Dr. AIX's POSTS — 07 December 2014

By

I heard from Robert Stuart of Meridian via email today. He sent a very nice note offering to provide additional information about Meridian’s new MQA technology AND to come by and demo the technology on his next visit to Los Angeles. We tried to connect during the AES sessions but his schedule, as you might imagine, is very crazy. He beat me to the punch. I was planning to write to him today and ask about the MQA process. I’ll keep you posted on what I learn. I may even be able to convince Robert to give me an online interview…I’m not promising, but I’m certainly willing to ask.

One thing that I did want to quibble about is the illustration that Meridian provided on its website showing the contrasting paths that “Quality” vs. “Convenience” have taken via the various formats that we’ve had over the years. I used the diagram yesterday but have reproduced it again below:

141206_mqa_3_image

Figure 1 – An illustration taken from the Meridian “Music is Changing” website.

Maybe the individual that put together this graph is unaware of the actual specifications associated with the various recording and delivery formats, but I think it’s important to get our relative “quality” scales closer to the truth.

The highest gold colored “Quality” line is associated with Reel-to-Reel analog tapes. Then comes Vinyl LPs followed by DVD-A/SACD, compact discs, cassettes and finally streams. In terms of frequency response, dynamic range, crosstalk, distortion, speed accuracy, and other less obvious factors, the graph is completely wrong. It seems designed to please analog advocates at the expense of recent digital formats. Perhaps I shouldn’t be so nit picky but I prefer that illustrations…like words…be precise and accurate.

I’ve remade the illustration to show the different formats and their potential fidelity or “quality” levels. Take a look:

141207_graphic_redone

Figure 2 – The same illustration with the “relative” qualities plotted more accurately.

There is room for interpretation in this redone graphic. I would probably add another category higher for downloads…after all there are sites that are making available 352.8 or even 384 kHz PCM downloads (they call them DXD but because DXD is a marketing term and not a format and I don’t want to confuse the issue). But clearly DVD-Audio has the potential to easily eclipse analog tape. And analog tape doesn’t suffer from the compromises associated with vinyl LPs. SACD/DSD…I’ll just leave that for now.

Compact Discs actually do a great job of delivering both quality and convenience but real high-resolution PCM audio cannot be surpassed.

Got to run to the Los Angeles and Orange County Audio Society Annual Christmas Gala…back tomorrow with a full report.

+++++++++++++++++++

I’m still looking to raise the $3700 needed to fund a booth at the 2015 International CES. I’ve received some very generous contributions but still need to raise additional funds. Please consider contributing any amount. I write these posts everyday in the hopes that readers will benefit from my network, knowledge and experience. I hope you consider them worth a few dollars. You can get additional information at my post of December 2, 2014. Thanks.

Forward this post to a friend and help us spread the word about HD-Audio Forward this post to a friend and help us spread the word about HD-Audio

Share

About Author

Admin

(28) Readers Comments

  1. Mark,
    You pretty much hit the nail on the head with your revised chart though I would grade the CD a little higher over all.
    Due to the heavy slide in quality in the quest of convenience, the Redbook CD has become the wiping boy for the analog community that was more deserved by the MP3 explosion in the marketplace. I also note the total absence of MP3 on the chart which should be close to leading the tan line of shame in the demise of sound quality.
    Time we stood up to the analog LP crowd in their constant berating of the CD medium. Given equal quality recording techniques and media mastering/manufacturing, the CD kicks butt over the LP, and that’s just the fact of it. 🙂

    Steve Dahl blows up hundreds of LP’s on the pitchers mound of the Sox Comiskey Park in Chicago July 12 1979
    http://www.csnchicago.com/sites/csnchicago/files/081313_disco_demolition_645.png
    Steve was truly a man before his time. LOL

  2. Yes your comments on the quality vs convenience graph are right on.
    Good humor in your comments.
    I really like the changes you made.

    Daron

  3. Mark wrote: “Compact Discs actually do a great job of delivering both quality and convenience but real high-resolution PCM audio cannot be surpassed.”

    Quantitative clarification please as this seems to go against your previous advocacy of 16/44.

    • I’m saying that CD can deliver a great fidelity experience but high-resolution PCM audio recordings can eclipse them.

      • How?

      • I have the same question. It’s not clear to me how any digital formats (which are quantified from master Reel-to-Reel analog tapes) can have a “higher quality” than the Reel-to-Reel analog tapes themselves. Perhaps I’m overlooking something. I would just place Reel-to-Reel at least on the same level as high-quality digitizations. I’m less concerned where LPs are placed.

        • Generally speaking, a digital version of a source that started out as an analog tape cannot have high quality. It is possible to remove some of the noise from analog tapes…I’ve used NoNoise and others in the past. However, a new digital recording blows analog tape away.

          • Thank you for clarifying that for me.

  4. Hi, Mark!

    Does use of MQA expect to reduce timing errors compared to CD or FLAC used for the playback part of the end-to-end recording-playback cycle?

    • I’m going to be speaking to Robert Stuart this week and getting a complete rundown on the new format. As I understand the format thus far, their technique is about delivering the same fidelity of 192 kHz/24-bit PCM in a much smaller bit bucket. This reduces the timing errors to below 5 microseconds.

  5. That completely destroys the credibility of Meridian in my eyes, just as I was saddened to see Bryston started to sell “audiophile” power conditioner snake oil they used to take a principled stand against.

    • I’m not in agreement here. I know the principals at both of the companies you mentioned and don’t believe they’re snake oil salesmen. I saw lots of that yesterday at the LA & OC Audio Society Gala and raffle.

  6. The individual that put together this graph and the management that approved it did NOT care about the actual specifications. They only cared about passing along the marketing message to support the campaign. As I implied before, there’s a suspiciously large amount of pseudo-scientific marketing mumbo jumbo in all this story obscuring the scientific and technical realities and confusing the issue. Maybe this is the result of trying to support a solution seeking a problem? Immediately when I first heard the stuff about frequency vs time domain and the implied need to record at sample rates much higher than 96kHZ only to then get rid of ‘all the zeroes’ (as if previous compression methods did not get rid of them in the first place) I was suspicious of this. Disappointed. I’m sure this thing works but I’m far from sure we really need it. Wake me up with a large bowl of pop corn when people start ‘debating’ the audio quality of this vs that and how ‘analog’ this new thing sounds. We still have not agreed about what ‘resolution’ means in the context of digitized audio and we are taking another step back towards universal confusion.

    • I believe this is an important step in the evolution of digital audio…it’s a component, not a fidelity revolution.

      • Mark,
        Sounds like MOA might offer a genuine advance in speed of delivery for large size HD files. But sadly as some replys here highlight, the high end industry has so degraded it’s credibility that even genuine advancements are now automatically being questioned by consumers that halved been conned one too many times.
        SAD

        • We’ll have to wait and see just how important the MQA technology is. I’ve been reminded that it sounds similar to the HDCD metadata tagging that Pacific Microsonics did in the low order bits of the 16-bits CD words.

  7. While it is difficult to find much information on MQA, Meridian has not claimed that MQA is a true lossless compression algorithm. Meridian is stating that if a master recording is encoded and then decoded with MQA, the MQA recording will be heard by a human as identical to the master recording. This claim is being made for analog master recordings and digital master recordings up to 768kHz sampling rates. Meridian is not saying that a 192/24 file encoded and then decoded by MQA back to 192/24 would be bit identical to the original file. Meridian is stating that the additional MQA information can be embedded within a 44.1/16 file without compromising the quality of the file when played back without MQA decoding.

    Meridian is stating that 96/24 PCM is not sufficient to provide this identical sounding capability. To me this is the more interesting statement. We are still learning, and at this time, we don’t completely understand how human hearing works, but we have known for a long time now that our hearing is not based solely on linear processing. Here’s a paper documenting recent research on this.

    http://physicsworld.com/cws/article/news/2013/jan/31/human-hearing-is-highly-nonlinear

    http://arxiv.org/pdf/1208.4611v2.pdf

    • Thanks Mark…I’ll be talking to Robert Stuart later this week. This is certainly interesting.

    • Interesting. I wish I could just read the mathematics behind his technique, to figure out exactly what he is doing here. While he may not be claiming that his coding technique is not _always_ lossless, I was under the impression that we was trying to achieve lossless compression under “reasonable” circumstances, like a resonable decay in the dB/frequency spectrum. An extremely naive way to do this is to simply discard anything outside of the reasonable dB/frequency spectrum on the Fourier Transform side, and then find a reasonable way to encode what’s left. This would be lossless in “most” scenarios, but certainly you could construct scenarios where they are not. But if the signal obtained is lossless up to 20kHz (say), then does it matter for the loss? While this is certainly true for the continuous Fourier Transform (but where efficiency of encoding makes less sense), this is less clear for the discrete Fourier Transform as to how this impacts what is “audible”.

      • I’m hoping to get more information when I spend some time on the phone with Robert of Meridian.

  8. I see in your revised graph that you put Download at the same level as DVDA and stream same as CD, while this is possible it is a small percentage of users, if you take the major percentage of users who are limited by the media supply I would rate it lower, JMHO about real life use.

    And for both graphs I would put stream lower that download from a convenience, I commute an hour across country and today I can not drive for 5 minutes without losing data 🙁 At home or in the city it is a different story.

    • Take a look at today’s post.

  9. With your Fig. 2 you throw away any worthiness your blog it might otherwise have had (and you’re looking for funding from your readers to attend CES?!). CD on a par with reel-to-reel and of higher quality than vinyl? It’s easy to be dazzled by theory, particularly when one only has a slight appreciation of it, but have you actually listened to vinyl in comparison to 16-bit audio, or is your level of inquiry satisfied by Nyquist theory? If you have and believe CD to be ‘better’ then clearly our Ears are cut from a different Cloth, as are Sony’s and the late Herbert V. K’s when, in his mid-70s, he was Sony’s mouthpiece for ‘Perfect Sound Forever’.

    When I first listened to CD in comparison to my vinyl almost three decades ago — on a modest STD, Hadcock, Nagaoka, Nytech, AR18 system with Phillips 2nd-gen CD player, I was aghast at what this ‘perfect’ new medium offered for treble and dynamism: the dynamics were not just crushed, they were gone — just as they are with a 1 7/8 ips cassette copy of a vinyl record, speaking of ab encoding system that is not transparent to the source. The treble of CD is laughably poor, even in comparison to a pre-recorded musicassette. You can perform the same easy test that I did: listen to the closing percussive tinkles of ‘Final Sunset’ from Brian Eno’s Music For Films album on high-speed duplicated pre-recorded cassette and then listen to the same on CD, using either the EG original or the Original Masters version. The difference is night-and-day, and if you can’t hear it then you really have no business writing on audio matters. Simply put, one (tape) sounds like real music, if a poor imitation of the vinyl version, whilst the other (CD) sounds like a cardboard pastiche.

    65,000+ samples per second sounds impressive enough, but if you think about the 15,000 cycles per second that define the 15 khz tone that most of us will be able to hear, then it’s no real surprise that CD sounds such a joke in the treble given a sample variation of just 4 at this frequency. Did anyone at Sony or Phillips ever sit down and listen to their delivery system at a range of bit-depths to compare it with a master tape to see at what level they could detect a difference? No, clearly not, and ever since others such as the author of this blog have been content to promulgate the superiority of this system in the media for anyone willing to isten (if not to the music itself).

    Anyone with access to the most mundane record deck can hear this difference for themselves against any CD player at any price imaginable — audiophile analogue (i.e., 70s vinyl without digital delay lines in the cutting head) is transparent to the master tape because it runs at a similar angular velocity and has processing limited to amplification and RIAA tonal correction, unlike the breaking-up and remaking of the signal of digital processing. It’s as plain as the difference between perfect analogue 625-line TV and the artifact-ridden Youtubey mess that that is Digital Freeview broadcasting in the UK.

    This high-frequency bandwidth problem with CD has bothered me ever since my first listen and effectively ruined music for me since CD put paid to readily available vinyl (and loudness-war equalised digital audio has turned this musical pastiche into a painful experience), but I’ve often thought that there should be some empirical evidence to substantiate this high-frequency problem.

    Fortunately, someone with the equipment has shown what digital sampling at CD and higher bit depths does to high-frequency analogue audio, to the degree that even 192 khz sampling is not enough. See

    http://www.channld.com/vinylanalysis1.html

    …and look at those graphs.

    Don;t be fooled by the low level of the signals involved — these are harmonic fundamentals that are essential to the music. I can hear the difference between the MOFI vinyl and CD of this album, and it sounded then all those years before this graph was published like it looks now — data truncation.

    I can hear it. Why can’t you?

    I will never forgive Sony and Phillips for what they did to one of the joys of my life. Ironically, in ruining music for us all, they created a binary system that lends itself perfectly to the faithful copying of that signal, the distribution of comprises fills so much of today’s internet traffic. Home taping never killed music. The purveyors and proponents of commodity digital audio did.

    • Dave, thanks for the thoughtful and lengthy comment. Analog tape CAN provide ultrasonic frequencies as the spectra at Channel D’s website show but they can’t even approach the dynamic range of a CD (60-72 dB for analog tape and 93 for CDs). I looked at the graphs…and I’ve seen lots of spectra before. I use them a lot to show the noise associated with DSD etc.

      Unfortunately, your grasp of digital sampling is incomplete and flawed. I’m actually beginning a series on the Nyquist Theorem with today’s post. You might find it an interesting read.

      There is no one telling you to throw away your vinyl LPs or analog tapes…but the question of sonic accuracy and fidelity is not going to be won by older analog technologies…as much as you and others might want them to.

      High-resolution PCM digital is by far the best audio encoding format humans have yet invented and implemented. If you’ve downloaded the 96 kHz/24-bit files from my FTP site, you’ll know what I’m talking about. There will always be people looking that want to hold on to the technologies of the past and if they work for you then continue.

      I can hear the difference between a piece of vinyl and a compact disc. And you’re right they do sound different but that doesn’t mean that vinyl LPs have greater fidelity than digital. It simply means that you prefer the sound of analog and vinyl. However, others disagree with you…including me. And Andrew Quint of The Absolute Sound. After hearing our tracks in my studio he wrote:

      “But the multichannel audio, emanating from five B & W 801 loudspeakers, is quite simply the most realistic and involving instance of recorded sound I can recall, from any source format. Mark Waldrep knows what he’s doing.” I’ll stick with his opinion.

      I got an email from a seasoned audiophile calling my recording of the Mozart Clarinet Quintet the “Holy Grail” of audio. He was glad to have lived long enough to finally experience that level of quality.

      It’s funny because I just got off the phone with a very large Asian electronics company that will be introducing a new electro static speaker at CES. They’ve been listening to a lot of audio files to demo their setup and eventually bumped into my recordings. The gentleman in charge (who happens to be an audiophile) called me to say that in all of his years of listening to music, he’s never heard music sound so real, so immediate, and so full of emotion. We’re going to be the sole supplier of audio for their new product introduction.

      CDs have flaws as do vinyl LPs and analog tape…but they are less harmful to audio fidelity. At the end of the day, the chart that was published the other day is fair and accurate. You can choose to believe it not.

      • Thanks for publishing and responding, although my main assertions regarding the non-fidelity of the CD standard are not addressed in that response. I presume the impressed TAS chap was not being played 16-bit/44.1 khz digital audio when he was so enthused in your studio, and this was my point: we’ve had a substandard medium imposed on us (yes, imposed, because it effectively wiped out vinyl production) and we (well, most of us) are only now waking up to that fact and seeking higher standards. Unfortunately, it’s too late for lots of the music, many of whose master tapes have deteriorated and only been backed-up to the inferior 16-bit system of your preference. Ummagumma? Avalon? Gone, and only ‘preserved’ on 16-bit. Same with the first Nick Drake album (although that at least made 24/192). Within the past year, all the Clash masters have gone too. Likewise ‘The Wall’.

        The suggestion that my preferences are just that rather than an interest in fidelity is a disingenuous one, as I don’t care if my music comes on an Edison cylinder or a holographic chip as long as it’s faithful to the source, which no digital format I have yet heard is. Neither is it a question of ‘belief’ — if you think that CD and vinyl issues of the same material from the same master sound merely “different” rather than one being streets ahead of the other then I would suggest again that you are in the wrong business or discipline of inquiry.

        Back in the day I begain copying my vinyl collection to tape using a Nakamichi CR7, pretty much the best tape deck ever made; I stopped after a couple of days as the fidelity was largely lost, and this was my other, equally unaddressed, point — the overwhelming majority of musical material that people want to listen to is on analogue R2R tape, and no system of transcription other than vinyl or R2R copy had yet proved faithful to that system (and certainly not mediocre 16-bit). Noodling in modern studios at 88.1/96/192 khz does not demonstrate anything, as such material has no analogue comparator over which to demonstrate its superiority. This is the nub of it I think — you’re suggesting with your appeals to measured dynamic range and so on that digital is a superior Recording system (I don’t see how that is ever likely to be real-world true when our understanding of fidelity ar any given time is prescribed/limited by theory-driven notions of the day regarding what is a suitable bit-depth and sampling frequency), but what good is that to someone who wants to listen to ‘Led Zepellin IV’ or ‘Piper at the Gates of Dawn’ or ‘Kind of Blue’ or ‘Peter Gabriel III’, all of which reside on high-speed open reel tape (not everyone is satisfied with a musical diet of Alison Krauss)? Ask Neil Young what he thinks about ‘Harvest Moon’ being recprded at 16-bit and listen to the difference between that and ‘Harvest’ on vinyl. His sorrow at a piece of his musical history being forever comprimised by the exponents of digital and their cheerleaders is palpable, and was probably instrumental in his establishment of Pono. I haven’t downloaded any of your 24/96 downloads, but I have a great number of them from other sources (e.g., HDTracks), not one of which holds a candle to the vital, dynamic sound that bursts from vinyl (including the ‘Crime of the Century’ example used by channelD).

        We have the space on our media now (think dual-layer BD, ~ 50 Gb) to encode our consumer audio at 64-bit/758 khz which proabaly would be faithful to (or imperceptably indistinguishable from) ALL primary sources, and that was my last point, implied rather than stated: I am not against digital audio per se, just the piss-poor, chopped-off-at-the-knees implementations we keep getting from the industry whilst our musical heritage continues to disintegrate around us (literally, not just figuratively). Uncritical blogs such as this one only take us further away from ‘perfect sound forever’ (but then we already had ir and we threw it away).

        @Mark: I use Grado SR-80s and AR18s — whose rudimentary paper tweeters make triangles and cymbals sound like real musical instruments when fed analogue vinyl, better than any 16-bit recording played through the Logan CRXs or Magicos of the digital-fans choice. Remember Tiefenbrun’s maxim? Garbage-in, garbage-out…

        • Dave, you need to download the free samples of my tracks. I believe that you will experience digital recordings produced with great care in a format that is accurate and faithful to the original sound coming in through the microphones. There should be more attention paid to the production of great recordings vs. the format on which they are delivered. That is why the reviewers and my customers are so enthralled with my recordings. I subscribe to a different way of making records.

          I would never agree to compromise the clarity and sonic quality of my recordings by cutting them on vinyl LP, although I have been told by other small labels that I could reap tremendous rewards in terms of sales and prestige.

          The Redbook CD specification is more than capable of capturing the natural sound of virtually all commercial recordings…if the production path and mastering is done with care. Better dynamics and plenty of high frequency response…not question. Preference for vinyl versions (folded LF mono, wow and flutter, scrape distortion, limited dynamic range etc) or well done digital at 44.1 kHz/16-bits is a personal statement NOT an absolute. I prefer the accuracy of PCM digital and have established a reasonable reputation among engineers and audiophiles.

          You bring up Neil Young (who I know and have worked with). His Pono initiative is 99.9% rips of 44.1 kHz/16-bit Redbook CDs, which he claims when played on his Pono player “allow you to rediscover the soul music”. How he squares that with his aversion to CD spec audio is rather curious IMHO.

          The downloads that you’ve heard are not high-resolution if they started on analog tape, which most have. The only real HD-Audio is that which was recorded on HD audio equipment when the musicians were present.

          It’s not worth arguing about the relative merits or shortcomings of one format over another until you’ve had a chance to listen to a proper PCM digital recording. Get the FTP credentials, download some tracks (there are comparisons for tape vs 96/24 and CD vs. HD-Audio of exactly the same recording) and have a critical listen. If you open to the idea that digital can actually deliver great fidelity, I think our conversation might change.

          In my 40 years being an audio engineers and in the past 15 years since I started recording in real HD-Audio, I heard from hundreds..even thousands…of audio enthusiasts that changed their mind regarding digital “harshness” and “lack of warmth”.

          Take a listen and let me know what you think. You’ve got the equipment necessary to hear what I’m talking about. I’ve been writing about the world of high-resolution audio and superior sound for a long time…I have no plans to stop.

        • Dave, I have respect for AR speakers, the first HiFi system I put together 47 years ago included AR-2ax speakers, and I am sure that it is easy to hear the difference between a properly produced and a poorly produced cymbal waveform transformed to sound by your AR-18 speakers, but the sound reproduced by the speakers will not include frequencies above 25kHz. The sound waveform that is produced by the AR-18 speakers will look like the CD format spectrogram example in the link you provided.

    • Dave, what speakers or headphones do you use when you listen to that MOFI vinyl album?

Leave a Reply to Sal Cancel reply

Your email address will not be published. Required fields are marked *

twelve − 8 =