Jump to content
  • The Computer Audiophile
    The Computer Audiophile

    My First 24 Hours With MQA

    thumb2.jpg

    1-Pixel.png

    It all started with an email on December 4, 2014. “Hi Chris, It is my great pleasure to provide details on Meridian’s breakthrough technology, MQA (Master Quality Authenticated). The press release is pasted below. And attached is a white paper …” That seems like forever ago. In the ensuing months MQA has been growing like a snowball rolling downhill. More manufacturers getting on board, more content partners signing up, and more chatter within Computer Audiophile community (among others). Based on objective site analytics, I can easily say that since CES 2016 the interest in MQA has grown immensely here on CA. Much of the talk since MQA’s first introduction has been speculative because only a relatively small number of people have actually heard MQA music. Even those who’ve heard it, have likely not heard it in their own audio systems. That was until Meridian officially released the MQA enabling firmware for its Explorer2, Prime, and select components (818v3,*808v6 and Special Edition Loudspeakers) Thursday February 4, 2016. I downloaded the firmware and updated my Explorer2 to v1717. It’s now MQA enabled and I have a DAC that decode and render this content through my own audio system in my own listening room. I’ve been waiting for this forever. I’ve heard MQA at shows plenty of times, but never in my own familiar environment. Now that the hardware was enabled for MQA playback, I needed some MQA music to play. Late afternoon I received an email with a link to download ten MQA FLAC files. Click, save, unzip, play, listen … MQA rules, it’s the best thing since sliced bread. If only it was that cut and dry.[PRBREAK][/PRBREAK]

     

     

    Listening To MQA

     

     

    Like most people, I wanted to listen to a single MQA track and have my mind blown by fidelity I could only dream of prior to MQA. I also wanted to compare MQA versions of tracks to non-MQA versions of the same tracks and come to sweeping conclusions that the MQA version was so much better I would never go back to such unsophisticated non-MQA music again. My list of wants was a bit unrealistic, but my expectations were set at a normal level while I hoped for the best.

     

    In addition to the ten tracks sent to me this afternoon, I purchased some content directly from the 2L record label’s website. This enabled me to purchase both the MQA and non-MQA versions of the same music. What could be more telling than two versions of the same thing? Or, so I thought.

     

    First up on my list to listen to was Stille lys (Quiet Light) by Jan Gunnar Hoff (link). I received the MQA version of track one titled Mitt Hjerte Alltid Vanker and I purchased the 24 bit / 192 kHz download of the same track. According to 2L the album was produced in DXD (Digital eXtreme Definition 352.8kHz/24bit). I would have downloaded the original DXD version but the Explorer2 doesn’t support sample rates over 192 kHz. The MQA version of the track appears in Roon as a 24/44.1 track because Roon sees the file like a DAC without an MQA decoder. Fortunately Roon, or any other application, simply needs to send the audio out to the DAC bit perfectly (unchanged) so an MQA enabled DAC can unfold the file into a higher resolution if needed. While playing this track through the Explorer2, the MQA light illuminates blue and the 4x sample rate lights are also illuminated. The LED lights up blue to indicate an MQA Studio file is playing. MQA Studio files are artist/producer-approved studio releases.

     

    Prior to this afternoon I had never heard this album at a show or in my own system. I would have preferred listening to music I am very familiar with, but at this point we have to take what we can get. I started with the MQA version of Mitt Hjerte Alltid Vanker and played it through three times. It sounded wonderful. Right from the beginning I noticed a clarity to the sound of each note as the hammers struck the strings and a superb decay as the tone faded into a black background. It really is a stellar sounding piece of music in all its MQA glory. That said, the 24/192 version of this track is also terrific. The main differences between the two versions of this track are 1) The MQA version has an uncanny clarity and sense of space around each individual note that is just not present in the standard 24/192 version. This space is specifically around each note, not necessarily presented as a larger or more airy soundstage as a whole. 2) The 24/192 version sounded like the microphone was closer to the strings and the sound was more narrow as if each note was compartmentalized its own silo. 3) On the MQA version, the tone of the decay of each note has a purity to it or an appropriate color to it that isn’t present in the standard version. I really noticed this sense of hearing the entire note, from the initial hammer strike to the last decibel of the decay, in all its glory.

     

    I’m not into hyperbole or writing something with which I am unsure. Thus, I gave myself a blind ABX test by putting the two versions of this track into a playlist, listening to them back to back, then setting the queue on repeat and random and pressing the next button several times without looking. I did this several times and immediately selected the correct MQA or non-MQA version of the track every time. Readers should keep in mind that just because I immediately picked the correct version of the track, doesn’t mean the differences are night and day. These things are subtle. But, once heard it’s hard not to hear the differences.

     

     

    Up next was the album Ein Song Frå Dei Utsungne Stunder by Berit Opheim, Nils Økland & Bjørn Kjellemyr, also known as The BNB (link). This album was originally produced at 16 bit / 44.1 kHz by 2L. Playback through the Meridian Explorer2 illuminated the MQA light in blue and didn’t light up the 2x or 4x LEDs. This MQA album remains at the same resolution seen by Roon, 16/44.1. The Explorer2 internally upsamples the audio to 4x (176.4) but that’s a topic for another time. This entire album sounds fantastic. Great vocals and great double bass accented by a sweet fiddle and viola. I noticed two subtle differences between the original and MQA versions of this album. 1) The original non-MQA version contained what I’ll call a plastic edge to the sound of some instruments. There was something synthetic about the sound that likely can’t be heard unless one has the MQA version for comparison. 2) The non-MQA version has a darkness or dullness to it that isn’t present in the MQA version. This isn’t darkness associated with the blackest of backgrounds or a low noise floor, rather its a deadness that’s heard with the sounds of the instruments. As with the previous album, the differences are not equivalent to bumping the volume by a few dB. They are subtle and may not be apparent all listeners, especially when listening to unfamiliar music.

     

     

    Switching to music that I am a bit more familiar with, I listened to a track titled When I Go from Judy Collins’ album Strangers Again. On this track Judy duets with Willie Nelson. Roon sees the track as 24/44.1 while the Explorer2 DAC sees it as 2x (most likely 88.2 as that’s the resolution of the HD version available from HDtracks and others (link)). The Explorer2 also illuminated the first LED as green rather than blue. Blue is the MQA Studio color, but green indicates that the unit is decoding and playing an MQA stream or file, and that the sound is identical to that encoded. I am not 100% sure what this means in terms of the MQA process to turn the music into an MQA album from a standard high resolution album. For all I know it may mean that the album was converted to MQA for its smaller file size, without much of the wizardry that goes into the MQA white glove process of creating MQA Studio files. Don’t quote me on that, it’s just a wild guess. (see edit 2 below) Perhaps that wild guess has something to do with the very small sonic differences I heard on this Judy Collins / Willie Nelson track. I thought if there was one track, out of the ten I received, in which I would really notice a difference, it would be this track. Most of us have heard Willie Nelson a million times and are familiar with folk music (more so than classical for many people). After listening over and over to the MQA and the original high resolution versions of this track I think the only noticeable difference I hear is a touch more natural or appropriately soft sound in Willie’s voice. On second thought, I believe there is also a difference in the sound of the opening drums. (I literally went back and listened a few more times). The MQA version of the track seems to reproduce more of the drum’s frequencies or make more of the drum audible. It’s not that the drum has a super wide frequency response, rather the non-MQA version seems to lose some of the drum sound into the background. The MQA version seems to reproduce a fuller drum sound with better decay than the non-MQA version. Either way, this track was a tough one for me as I struggled to hear the differences I wanted to and I thought I would hear.

     

    Edit 1: I just received a quote from Alan Silverman, Mastering Engineer on the Judy Collins track When I Go:

     

    “We have done many blind comparisons of my original high-resolution masters with and without the MQA process. MQA is the consistent winner. What mystifies me about the technology is the purity of tone and natural realism that MQA unlocks from my high-resolution recordings. The MQA playback is more satisfying and not by just a subtle shade. MQA has educated my ear to digital artifacts that still exist, in spite of the best practices with the best equipment, by eliminating them. It is perhaps a holy grail of digital audio.”

     

    More specifically about the track When I go Alan said, "I’ve just compared the MQA playback with my original 88.2k 24-bit master and find the MQA to be mystifyingly more satisfying, and not by just a subtle shade. Listening to Willie and Judy, their voices sound much more real, at the same time, they have a textural filigree and detail of tone that I am not hearing in the original master! The same holds for the banjo and the subtle electric guitar in the right channel. I am delighted and extremely enthusiastic about the MQA process.”

     

     

    Edit 2: This just in from MQA ltd., "There is no sonic difference between files marked as green or blue, it is only about Provenance or Approval." In addition, "Today Alan Silverman asked us to move the Judy Collins [album] up to Studio."

     

     

    Wrapping Up The First 24 Hours

     

     

    Overall I am happy with the MQA music I’ve heard. I wish I could render an opinion, that would carry across all MQA products and music, that MQA is always better by a wide margin, but this isn’t the case. The differences I’ve heard so far are subtle and my opinions are limited to the music and hardware I used in the last 24 hours. I also have a suspicion that the MQA process will be more beneficial to recordings that were done under less than stellar circumstances (i.e. lesser quality A to D converters, etc…). The 2L recordings are done with the utmost care using very good equipment and very good engineers. While there is still improvements MQA has made to the original 2L masters, I’m willing to bet there are greater improvements to be made to more traditional popular recordings or very old recordings. On the other hand, it may not be easy to compare an MQA version and non-MQA version of some old recordings because the MQA version has been done with the white glove process. It would be the same as comparing two difference masters of the same album, of course they’ll sound different. There will be clear differences with or without MQA. The real question many people will want answered is, how much of the difference is MQA and how much is the white glove process? But, does this question really need to be answered? I’m not so sure because we don’t have the option of getting new white glove masters of some of our favorite music. If MQA is the impetus to get us better sounding music, that’s all that really matters. In a dream world we may have the option of a white glove MQA and white glove non-MQA, but this is the real world. The options are, MQA or live with what we already have. Anyway, the MQA train is finally leaving the building. I’m cautiously optimistic that everything will work out and we’ll have better sounding music without too much trouble.

     

     

     

     

     

     

     

    System I used for playback:

     

    Roon software running on SOtM sMS-1000SQ Windows Edition server and SOtM sPS-1000 power supply > Wireworld Platinum Starlight 7 USB 2.0 Cable > Meridian Explorer2 DAC > AudioQuest Yosemite 3.5mm to RCA Cable > Constellation Audio PreAmp 1.0 > Wireworld Platinum Eclipse 7 Interconnects > Constellation Audio Mono 1.0 Amplifiers > Wireworld Platinum Eclipse 7 Speaker Cables > TAD CR1 Loudspeakers.

     

     

     

    1-Pixel.png




    User Feedback

    Recommended Comments



    But where is todays MQA SW decoding happening ? In Roon or in the Tidal PC controller app ?

     

     

    Sent from my iPad using Computer Audiophile

    Share this comment


    Link to comment
    Share on other sites

    But where is todays MQA SW decoding happening ? In Roon or in the Tidal PC controller app ?

    Today: either in the Tidal PC/mac app or in the DAC if you choose passthrough. Roon currently does NOT decode MQA (supposedly it will).

    Share this comment


    Link to comment
    Share on other sites

    Nonsense!

     

    For all intents and purposes, think of MQA as a way to zip a high res file into a smaller package by "throwing a few things that cannot be heard out". If you don't decode the MQA information you get something very similar to a CD. The additional information is encoded in the least significant 8 bits - MQA samples are 24 bit, CD is 16 bit, so you have 8 extra bits. The software decoder is able to make sense of those extra 8 bits and produce the higher frequency information, giving you a high res PCM stream.

     

     

    BTW... I auditioned MQA at Merdian in NYC in March 2015. The files they played sounded markedly better in MQA form. The most amazing of those was one that was 24/192 originally and MQA'd to 24/192. I would deem this demo not very clear or relevant, as it was impossible to tell whether the MQA versions sounded better due to MQA or to careful remastering from the original source. I doubt that the effect was just the time deblurring, it was too big a difference. Ever since I have been very suspicious about any MQA claims (as people might recall from my other posts).

     

     

    However, I am a very pragmatic person, and if Tidal ends up streaming in MQA and I enjoy that sound better, by all means gimme MQA (just don't ask me to buy a shitty DAC).

     

    I would think de-blurring is the only benefit of MQA. And can't be done without an MQA DAC. The rest of it is just a lossy compression for nothing else but to reduce bandwidth required for streaming.

    Share this comment


    Link to comment
    Share on other sites

    I got Tidal MQA streaming to the Explorer2... set both Exclusive and Passthrough...

     

    638944e7f1e0ac2ed2755f068f307c450bb97250_1_660x500.jpg

    Share this comment


    Link to comment
    Share on other sites

    I would think de-blurring is the only benefit of MQA. And can't be done without an MQA DAC. The rest of it is just a lossy compression for nothing else but to reduce bandwidth required for streaming.

    You're wrong. Read what I wrote.

    Share this comment


    Link to comment
    Share on other sites

    But where is todays MQA SW decoding happening ? In Roon or in the Tidal PC controller app

     

    Seems that you can play the MASTER albums on both OSX and Windows desktop Tidal apps.

     

    If the Passthrough is selected, you have to have a MQA capable DAC.

     

    No MQA decoding occurs in Roon. Tidal just passes the MQA file through to Roon and the DAC has to handle the decoding.

     

    I cannot detect any difference playing Moondance through Tidal or through Roon... happy so far.

    Share this comment


    Link to comment
    Share on other sites

    You're wrong. Read what I wrote.

     

    What you wrote is your opinion. I've actually read the MQA white paper! And I've read comments by experienced reviewers who said they couldn't reliably pick MQA over non-MQA files in a blind test. And that's with an MQA DAC. So how is software decoding gonna be even equal to the file before the lossy compression? Really? De-blurring is the key to the entire process and it's wrapped up in a package with the compression scheme so it sells better.

     

    I agree with the guys at Schiit. Show me a demonstrable improvement?

    Share this comment


    Link to comment
    Share on other sites

    What you wrote is your opinion. I've actually read the MQA white paper! And I've read comments by experienced reviewers who said they couldn't reliably pick MQA over non-MQA files in a blind test. And that's with an MQA DAC. So how is software decoding gonna be even equal to the file before the lossy compression? Really? De-blurring is the key to the entire process and it's wrapped up in a package with the compression scheme so it sells better.

     

    I agree with the guys at Schiit. Show me a demonstrable improvement?

     

    It's up to you to do your own listening test. Trust your ears.

    Share this comment


    Link to comment
    Share on other sites

    What you wrote is your opinion.

    No sweetie, it is not.

     

     

     

    MQA starts from a high resolution master, call it M0, applies the famous "time deblurring" to "correct for ADC artifacts" getting another high resolution file, call it M1, which is then "folded" into the MQA package as I explained previously (and is otherwise detailed in the paper). The resulting file, call it M2, is a 24bit/48KHz file where the most significant 16 bits of every sample represent the equivalent most significant 16 bits of the M1 file's equivalent sample rate of 48KHz (hence you get the initial deblurring benefit) and the 8 least significant bits are used to encode the information above the 48KHz sample frequency, with a "recipe" that only MQA knows. What the software decoder does is "unfold" the M2 file into a file that is "close" to the M1 file save for some lost information that Bob Stuart argues cannot be heard. So you get something with more information than 16bit/48KHz by using the decoder.

     

     

     

    If you instead where to play the M2 file undecoded, the least significant 8 bits would average out as noise, losing the information encoded in them. This is also the reason the file does not compress to an equivalent 16bit/48KHz size FLAC: there is information in those 8 bits.

    Share this comment


    Link to comment
    Share on other sites




    Guest
    This is now closed for further comments




×
×
  • Create New...