a few non-Meridian companies do support MQA so far that I can see.

    So why does MQA say these companies support the new format if they don’t? From what I’ve seen and heard, maybe MQA means that they are talking to these companies and are hoping they will come on board. Even so, that isn’t what their site says and it is misleading. Still waiting, wondering if an MQA investment would pay off.

    MQA I think, needs to lead the pack in creating lots of MQA software (downloads) for us, and at first supply it all for free, just to get people to jump in and enjoy the benefits. Of course, MQA files need to come from master tapes, not just converted from redbook files, and that probably means licensing fees if MQA doesn’t own those master tapes.

    So for now I’m still on the fence.

    If I’m not up to date about this, please update the list, and what’s going on. Things change.

    Gary

  18. I am fascinated with any change and sometimes change does create an improvement over the status quo. The discussion of the wide variety of audio streaming methods is wonderful and the resolve will , probably, result in more enjoyable listening. However, the method of selection of a process to be used for the recording/streaming based on “listening” tests
    are extremely inaccurate since auditory memory is extremely inaccurate and unreliable. The deviation in data used to analyze auditory differences has never been data with small standard deviations.

    To further complicate auditory memory, the time from listening to listening sample presentation can very easily fall into a abyss that is characterized by the time that transfers auditory data from the temp memory bank to the permanent
    memory bank. It is well known that there is a “dip” in memory (recall) when this transfer occurs and that the resolve when the data is placed in permanent memory bank

  19. I
    am fascinated with any change and sometimes change does create an
    improvement over the status quo. The discussion of the wide variety of
    audio streaming methods is wonderful and the resolve will , probably,
    result in more enjoyable listening. However, the method of selection of a
    process to be used for the recording/streaming based on “listening”
    tests
    are extremely inaccurate since auditory memory is extremely inaccurate
    and unreliable. The deviation in data used to analyze auditory
    differences has never been data with small standard deviations.

    To further complicate auditory memory, the time from listening to
    listening sample presentation can very easily fall into a abyss that is
    characterized by the time that transfers auditory data from the temp
    memory bank to the permanent memory bank. It is well known that there is a “dip” in memory (recall)
    when this transfer occurs and that the resolve when the data is placed
    in permanent memory bank is usually below by as much as 20% accuracy.

    So we have another variable in the mix-human memory insufficiency(HMI). If we want to have fun we can always
    call this HMI, try to calculate it and create an equation to include all the variables such as IMD, HD, etc etc etc.
    and wind up with a new measurement. The measurement can be called………………………………”.IF” for imagination factor.
    I can visualize a totally new ANOV.

    What a Hobby!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

    Larry

Leave a Reply

Stop by for a tour:
Mon-Fri, 8:30am-5pm MST

4865 Sterling Dr.
Boulder, CO 80301
1-800-PSAUDIO

Join the hi-fi family

Stop by for a tour:
4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram