Is MQA a Friend of the Devil? A sonic evaluation...

[Jud]

The DVD-A was Mickey Hart's mix, I think, so if the MQA master is supposed to be "as the artist intended...."

 

This provenance stuff gets complicated.

 

 

Sent from my iPhone using Computer Audiophile

[bobbmd]

Jud-interesting where do you get your info? I never delved that deeply into the esoterics of Dead recordings-by the way 'delve' is according to dr Jerome Berrigan(of the famous anti-war Berrigan brothers) and my 11th grade advanced placement english teacher a 'semi archaic' expression which I used in the opening lines of my thesis on Samuel Taylor Coleridge and the Rime of the Ancient Mariner!

I feel the DVD-A recordings of both American Beauty and Workingman's Dead to be their best recordings and that MQA of American Beauty approximates those.Just can't believe YOUR knowledge of how these recordings were done!

[sdolezalek]

Ok, after a day of excitement and hearing clear differences between the Master and regular 16/44 versions, I'm now strongly leaning to preferring my upsampled-to-DSD128 versions of certainly 24/96 starting material, and maybe even 16/44 Tidal starting material, to Master MQA versions of the same albums through Tidal.

 

Particularly on Classical music, and individual string instruments, the MQA versions have a harshness that the DSD128 upsampled versions don't have. The latter has a silken sheen, the former much more of an in your face brightness that just doesn't convey the same level of "realness" in my system.

 

Admittedly, I'm just using the software decoding fed directly into the DAC in my Bryston SP3, so YMMV for in-DAC decoding. It will be interesting to hear other opinions develop over the next few weeks. I also haven't been able to try upsampling the MQA decoded files through HQPlayer into DSD128 (that will require Roon 1.3 for me to do it), so my comparison is not a true apples-to-apples comparison.

[ckpiv]

Ok, after a day of excitement and hearing clear differences between the Master and regular 16/44 versions, I'm now strongly leaning to preferring my upsampled-to-DSD128 versions of certainly 24/96 starting material, and maybe even 16/44 Tidal starting material, to Master MQA versions of the same albums through Tidal.

 

Particularly on Classical music, and individual string instruments, the MQA versions have a harshness that the DSD128 upsampled versions don't have. The latter has a silken sheen, the former much more of an in your face brightness that just doesn't convey the same level of "realness" in my system.

 

Admittedly, I'm just using the software decoding fed directly into the DAC in my Bryston SP3, so YMMV for in-DAC decoding. It will be interesting to hear other opinions develop over the next few weeks. I also haven't been able to try upsampling the MQA decoded files through HQPlayer into DSD128 (that will require Roon 1.3 for me to do it), so my comparison is not a true apples-to-apples comparison.

 

I think I am coming to the same conclusion. The MQA played through the Tidal app do sound very good but seem to be fatiguing after awhile. I keep going back to my 16/44, 24/88, 24/96, 24/176, 24/192 up sampled to DSD128 or 256 with HQPlayer. The sound is very smooth and not fatiguing, where I just want to keep listening. Upsampling the Tidal MQA non-decoded to DSD from Roon does not do as well so like you I am waiting on the Roon update to see how they do then. For now I find up sampling the MQA to 768pcm with the HQplayer Poly MQA filter sounds better then DSD.

[Jud]

Jud-interesting where do you get your info? I never delved that deeply into the esoterics of Dead recordings-by the way 'delve' is according to dr Jerome Berrigan(of the famous anti-war Berrigan brothers) and my 11th grade advanced placement english teacher a 'semi archaic' expression which I used in the opening lines of my thesis on Samuel Taylor Coleridge and the Rime of the Ancient Mariner!

I feel the DVD-A recordings of both American Beauty and Workingman's Dead to be their best recordings and that MQA of American Beauty approximates those.Just can't believe YOUR knowledge of how these recordings were done!

 

Got the info from other members here.

 

For confirmation, see for example Blair’s Golden Road Blog—New Twists on Three Classic Albums | Grateful Dead :

 

Right off the bat, I should mention that the hi-res Workingman's Dead (which is out) and American Beauty (coming November 6) are not precisely the albums that 99 percent of you are familiar with. You may recall that back in 2001, when DVD surround recordings looked like they might be the Next Big Thing (and an important new revenue stream for record companies), Warner Bros. hired Mickey Hart and his engineer in that era—the great Tom Flye—to go back to the 16-track master tapes of those two albums and remix them in 5.1 (front left-center-right, rear left and right, plus a subwoofer channel). It is stereo versions of those remixes—which all but vanished from the planet when the Warners-favored DVD-A format did not succeed—that are now being offered as downloads through HDtracks.

[The Computer Audiophile]

While I understand the desire to compare and pick a winner, all bets are off when we don't have identical masterings and matched levels. It's a disservice to suggest the difference in sound comes from MQA, when we have no clue if the difference is solely related to differences in masters.

[crenca]

While I understand the desire to compare and pick a winner, all bets are off when we don't have identical masterings and matched levels. It's a disservice to suggest the difference in sound comes from MQA, when we have no clue if the difference is solely related to differences in masters.

 

Not sure I follow. A "disservice" to what, or whom, exactly? Was not the "difference in masters" exactly the sort of problem that MQA is supposed to solve - "what the artists intends" and "end to end" and all that? It sounds like you are saying that by putting a critical ear to the MQA that we have (Tidal right now) that we are being somehow "unfair" to MQA. If so, how and when exactly do you think think it will be fair or time to pick a winner?

 

Perhaps as an "insider" you can ask Tidal/MQA/Bob which albums do have a consistent provenance so that we can be confident in our "apple to apple" comparisons. One example might very well be Bob James & Nathan East's "The New Cool" since it is just over a year old and thus probably without any variance in master/mix between the Tidal 16/44, Tidal MQA, and the 24/48 Hi Res download...probably...

 

p.s. the level thing is easy to deal with (all else being more or less equal - a real question here)...

[The Computer Audiophile]

Not sure I follow. A "disservice" to what, or whom, exactly? Was not the "difference in masters" exactly the sort of problem that MQA is supposed to solve - "what the artists intends" and "end to end" and all that? It sounds like you are saying that by putting a critical ear to the MQA that we have (Tidal right now) that we are being somehow "unfair" to MQA. If so, how and when exactly do you think think it will be fair or time to pick a winner?

 

Perhaps as an "insider" you can ask Tidal/MQA/Bob which albums do have a consistent provenance so that we can be confident in our "apple to apple" comparisons. One example might very well be Bob James & Nathan East's "The New Cool" since it is just over a year old and thus probably without any variance in master/mix between the Tidal 16/44, Tidal MQA, and the 24/48 Hi Res download...probably...

 

p.s. the level thing is easy to deal with (all else being more or less equal - a real question here)...

 

The only thing I'm saying is this - You're comparing Apples to Oranges. You have no idea if the MQA version sounds different because of MQA or because of different mastering. Thus, there's no way to make a technology judgment. The only conclusion supported by your listening experience is that two versions sound different.

 

It's possible this is a new master and MQA made it sound much worse or much better than the straight PCM. We have no clue.

[sdolezalek]

While I understand the desire to compare and pick a winner, all bets are off when we don't have identical masterings and matched levels. It's a disservice to suggest the difference in sound comes from MQA, when we have no clue if the difference is solely related to differences in masters.

 

Chris: I fully agree with your position that it is too early to suggest that sound differences are a result of MQA. I am hearing some things fairly consistently across a number of the MQA Masters, but those could easily just be be different mastering/editing choices (such as generally higher levels of bass) and may not be an artifact of the MQA process per se. Others have also mentioned greater separation of certain instruments and voices that seem to increase clarity but seemingly also grain. Again, these may be conscious choices made by Meridian and not a function of the coding/decoding process. The result may be that there is no true apples-to-apples comparison because two, not just one (MQA), changes are being made as compared to the prior Tidal 16/44 file.

[abrxx]

Agreeing with Chris and others here. I don't see how we can assess MQA by listening to Tidal Masters. Its clear many of the masters chosen differ from the regular flac master. For some songs the entire mix seems.different even the track length is different. Its almost hopeless.

 

However we can as consumers say whether we prefer listening to the Tidal master or existing redbook, at least while Tidal maintains duplicates.

 

To my ears with my fairly basic stereo at home, the Tidal masters sound fantastic. The MQA tech may contribute little to the final sound. But without MQA Tidal would have not have introduced its Master quality. Even if MQA contributed 0% change to audio quality, I still get to listen to remastered hi rez albums that sound fantastic. That can only be a good thing.

[One and a half]

First off, I cancelled my subscription to Tidal two weeks ago and do not have a MQA compliant DAC, these observations are from the 'outside', so comments are all 1/2c worth.

 

- Due to bandwidth restrictions, >44.1kHz files are worth to MQA encode, simply due to the bitrate, OK, I understand to a point. DSD128 (perhaps more) has experimentally being transmitted, and this has thousands of b/s but does not require MQA encoding.

 

- What's to stop Tidal, Napster from encoding MQA themselves, SRC a plain vanilla redbook file to 96/24(17), make some EQ adjustments and transmit down the pipe as MQA? Audirvana has a software decoder, surely there's a software encoder, which brings me to:

 

- The provenance of the file is still unknown and a mystery, unless the stream has a sub code that identifies it as an authenticated master that's in a registry to look up, by software somehow, like a UPC. Seems that there are different masterings of tracks/albums being transmitted as MQA, but MQA may not necessarily responsible for the SQ as Chris mentioned.

 

- 30,000 tracks is still not enough, several million would be a suitable of a target for a wide scope of music, but is after all only one label, Warner, for the moment. Where's the rest then?

 

- If the SQ of the Tidal masterings is something positive, would the Schitts of this world have a change of heart and grudgingly add in MQA after consumer pressure? I'd say not, if a software solution developed by someone else, other than the hardware manufacturer is a preferred option.

 

- Aurender and Lumin have been very quiet for MQA support although initially the A10 could have MQA ability. Lumin has stated, no one in the last six months has asked for MQA support.

 

- Tidal could have informed customers in advance of the MQA switch on, but advertising budget constraints may have put the brakes on , or some nerdy type problem prevented some switch on which was unavailable to fix on time.

[crenca]

Well (in response to the direction Chris, Sdolezalek, and abrxx are going - not unreasonably of course), I guess I will ask again - when do we get to evaluate MQA and it's multiple claims of sonic improvement/nirvana? What are the necessary conditions for a reasonable evaluation? Up until now, the common answer was when Tidal was going live with MQA, but now that rug is being pulled out from under us and MQA (like a greased snake) is escaping yet again and we are told once more "just trust us - it IS the greatest thing since..."

 

In any case, you guys sure are making a good case that MQA (at least when it comes to streaming - scratch that, when it comes to everything) is NOT about SQ at all, but rather was always about DRM all day, everyday - as in:

 

"Even if MQA contributed 0% change to audio quality, I still get to listen to remastered hi rez albums that sound fantastic. That can only be a good thing." (abrxx)

 

But, to push in the other direction I do think we have at least a handful of albums (such as the Bob James "The New Cool" I mentioned above) in which all versions (Tidal - both MQA and 16/44, Hi Res download from HDTracks and elsewhere, etc.) are from the same mix/master and so in point of fact it IS possible even now to do an apple to apple comparison...

[abrxx]

Well (in response to the direction Chris, Sdolezalek, and abrxx are going - not unreasonably of course), I guess I will ask again - when do we get to evaluate MQA and it's multiple claims of sonic improvement/nirvana? What are the necessary conditions for a reasonable evaluation? Up until now, the common answer was when Tidal was going live with MQA, but now that rug is being pulled out from under us and MQA (like a greased snake) is escaping yet again and we are told once more "just trust us - it IS the greatest thing since..."

 

In any case, you guys sure are making a good case that MQA (at least when it comes to streaming - scratch that, when it comes to everything) is NOT about SQ at all, but rather was always about DRM all day, everyday - as in:

 

"Even if MQA contributed 0% change to audio quality, I still get to listen to remastered hi rez albums that sound fantastic. That can only be a good thing." (abrxx)

 

But, to push in the other direction I do think we have at least a handful of albums (such as the Bob James "The New Cool" I mentioned above) in which all versions (Tidal - both MQA and 16/44, Hi Res download from HDTracks and elsewhere, etc.) are from the same mix/master and so in point of fact it IS possible even now to do an apple to apple comparison...

 

I don’t think that MQA does nothing at all, I was just making the point that if it that were true we still have a whole bunch of great sounding albums that to my ears sound way better than the existing redbook versions. I can hardly complain about that! My gut feeling is that compared to the existing redbook flacs on Tidal the new Masters sound better 70% due to improved mastering 20% due to higher sample rate and 10% due to MQA magic. But thats just my gut feeling.

[crenca]

I don’t think that MQA does nothing at all, I was just making the point that if it that were true we still have a whole bunch of great sounding albums that to my ears sound way better than the existing redbook versions. I can hardly complain about that! My gut feeling is that compared to the existing redbook flacs on Tidal the new Masters sound better 70% due to improved mastering 20% due to higher sample rate and 10% due to MQA magic. But thats just my gut feeling.

 

Based on my listening I am can not say that the MQA "Hi Res" (I put it in quotes because it is debatable as to how much the 24/96 shown in my DAC's control panel really indicates something - as others have pointed out this is more a marketing coup by MQA than the truth) version is always, or even usually, "better". However, I will say that you are on to something with your "gut feeling" percentage breakdown. That said, MQA is designed (as all DRM/IP protected black boxes are) to thwart inspection and clear evaluation. This is yet another reason to keep it out of your digital eco-system - why would you give up this aspect of your system to a black box when you (and every other audiophile) is obsessive about controlling this or that minute aspect of their system/sound?

 

But I don't think MQA is going to prove bullet proof - like I said I think we are going to find multiple examples (even on Tidal) that are going to be real apple to apple comparisons and we will be able to make honest, reasonable comparisons that will allow a winner to be picked.

 

I spent part of this evening with "The New Cool" (Tidal MQA & 16/44, Hi Res download - straight and upsampled with HQPlayer) and I am going to write up my impressions tomorrow. Unless someone wants to make a convincing case that this album has been re-re-mastered/mixed in the short time it has been out, I am going with the assertion that this is an "apples to apples" eval...

[crenca]

Ok, I have spent quality time with what I will call 4 "playback versions" of Bob James "A New Cool" which I take to be an example of a recording with the same mix/master across all available versions:

 

1) 16/44 via Tidal

2) MQA via Tidal

3) 24/48 via JRiver

4) 24/48 upsampled to DSD x 4 (11289600) via HQPlayer (using ASDM7 and poly-sinc-short)

 

And the winner is....wait for it.... NONE! Fact is, I can not be confident in any of the differences I am hearing between all four of these playbacks, as I believe they are all within the 'margin of error' of my hearing acuity and expectation bias. Actually, I can hear just a touch of difference with the HQPlayer/upsampled version in that it gives me that strange mix of rounded off decay and slightly "suppressed" (or is it just right?) tone of the HF harmonics of the (well recorded) and yet more "natural" sounding piano of this album. This fact is counter intuitive because one of the things I miss on an average/poorly recorded piano is the quality of the HF harmonics. The stand up bass in this recording also sounds very natural or live - they captured the "woodiness" and complexity of the instrument that is missing in so many recordings and is the first thing you notice when you hear a stand up played in person. At first I thought the MQA version was capturing the bass the best, but after more few more listens I can only say that the MQA might (just might) be activating the very lowest harmonics (out of my subwoofer) just a tad more than the other versions - almost as if a 1 db shelf has been added starting around 38 cycles or so to the MQA version. I could distinguished nothing significant around the various appearances of strings or voice on various tracks.

 

 

My conclusion is that on certain recordings that are done well in the first place, as most "audiophiles" know, there just is not that much difference between 16/44 and Hi Res. MQA does nothing to change this fact and appears not to significantly change the way the 16/44 sounds either for better or worse. Would a 24/96 or higher version of this recording made a difference? Perhaps, but my guess it would be quite small.

 

I would be interested in others opinions, but I am going to have a hard time believing that anyone can hear significant differences between the MQA and 16/44 here - if you claim to, please be as specific as possible because I want to hear it also! I know that the MQA version "unfolds" to 24/96 as indicated by DAC control, but what does this mean really? That they used the actual "master" PCM 24/96 (or higher) in the creation of the MQA, and if so why does it not sound "better" than my 24/48, or the Tidal 16/44? So the Hi Res 24/48 download is downsampled from a higher res master - does not appear to make any difference with this recording.

 

What all this means is that there might actually be no value in the alleged MQA SQ improvement at all, excepting in older and/or poorly done recordings (which I recognize are the majority that most listen to), but then we are back to the "which is it - a remaster/mix or the MQA that is effect here" question. Which version did I choose to listen to when typing out this comment? The HQPlayer upsampled 24/48 of course...which really means it "wins" but I admit this preference might not be meaningful to someone else...

[bobbmd]

Thanks Jud went to the dead's web site you quoted I hadn't been there for some time-very interesting and it just goes to what I said elsewhere- it's everyone's else's ears for themselves I just enjoy the music and the reproduction of it. Am so glad my kids in the '90's reintroduced me the DEAD and entrenched in me in americana/country-blues/bluegrass. Do you think all the rest of Jerry's music and styles with Grisman etc will be in MQA? I hope so and what about Sinatra and other RatPack members music?

[Ralf11]

The only thing I'm saying is this - You're comparing Apples to Oranges. You have no idea if the MQA version sounds different because of MQA or because of different mastering. Thus, there's no way to make a technology judgment. The only conclusion supported by your listening experience is that two versions sound different.

 

It's possible this is a new master and MQA made it sound much worse or much better than the straight PCM. We have no clue.

 

I think only almost all bets are off - if numerous MQA releases sound different but not better than "regular" releases, then we know something. This is what the post just above says.

 

Similarly, if the MQAs only sound a bit better then we also know something. Several here have reported this.

 

Only if we find that MQA sounds a lot better (as Stereophile claimed) do we both care, but not know why.

 

 

From the consumer's point of view, why a MQA release sounds much better may not matter either.

 

Now, the important thing is: which Workingman's Dead do I buy? Should I get the DVD-A or not?

[abrxx]

Based on my listening I am can not say that the MQA "Hi Res" (I put it in quotes because it is debatable as to how much the 24/96 shown in my DAC's control panel really indicates something - as others have pointed out this is more a marketing coup by MQA than the truth) version is always, or even usually, "better". [snip]

 

From what I have read, based largely on the available patents (which appear to show MQA 0.9 or some kind of previous version), and interpolating what Bob has said at previous points, for fully decoded MQA, lets say for a original master of 192/24

 

- The core 48/24 layer is reproduced audibly losslessly, with perhaps a a bit depth reduction to 48/21, which as we all know cannot be heard.

 

- The next layer for the ultrasonics between 48 and 96 Khz (corresponding to audible frequencies 24-48 Khz) is encoded under the noise floor is a almost lossless fashion. I have no qualms about a 96 Khz sample rate light being on with an MQA file. According to the patents its almost a lossless representation of the original 96 Khz file.

 

- Further levels of ultrasonics (96-192 Khz, corresponding to audible frequencies of 48 Khz-96 Khz, and beyond for further foldings) are preserved in a lossy way. Since we can't hear these frequencies, and they barely exist in music, I'm not sure how much of an issue this actually is. The reason MQA works at this frequency is to preserve temporal data, not frequency data (their theory).

 

How this all works no-one is 100% sure since the patents do not fully align with what we know and what Bob has claimed.

 

But the point I want to make clear, is that MQA has at least three different approaches to encoding the file, depending on the frequency range. This is very clear from the two different core patents involved. How I wish Bob would produce a white paper that once and for all would explain the process. Oh well....

[crenca]

From what I have read, based largely on the available patents (which appear to show MQA 0.9 or some kind of previous version), and interpolating what Bob has said at previous points, for fully decoded MQA, lets say for a original master of 192/24

....But the point I want to make clear, is that MQA has at least three different approaches to encoding the file, depending on the frequency range. This is very clear from the two different core patents involved. How I wish Bob would produce a white paper that once and for all would explain the process. Oh well....

 

I wonder what the implications are of mansr's plots:

 

http://www.computeraudiophile.com/f8-general-forum/mqa-spectrum-plots-31199/

 

If I understand mansr, Miska, etc. correctly they are calling into question the straightforwardness of Bob's claims about what and how the "ultrasonic" sound/data is being captured and played back to the end user.

 

Also, what would a "white paper" reveal given that the DRM/IP/Software Patent nature of black boxes such as MQA is designed to obscure and hide what is really going on? If I am not mistaken, it would be (yet more) "just trust us"...

[abrxx]

I wonder what the implications are of mansr's plots:

 

http://www.computeraudiophile.com/f8-general-forum/mqa-spectrum-plots-31199/

 

If I understand mansr, Miska, etc. correctly they are calling into question the straightforwardness of Bob's claims about what and how the "ultrasonic" sound/data is being captured and played back to the end user.

 

Also, what would a "white paper" reveal given that the DRM/IP/Software Patent nature of black boxes such as MQA is designed to obscure and hide what is really going on? If I am not mistaken, it would be (yet more) "just trust us"...

 

Yeah I saw those plots. I appreciate folks posting real date rather than just muttering conspiracy theories. I haven't got time to look into at the moment though. But maybe at one point Bob himself can respond at some point. Perhaps Chris can get another round of Q&A after we've had time to listen to the music and do some testing as well.

 

Regardless of what people think of MQA (personally I am cautiously for it), I think we can agree that the marketing and promotion towards audiophiles and tech savvy folks has been a complete mess. I'm glad MQA is now firmly here, but what a crazy ride the last two years have been, with all the changes of direction etc.

[[email protected]]

DVD-A version of American Beauty is different than the SACD version. Is there more than one version on the DVD-A (is there a stereo version as well as downmix from quad)?

 

So which of these (if any) is the MQA version taken from?

 

MQA needs to be transparant and tell us what source/mix was used to make the MQA/MASTER. This should be in the Metadata so that Roon and Tidal can display it for us and eliminate all the speculation as the source.

 

Apart from this pet peeve I am enjoying the MQA/MASTER tracks through my Explorer2 DAC. Thanks Tidal.

[agladstone]

Jud-interesting where do you get your info? I never delved that deeply into the esoterics of Dead recordings-by the way 'delve' is according to dr Jerome Berrigan(of the famous anti-war Berrigan brothers) and my 11th grade advanced placement english teacher a 'semi archaic' expression which I used in the opening lines of my thesis on Samuel Taylor Coleridge and the Rime of the Ancient Mariner!

I feel the DVD-A recordings of both American Beauty and Workingman's Dead to be their best recordings and that MQA of American Beauty approximates those.Just can't believe YOUR knowledge of how these recordings were done!

I have the DVD-A (also ripped to 24/96 FLAC ), the red book (ripped to 16/44 from late '80's CD) and the MoFi SACD's (ripped to DSD 128) of both American Beauty and Working Mans Dead), I much prefer the MoFi SACD / DSD versions of both, perhaps that's why I did not care for the MQA streams of them?

In my system the Low end of the MQA and DVD-A sounds bloated to me in comparison to the MoFi DSD versions?

[agladstone]

Yeah I saw those plots. I appreciate folks posting real date rather than just muttering conspiracy theories. I haven't got time to look into at the moment though. But maybe at one point Bob himself can respond at some point. Perhaps Chris can get another round of Q&A after we've had time to listen to the music and do some testing as well.

 

Regardless of what people think of MQA (personally I am cautiously for it), I think we can agree that the marketing and promotion towards audiophiles and tech savvy folks has been a complete mess. I'm glad MQA is now firmly here, but what a crazy ride the last two years have been, with all the changes of direction etc.

"What a long strange trip it's been" Lol!

[The Computer Audiophile]

MQA needs to be transparant and tell us what source/mix was used to make the MQA/MASTER. This should be in the Metadata so that Roon and Tidal can display it for us and eliminate all the speculation as the source.

 

Apart from this pet peeve I am enjoying the MQA/MASTER tracks through my Explorer2 DAC. Thanks Tidal.

 

It's not MQA Ltd. that needs to be transparent, it's the record labels. MQA only offers its technogy to record labels that use it how they want.

[maelob]

Put me in the camp of I am just happy that I am listening to higher resolution music - I am just happy to see my DAC change sample rates LOL without shelling additional money. I know you guys are deep in the weeds about sourcing, timing, mastering etc etc, but in my book it is a win win, I didnt have to buy a MQA DAC and finally my "regular' DAC is getting some exercise. Also, I think some of you are taking the "master" label too literal. I am guessing it is just a label representing higher resolution "whatever that means". Either way I am just enjoying my new content on Tidal. cheers