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

[crenca]

I stand corrected: right now, Tidal is keeping the MQA and the 16/44 version of the same album (perhaps not the same master though...see below) which make real A/Bing possible.

 

I want to comment on a single track, "Friend of the Devil" by the Grateful Dead (American Beauty album). I am intimately familiar with this track, having heard it countless (surely more than a 1000) times. For years, at family gatherings I would hack it out it on guitar and harmonize with my sister in the lead - I actually think most of my family liked it because the kept asking us to sing it...or perhaps they just thought we were "cute"

 

The MQA differences:

 

1) Level is louder (isn't this cheating?), I want to say 2db or so.

 

2) Stereo separation is "more". I like it at first - like when I first push those buttons "boast" surround, bass, and other DSP enhancement buttons on cheap equipment, and then after a few listens it strikes me as exaggerated and mechanical. Was the "original" mix intended to sound like this? Will anyone ever know? This obviously effects the sound-stage, which is "more" but again...

 

3) Instrument separation and timbre detail is better, more "air", etc. I am asking myself about treble EQ tweak here as it sounds "better" in the way I can accomplish with some recordings with EQ, but it is also more than that. This is a real improvement I believe.

 

4) Rhythm section has been re-miked and re-recorded Really, the drums and bass "pop" out of the mix instead of being a more muddled "behind the scenes" thing. At 40 seconds in, you can here the drums come in clearly and strongly - on the non-MQA this is almost a non-event.

 

5) The vocals are (like the instruments) more open, clear, and have a bit more detail. It is easy to over-emphasize this - it is more subtle than obvious but it is noticeable. However, after the first few listens I find myself questioning something about it - it has a DSP "69.2 surround sound -THX v 137.8" kind of "digititus" to it...or something...hard to pin down.

 

 

If I had to choose a way to express my overall impression, I would say that the 16/44 mix sounds like "vinyl" (veil and all) and the MQA sounds like digital (more transparency, detail, etc.). I would say it is a largely subtle except that really it is not - not on an "audiophile" level. However, I have this bad aftertaste of processed/DSP/digititus that I am not quite sure what to do with.

 

After the first dozen or so listens, I was going to say the MQA version is the clear winner, however I am now realizing that the MQA version is the "fatiguing" version and that the 16/44 sounds a bit more "natural" (sorry, best I can come up with). The improvements on the bass/drums in the MQA version is significant enough to pull it ahead I think and are the most obvious non-qualified "improvement". If you were to play these tracks side by side for me and not tell me one was MQA, I would guess that one is the RVG version and the other is not, and excepting for the rhythm section "fix" I would have to hedge as to which one I would prefer over the long term.

 

Does any of the SQ evaluation change the objective fact that MQA is DRM and there is almost no chance (I am old enough to never say never) it will ever enter into my digital ecosystem? Nope.

 

I used my lessor "desktop computer speaker system" (that's what the wife thinks it is ) which consists of an a Windows computer feeding an Ifi iDSD nano feeding an Emotiva amp feeding Legacy Audio Studio HD's...

[kilroy]

Of course even a backyard statistician knows that a sample size of 1 ain't worth much and certainly is not representative of the population. Just saying.

[crenca]

Of course even a backyard statistician knows that a sample size of 1 ain't worth much and certainly is not representative of the population. Just saying.

 

Um, ok, I will see your "objectivist" objection and raise you one "subjectivist" trump card from the Founder :

 

"All reviews are subjective. Objective measurements are data sheets and data points that may help so done review a product.

 

Product A has distortion of 2%. That's not a review, that's a data point available in a spreadsheet. The 2% distortion sounded terrible. That's a review."

[abrxx]

Listening with a $30 pair of headphones, I can hear much of what crenca reports here. Although I don't think the drums have been re-recorded. It sounds like the original multi-track master (or is that pre master>) was pulled out and the drum section and the drum reverb levels were changed? That would be a new mix, not just a new master right?

 

crenca, what do you think about the Tigerlily album differences? To my ears even with cheap headphones the differences are startling. Sometimes the MASTER has more high HF, but for one track it appeared to have less! Quite strange really.

[blownsi]

Since you are so familiar with this track, you have the way it should sound memorized. This memory could easily skew your results.

 

I have found this numerous times with remastered CDs. They sound inferior to me because they sound different not necessarily worse. The opposite could be happening with clarity. Years ago there was a device call the BBE sonic maximizer which essentially jacked up the treble. People were led to believe this was enhanced clarity. Upon the first few listens you were blown away by the new sound but later became fatigued and annoyed by it. I'm still not convinced that MQA is anything more than just a fancy equalization which may or may not yield better sound.

[barrows]

I stand corrected: right now, Tidal is keeping the MQA and the 16/44 version of the same album (perhaps not the same master though...see below) which make real A/Bing possible.

 

I want to comment on a single track, "Friend of the Devil" by the Grateful Dead (American Beauty album). I am intimately familiar with this track, having heard it countless (surely more than a 1000) times. For years, at family gatherings I would hack it out it on guitar and harmonize with my sister in the lead - I actually think most of my family liked it because the kept asking us to sing it...or perhaps they just thought we were "cute"

 

The MQA differences:

 

1) Level is louder (isn't this cheating?), I want to say 2db or so.

 

2) Stereo separation is "more". I like it at first - like when I first push those buttons "boast" surround, bass, and other DSP enhancement buttons on cheap equipment, and then after a few listens it strikes me as exaggerated and mechanical. Was the "original" mix intended to sound like this? Will anyone ever know? This obviously effects the sound-stage, which is "more" but again...

 

3) Instrument separation and timbre detail is better, more "air", etc. I am asking myself about treble EQ tweak here as it sounds "better" in the way I can accomplish with some recordings with EQ, but it is also more than that. This is a real improvement I believe.

 

4) Rhythm section has been re-miked and re-recorded Really, the drums and bass "pop" out of the mix instead of being a more muddled "behind the scenes" thing. At 40 seconds in, you can here the drums come in clearly and strongly - on the non-MQA this is almost a non-event.

 

5) The vocals are (like the instruments) more open, clear, and have a bit more detail. It is easy to over-emphasize this - it is more subtle than obvious but it is noticeable. However, after the first few listens I find myself questioning something about it - it has a DSP "69.2 surround sound -THX v 137.8" kind of "digititus" to it...or something...hard to pin down.

 

 

If I had to choose a way to express my overall impression, I would say that the 16/44 mix sounds like "vinyl" (veil and all) and the MQA sounds like digital (more transparency, detail, etc.). I would say it is a largely subtle except that really it is not - not on an "audiophile" level. However, I have this bad aftertaste of processed/DSP/digititus that I am not quite sure what to do with.

 

After the first dozen or so listens, I was going to say the MQA version is the clear winner, however I am now realizing that the MQA version is the "fatiguing" version and that the 16/44 sounds a bit more "natural" (sorry, best I can come up with). The improvements on the bass/drums in the MQA version is significant enough to pull it ahead I think and are the most obvious non-qualified "improvement". If you were to play these tracks side by side for me and not tell me one was MQA, I would guess that one is the RVG version and the other is not, and excepting for the rhythm section "fix" I would have to hedge as to which one I would prefer over the long term.

 

Does any of the SQ evaluation change the objective fact that MQA is DRM and there is almost no chance (I am old enough to never say never) it will ever enter into my digital ecosystem? Nope.

 

I used my lessor "desktop computer speaker system" (that's what the wife thinks it is ) which consists of an a Windows computer feeding an Ifi iDSD nano feeding an Emotiva amp feeding Legacy Audio Studio HD's...

 

Any difference in level will invalidate any comparisons, additionally what are we comparing here? I am assuming that the MQA version is hi res, folded down for streaming, and then re-constitued for playback, is this correct? As 24/96 already exists for American Beauty it seems likely. For me the fair comparison would then be level matched comparisons of the MQA stream vs. a straight 24/96 file playback. BTW, I think that American Beauty is quite well recorded and should be a good test album.

If they are boosting level along with MQA this makes me feel like they are being quite disingenuous, as this will trick almost any listener into preferring the louder track (more apparent detail, bigger soundstage, more dynamics, etc)

[crenca]

Any difference in level will invalidate any comparisons, additionally what are we comparing here? I am assuming that the MQA version is hi res, folded down for streaming, and then re-constitued for playback, is this correct? As 24/96 already exists for American Beauty it seems likely. For me the fair comparison would then be level matched comparisons of the MQA stream vs. a straight 24/96 file playback. BTW, I think that American Beauty is quite well recorded and should be a good test album.

If they are boosting level along with MQA this makes me feel like they are being quite disingenuous, as this will trick almost any listener into preferring the louder track (more apparent detail, bigger soundstage, more dynamics, etc)

 

That's just it, the whole "provenance" question is folded into the MQA marketing claim of "THIS is the "master", what the artist intended, trust us". I also assume that the MQA version is the (or "a") 24/96 version "folded" but who can or will answer such questions? What about re-re-mastering? It's just all part of the MQA black box equation...

[crenca]

Listening with a $30 pair of headphones, I can hear much of what crenca reports here. Although I don't think the drums have been re-recorded. It sounds like the original multi-track master (or is that pre master>) was pulled out and the drum section and the drum reverb levels were changed? That would be a new mix, not just a new master right?

 

crenca, what do you think about the Tigerlily album differences? To my ears even with cheap headphones the differences are startling. Sometimes the MASTER has more high HF, but for one track it appeared to have less! Quite strange really.

 

I am going to do a separate write up about Tigerlily because your right, it is more "complicated" as far as the differences - for now I will just say that it comes across as less of an "improvement" and more of a "different master" or "different DSP profile" mix...

[barrows]

That's just it, the whole "provenance" question is folded into the MQA marketing claim of "THIS is the "master", what the artist intended, trust us". I also assume that the MQA version is the (or "a") 24/96 version "folded" but who can or will answer such questions? What about re-re-mastering? It's just all part of the MQA black box equation...

 

While not necessarily a definitive way to test MQA, another test which would be valid in a subjective way for an individual listener in their system would be to compare your favorite version of something (in this case let's say it is the straight file playback of the 24/96 FOTD) with whatever is being delivered by MQA/Tidal. While it might not be fair (because we do not know what master is used) it will be a legit comparison for one's own self.

I have lots of hires and good quality recordings of, King Crimson, for example, but I am pretty sure there will not ever exist MQA King Crimson for comparison.

[abrxx]

For Duke Ellington, Afro Bossa, at least one of the tracks is a different length, by over a minute! ("Afro-Bossa"). It certainly cannot be from the same original master!

 

Track lengths for "Afro-Bossa", track 1 on the album "Afro Bossa":

 

HIFI 3.05

MASTER 4:22

 

Nobody can claim MQA mastering improves on the length of the performance

[HalSF]

Another very basic and literal-minded question from me (sorry): What exactly was being compared in the A/B listening test of this Dead track? It seems like two simultaneous comparisons are happening:

 

1) Comparison of Tidal’s 16/44 Red Book file to Tidal’s MQA version.

2) Comparison of two completely different masterings of “Friend of the Devil"

 

Apart from whatever contribution was made by having one track being louder than the other, how can a comparison of two starkly contrasting masterings possibly help isolate and identify the contribution being made by the MQA secret sauce? Seems like the mastering may be the key source of difference, not MQA itself.

[Dr Tone]

To be 99.9% certain you have the same mastering, you pretty much have to start with something that was released in the last year and is also available in MQA on Tidal.

[abrxx]

Another very basic and literal-minded question from me (sorry): What exactly was being compared in the A/B listening test of this Dead track? It seems like two simultaneous comparisons are happening:

 

1) Comparison of Tidal’s 16/44 Red Book file to Tidal’s MQA version.

2) Comparison of two completely different masterings of “Friend of the Devil"

 

Apart from whatever contribution was made by having one track being louder than the other, how can a comparison of two starkly contrasting masterings possibly help isolate and identify the contribution being made by the MQA secret sauce? Seems like the mastering may be the key source of difference, not MQA itself.

 

That's just it. We can't tell whether the improvements are from the MQA process, from the high resolution file, from the new master/mix, or from some combination of all three. And its unlikely we will ever know for sure. But as consumers we can A/B the new Tidal MASTERS vs the HIFI tracks, and talk about the differences, and whether we prefer MASTER, and why. That's worth doing, especially if Tidal start charging more for MASTER at some point.

[weldp]

I am listening on the Meridian Explorer 2 which has hardware MQA decoding and HD650 headphones.

 

I have been A/B listening to HiFi vs MASTER across 10 or so albums today. The starkest difference I am hearing is on the "American Beauty" album. On MQA "Box of Rain" the bass is clearly deeper and more detailed. High end is smoother and less harsh.

 

-weldp

[Ralf11]

I hope you'll adjust SPLs (use a good meter) and let us know if it's a friend of mine...

[bobbmd]

crenca- My observations are similar-I also feel I am intimately familiar with American Beauty (and Workingman's Dead) through the Dead's dvd-a versions. This MQA version is similar in SQ, separation etc to the dva-a disc. I don't find it fatiguing tho but that is just my 'subjective' opinion. I don't care how they got to this level-I like it. Another example is Running on Empty I don't think I ever heard with such clarity the background talking /comments during before and after some the songs on CD's or otherwise-maybe I being tricked somehow on the sound reproduction but its 'better' and yes it is 'louder' I don't have to turn the volume up as much on my old AVR or on my Schiit FULLA or Magni with either my Sennheiser RS 195's or HD 598's-not necessarily a bad thing and I think the bass overall is much better.

IMHO this maybe a keeper and I won't have to buy any more expensive hardware(I have a premultibit Gungnir) but will spring for an Audioquest DF RED for the MQA ability.

Really nice observations thanks bobbmd

[abrxx]

I hope you'll adjust SPLs (use a good meter) and let us know if it's a friend of mine...

 

I don't need an SPL meter to tell that (for "Box of Rain"):

 

- the track lengths are different

- the stereo separation on the vocals is different (centre on MASTER, centre-right on hifi)

 

Whether the mastering level is connected to the following I am not so sure:

 

- appears to be more space in the mix, greater dynamic range possibly, bass is more solid, and treble clearer.

 

My gut feeling is that the differences for many of the MASTER tracks are 70% new mix, 20% hi-rez and 10% MQA magic.

[sete]

Does any of the SQ evaluation change the objective fact that MQA is DRM and there is almost no chance (I am old enough to never say never) it will ever enter into my digital ecosystem? Nope.

 

 

What's wrong with DRM? Does it affect sound quality? No, right? (I'm not asking in jest, I don't know much from the tech side of things).

 

I understand the reluctance with DRM if you were actually buying the stuff, and of course Apple's removal from the iTunes Store arguably made the download market possible. But my feeling is that MQA is the perfect way to serve high-res stuff over streaming, as with this Tidal example.

[Samuel T Cogley]

crenca- My observations are similar-I also feel I am intimately familiar with American Beauty (and Workingman's Dead) through the Dead's dvd-a versions. This MQA version is similar in SQ, separation etc to the dva-a disc. I don't find it fatiguing tho but that is just my 'subjective' opinion. I don't care how they got to this level-I like it. Another example is Running on Empty I don't think I ever heard with such clarity the background talking /comments during before and after some the songs on CD's or otherwise-maybe I being tricked somehow on the sound reproduction but its 'better' and yes it is 'louder' I don't have to turn the volume up as much on my old AVR or on my Schiit FULLA or Magni with either my Sennheiser RS 195's or HD 598's-not necessarily a bad thing and I think the bass overall is much better.

IMHO this maybe a keeper and I won't have to buy any more expensive hardware(I have a premultibit Gungnir) but will spring for an Audioquest DF RED for the MQA ability.

Really nice observations thanks bobbmd

 

+1

 

I listened to the title track last night in MQA and my first thought was, "I don't remember it sounding like this in '77". Don't misunderstand, I found the presentation quite enjoyable.

[Samuel T Cogley]

What's wrong with DRM? Does it affect sound quality?

 

MQA absolutely affects sound quality compared to the original hirez master. Is this change preferable? Time and the market will tell.

[Jud]

While not necessarily a definitive way to test MQA, another test which would be valid in a subjective way for an individual listener in their system would be to compare your favorite version of something (in this case let's say it is the straight file playback of the 24/96 FOTD) with whatever is being delivered by MQA/Tidal. While it might not be fair (because we do not know what master is used) it will be a legit comparison for one's own self.

I have lots of hires and good quality recordings of, King Crimson, for example, but I am pretty sure there will not ever exist MQA King Crimson for comparison.

 

MQA sez, "All your King Crimson are belong to us."

 

(Kidding, folks.)

[Jud]

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?

[witchdoctor]

The only evaluation I needed was to listen to MQA tracks for an hour and then go back and listen to redbook tracks. I didn't like the redbook so much after switching. I certainly wouldn't want to go back to 320kb MP3.

[new_media]

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?

 

There was a hi-res PCM/vinyl re-release 2 or 3 years ago, which I believe was a new master. I assume the MQA is encoded from that.

[Ralf11]

GDO is re-releasing all the albums - not sure when they'll get to Am. Beauty, but that will add to the various releases out there...