MQA bails on Rocky Mountain Audio Fest*

Discussion in 'Audio Hardware' started by ls35a, Oct 7, 2017.

Thread Status:
Not open for further replies.
  1. Archimago

    Archimago Forum Resident

    Hmmm not so sure about that... Wouldn't be too difficult to compare the DSD --> hi-res PCM and 16/44 PCM waveforms if one really cared. If it's audible, you'll for sure see the difference in the data...
     
  2. Rolltide

    Rolltide Forum Resident

    Location:
    Vallejo, CA
    I've always felt analyzing waveforms would be able to tell all differences, including those from cables, etc. The problem is, even if this is true, the subjectivist get-out-of-jail-free card of "...well then there's just more to the story" would still get played.
     
    Tim Müller, LarryP and dchang81 like this.
  3. Steve Hoffman

    Steve Hoffman Your host Your Host

    Anyone tried my little test yet? So easy to do and yet not a word.
     
    Metralla likes this.
  4. Claude Benshaul

    Claude Benshaul Forum Resident

    Exactly, mobile bandwidth isn't a problem anymore and it wasn't such a big one to start with. OTOH, HiFi heaven on the move is often an exercise in futility which is dictated by the environment. MQA isn't going to do much to improve car audio for example, if only because I believe that the problem there has nothing to do with bit resolution or sampling frequency.

    At home it's the exact opposite. In a perfectly personalized environment you get the revelation that the source is often the problem and, again, it has nothing to do with bit depth or sampling frequency. Will MQA technology be able to overcome a studio decision to release a master that is just good enough for Gov. purposes? Somehow I doubt it will be able to eliminate even the most trivial defect. Now if MQA is able to extend it's temporal filter to reach out in the past with a really big hand and squeeze the life out of the idiots that decided to brick-wall an album or that leaving a slight hiss on one channel is cost effective, now that's something that will color me impressed!

    Right now I really struggle to find a single scenario where I'll lean back, think about it and say to myself that they really have something interesting going there.
     
  5. markc2

    markc2 Forum Resident

    Increasing the amount of information gets you closer to the true analog form.

    Increasing detail in the picture is made by making more pixels per given space. That to me is kind of like bit depth.


    "digital audio using pulse-code modulation (PCM), bitdepth is the number of bits of information in each sample, and it directly corresponds to the resolution of each sample."

    FIG.1: If the bit depth is low (a), the signal will be inaccurately converted because it’s sampled in large increments. By increasing the bit depth (b), you get finer increments and a more accurate representation of the signal.

    FIG. 2: Increasing the sample rate but not the bit depth (a) improves the accuracy of the representation because the converter is taking “snapshots” of the signal more frequently. However, increasing both the sample rate and the bit depth (b) produces much more accurate results.


    fig 1.............................................................................................................fig 2
    [​IMG] [​IMG]
     
    Tim Müller likes this.
  6. Juan Matus

    Juan Matus Reformed Audiophile

    My take is they messed something up in the cd transfer part of the process. It's an old analogue non hi rez recording. Regular cd is perfectlty capable of capturing what you are talking about. At least in theory hence my conclusion it's a production snafu not a format issue. If it was a digital hi rez recording that would be another story.
     
    Tim Müller likes this.
  7. Ham Sandwich

    Ham Sandwich Senior Member

    Location:
    Sherwood, OR, USA
    Just for clarification. This Creedence Clearwater Revival album was mastered using a split feed. One feed for DSD using a DSD converter and one feed for PCM CD using a PCM converter?
     
    Tim Müller likes this.
  8. missan

    missan Forum Resident

    Location:
    Stockholm
    There are usually reasons for us writing what we do.
     
    Tim Müller likes this.
  9. Higlander

    Higlander Well-Known Member

    Location:
    Florida, Central
    Perhaps you are confusing "Been able" to "Actually bothered (or attempted) to even try"
    Anyone with decent test equipment can easily show this stuff. The fact that many or most on the forum do not possess such equipment, nor are professional audio engineers, does not mean a lot when you say "No one has been able".......
     
    Tim Müller and rbbert like this.
  10. harby

    harby Forum Resident

    Location:
    Portland, OR, USA
    Your pictures are the kind of misdirection used in marketing, and are incorrect. The "inaccurate conversion" of bit depth only results in quantization noise. Noise like is created by the inaccurate transcription of audio signals on magnetic tape by the random organization of metal domains, except magnitudes less.

    There are no "pixels", there are no stairsteps, and spectral content is reproduced accurately up to the Nyquist frequency by PCM. Your ear just doesn't have receptors to respond to frequencies or harmonics higher than redbook audio's capabilities - just like it wouldn't matter to your eyes if you had a TV with accurate infrared and ultraviolet reproduction (unless you want to feel warm and get a tan from pictures of the sun on your TV).

    There CAN'T be stairsteps if you understand Fourier analysis of waveforms and how any waveform is composed of a series of harmonics; in fact a square wave is an audio impossibility. Your speaker can't instantaneously move from one position to another and create an infinite amount of air compression.

    For example we take a 12KHz sawtooth waveform depicted below:
    [​IMG]
    as we add more and more harmonics to a fundamental sine wave (red), the output starts to look like the square wave we imagine or computer-generate. However, at 12KHz, any harmonics are ultrasonic and inaudible, and so we are only digitizing and reproducing the fundamental sine wave, and your ear can't tell the difference.



    Anyway, that distracts from the fact that MQA is a pointless scam, that neither reduces the bitrate at a given quality any more than a lossless compression method, nor sounds "better" than the barely larger 96/24 files or even 44/16 without adding audio filters and bunk.
     
    Last edited: Oct 14, 2017
  11. cdash99

    cdash99 Senior Member

    Location:
    Mass
    Put another way, MQA should have been sold as an OEM product rather than the greatest listening enhancement since the invention of the ear.

    This is similar to my comment in another thread, indicating that MQA would have been viewed more favorably if it was positioned as an upgrade over mp3, ACC, OGG, etc, rather than as a new expensive format.
     
  12. Higlander

    Higlander Well-Known Member

    Location:
    Florida, Central
    Except it does not actually work that way.
    Bit depth=low limit/Noise level
    Sampling Rate=Upper frequency limit.

    Sampling a sine wave "More" does not result in a more "Detailed sine wave", but does allow for higher frequencies, that the wave is smaller and closer together.
     
    Tim Müller, dchang81 and basie-fan like this.
  13. Agitater

    Agitater Forum Resident

    Location:
    Toronto
    Yes. I have. Repeatedly. With MQA and 16/44 files from identical sources on 2L, among other places. Accurately identifying the MQA source stream was no better than guessing/coin flips. Member @Archimago also did extensive testing using identical source material. So have others, independently, who ended up with the same results - sometimes differences, rarely improvements, frequently nothng of note at all.

    Comparing identical sources between CD and 24/192 and DSD is another matter entirely. As you posted earlier in the thread, improvements are there and audible, but subtle. The very slow uptake of hi-res by audiophile musc lovers over the past 15 years though, seems to indicate that the subtle improvements aren’t enough to fuel large numbers of purchases by those consumers sufficient to make hi-res the standard of the future so far.

    The problem for hi-res is that it’s only a bit better than a solution in search of a problem. That’s based on the fact, IME, that great CD-quality mastering done by you and many others for over 30 years is not to be denied. I have hundreds and hundreds of great CDs and LPs in my still-growing collection, and a few dozen hi-res files that offer further, subtle improvements. Basically, the subtle improvements can’t justify the expenditure of a couple of hundred thousand dollars to repurchase all the great stuff I’ve already got. I’m already enjoying my music collection too much. I buy the CDs and LPs I want to have. Some of them, most likely, will never be offered in hi-res because the market for them isn’t large enough to jusitfy production cost. I can find most of them through my Tidal HiFi account though, which is great!

    I make new purchases in hi-res too - mainly jazz and classical. Got to look up the liner notes online, unlike good CD and good LP releases which come with liner notes. So less is more in some aspects of hi-res, but subtle sound improvements are definitely there. The longer and more detailed decay that is sometimes audible is also sometimes unnatural in that such long detailed decay on certain phrases aren’t audible when attending a live performance of the material. So such wholly audible decay on a hi-res mastering sometimes feels like a studio special effect rather than a part of the music intended by the musicians.

    None of that alters the fact that MQA remains a scam on consumers. And maybe that’s one of the reasons that MQA bailed on RMAF?

    BTW, at TAVES 2017 in Toronto (on all weekend at the Toronto Congress Centre), I could not find a single mention or demo of MQA. Spent almost 7 hours at the show yesterday (Friday, October 13) - not one mention or demo or marketing poster or sign about MQA. Maybe I just missed it? I didn’t check the lecture and discussion tracks, so maybe there was something about MQA in the schedule.
     
    Last edited: Oct 14, 2017
  14. Dreams266

    Dreams266 Forum Resident

    Location:
    NJ
    I have the AP Willy and Poor Boys on vinyl only so can't do the A/B with the SACD. It's one of the best remaster I've ever heard btw on vinyl
     
    Tim Müller likes this.
  15. RandyHat

    RandyHat Senior Member

    Location:
    Denton, Texas
    Too much analyzing waveforms and too little listening. This reminds me of my other hobby, photography. Guys taking hi-rez photos of brick walls and zooming in to "analyze" the pixels. Breaking down the photo into it's most basic component parts but failing to see the picture. I am NOT an MQA proponent but I do have a modest streamer that allows me to stream MQA from TIDAL. Sorry I don't have a chart or graph to post but only the test results obtained from a pair of un-reliable ears. Some MQA sounds great. Some not so much. Kinda like the results I get from listening to my CDs and LPs. I like having these choices.
     
  16. Claude Benshaul

    Claude Benshaul Forum Resident

    A little bit off topic: Years ago I was involved in the periphery (practically the sideline) of an issue concerning the next standard after VGA. HD was right around the corner, the US component and the EU RGB SCART standards were unsatisfactory and the decision was to select a new interface and cabling solution for HD TV. At the time there were two contenders left: HDMI and SDI. HDMI was more complex, more prone to interference, (still) had a ridiculously delicate connector and was difficult to extend beyond 3 meters. SDI suffered none of these limitations and used a single robust coaxial cable with a bulletproof connector. HDMI won due to the pressure from broadcast groups that saw the DRM (sorry, HDCP) "feature" as a lifesaver against piracy.

    Perhaps that's what MQA Ltd hope to achieve - push for the adoption of MQA through the big studio and force feed this standard globally. Their problem is that we are not living in 2002 anymore and the public as a whole is no longer depended on the few anointed self proclaimed experts (AKA sock puppets) or a selected group of content providers.
     
    jfeldt, Tim Müller, gd0 and 3 others like this.
  17. missan

    missan Forum Resident

    Location:
    Stockholm
    No "we" are all of us who write here. You also have some reasons for writing like You do, but I'm not part of the reason. And for the records, I won't tolerate personal attacks.
     
    Tim Müller and Dreams266 like this.
  18. basie-fan

    basie-fan Forum Resident

    Your experience seems to contradict the AES study I cited earlier in this thread. The study looks to be done properly but maybe there's some flaw that's not obvious. What's your opinion of this paper? Did they get it wrong?

    An interesting experiment would be to downsample that SACD you mentioned to 16/44 and then see if blinded listeners can hear a difference between these two files (and also the CD layer.)
     
    Tim Müller likes this.
  19. No Static

    No Static Gain Rider

    Location:
    Heart of Dixie
    Wow. Welcome to the Forum.
    :wave:
     
    Last edited: Oct 14, 2017
    Tim Müller and Dreams266 like this.
  20. Steve Hoffman

    Steve Hoffman Your host Your Host

    Correct.
     
    Carl Swanson likes this.
  21. Steve Hoffman

    Steve Hoffman Your host Your Host

    I
    I truly give up.
     
  22. McLover

    McLover Senior Member

    DSD preserves more of the low level details. The room sound, the air and space around the instruments. 16 bit doesn't quite get the last level of refinement. This is why I prefer SACD for digital when digital is done. It preserves more musical nuance and lower level details, and keeps it sounding more like music. Steve and Kevin's "Fortunate Son" sounds best on LP and SACD, the RedBook 16 bit layer on the SACD is a third place finish.
     
    Tim Müller, Guss2 and Rolltide like this.
  23. testikoff

    testikoff Seasoned n00b

    I did something a tad different. Here is the result of my unsuccessful ABX test from a friend's DSD rip (converted to 24/88 using foo_input_sacd plug-in with FIR30 filter) of Fortunate Son track & my own down-conversion of it to 16/44 using fb2k with MDA VST dither (moderate noise-shaped) & then upsampled to 24/88 (both SRCs were done using SoX high-quality linear-phase algorithm):

    Code:
    foo_abx 2.0.4 report
    foobar2000 v1.3.16
    2017-10-14 12:05:28
    
    File A: 06 - Fortunate Son (2488).flac
    SHA1: ca3358455d78b33bef66c0a9a09cb9234a752840
    File B: 06 - Fortunate Son (1644-2488).flac
    SHA1: 550d584f3ae2dcd8b3762e19174e20e07174ec68
    
    Output:
    ASIO : ASIO4ALL v2
    Crossfading: NO
    
    12:05:28 : Test started.
    12:08:28 : 01/01
    12:08:38 : 01/02
    12:08:46 : 02/03
    12:08:55 : 03/04
    12:09:03 : 04/05
    12:09:20 : 04/06
    12:09:40 : 05/07
    12:09:55 : 05/08
    12:10:06 : 05/09
    12:10:18 : 06/10
    12:10:18 : Test finished.
    
     ----------
    Total: 6/10
    Probability that you were guessing: 37.7%
    
     -- signature --
    4fceeaae4d6b314f5fe7c81c5df4cca196338594
    

    Here the audio stats of the delta of 2 tracks:

    Code:
    Creedence Clearwater Revival - Willy And The Poor Boys [DSD]
    
    Peaks
    =====
    1.1
    
    Peak Table
    ==========
     1|  1.1
    
    Full Statistics
    ===============
     #|Chan|Peak %|  Peak dB|AvgRms dB|TotRms dB|MinRms dB|MaxRms dB| MinSample| MaxSample| DCOffset
    --+----+------+---------+---------+---------+---------+---------+----------+----------+---------
     1|   1|   0.3| -49.7827| -70.6255| -69.8878| -78.2520| -59.4832|-0.0031841| 0.0032424|  -0.0000
      |   2|   1.1| -39.0352| -67.6995| -66.2779| -75.7970| -54.2963|-0.0095438| 0.0111748|  -0.0000
    
    0 dBFS: Sine
    RMS Window Size: 50 milliseconds
    

    I found the difference in sound of 24/88 master & its 16/44 RBCD downcoversion very subtle & not that easy to reliably discern in a blind level-matched comparison...
     
    Tim Müller and basie-fan like this.
  24. No Static

    No Static Gain Rider

    Location:
    Heart of Dixie
    "Never give up. Never surrender."
     
    basie-fan and Rolltide like this.
  25. Rolltide

    Rolltide Forum Resident

    Location:
    Vallejo, CA
    The only Steve Hoffman SACD I have (to my knowledge) is the Donovan Storyteller disc, which is an absolutely incredible remaster. I have no idea how Steve made mid-60's folk music sound that good. It's never really occurred to me to downsample it and do an ABX comparison, because that would take a lot of time and be boring as hell - I'd rather just listen to it in the format I bought it.

    This is sort of my thoughts on High Res PCM and DSD in general. Are the improvements subtle? Maybe, but I have the hardware to play them properly and enough cheap disk to store them, so what is actually the counter-argument here? That I should stick to 44/16 as an article of faith? Heck, if the only thing I was doing was paying $3 or so more for the convenience of a high res download vs. ordering a CD, waiting for it to arrive, and ripping and tagging it, that would be fine with me.

    MQA is a different story for me though, as it's not claiming to be a better mousetrap, just a different mousetrap.
     
    Sevoflurane, Guss2 and No Static like this.
Thread Status:
Not open for further replies.

Share This Page

molar-endocrine