MQA vs HiRez: an apples-to-apples comparison - FINAL

[manisandher]

I can't imagine there's any appetite for yet another hires/MQA comparison... but I wanted to do this for myself anyhow, and decided I may as well share. You see, from a purely personal perspective, the score from the previous two apples-to-apples comparisons is: MQA 1 _ Hires 1. I can't leave it there - this has to be a best of three.

 

For all those who participated in the previous two threads, don't feel at all obliged to share your thoughts, or indeed do the comparison at all. But the outcomes of the previous two were inconclusive to my mind, and it might be nice to take a final listen to another apples-to-apples comparison.

 

As before, I played and captured 3 files:

- an original hires 24/96

- an MQA 24/48, decoded to 24/96

- a 16/44.1, converted to 24/96 with iZotope

 

Each file was played back on Roon (in WASAPI Exclusive Mode), sent to a USB-to-spdif converter and captured at 24/96 on a Tascam recorder.

 

The original hires and the hires capture null perfectly:

 

The original and the MQA capture show the following difference:

 

So everything looks perfectly in order.

 

Here are the 3 captures:

 

Sample A

https://drive.google.com/open?id=1Mbg55FWQjs98qfXlZG3cnq4U6FetAaxk

 

Sample B

https://drive.google.com/open?id=1OTOpIQcYjd3_DVAvJXtVFkAIjBJui7uv

 

Sample C

https://drive.google.com/open?id=10wuw9xu9sR0pmaSb6k5Edbz67bpYF5h4

 

(It's trivial to identify the 16/44.1 by analysis. By all means go ahead and do this, but please refrain from sharing the results of your analysis until I reveal the answer, in a few days' time.)

 

To my ears, there's a clear difference in sound between the 3. But which is the MQA?

 

Mani.

[manisandher]

On 10/13/2018 at 3:54 AM, Lee Scoggins said:

Peter's recordings are really excellent without the MQA encoding but the MQA encoding made a pretty noticeable difference.  With the MQA encoding, the sound of the room became much more clear.  There was more fullness in the mids and bass and transients seemed more lifelike.  It seemed like the soundstage got both wider and deeper.  It wasn't subtle.  The MQA files sounded much better and more natural like live music in a real space.

 

@Lee Scoggins, I wonder if you can hear the same 'MQA effect' in one of the three samples I've linked in the OP? I'd say this is a more realistic test of MQA's efficacy.

 

Care to take a listen and let me know?

 

Mani.

 

[manisandher]

23 minutes ago, Lee Scoggins said:

**I am not sure a pure play A/B exists in Tidal where the mastering has been identical except for the MQA encoding.

 

The original hires file (not from Tidal), the MQA (from Tidal), and the 16/44.1 (from Tidal) used in this thread are from the same master.

 

Mani.

[manisandher]

19 minutes ago, esldude said:

How do you know this for sure?

 

Musicscope analysis.

 

19 minutes ago, esldude said:

And is the 16/44.1 from Tidal the MQA version undecoded or actually just straight redbook?

 

It's redbook - undecoded MQA would be 48, not 44.1.

 

Mani.

 

[manisandher]

1 hour ago, esldude said:

But then I don't trust myself in sighted listening anyway.

 

I think you're doing yourself a disservice.

 

1 hour ago, esldude said:

Maybe I'll abx them tomorrow. 

 

I suspect any differences you heard sighted will vanish. The question then is: were you just hearing differences sighted that didn't actually exist, or... is ABX is less effective way of judging subtle differences? I'm squarely in the latter camp when there is absolutely no reason to be biased one way or another, as is the case in this comparison.

 

Mani.

[manisandher]

57 minutes ago, esldude said:

I'm squarely in the opposite camp.

 

Bias doesn't need a reason.  It is inherent in how the human organism works.  

 

Once you've analysed them, then your bias is going to set in. "I couldn't possibly have heard such subtle differences. I must have imagined them."

 

But as long as you're happy...

 

Mani.

[manisandher]

2 hours ago, Le Concombre Masqué said:

about bias : seeing austinpop going for 16/44 (as I did) but almost seduced by MQA and interested in the recording pianist (I would very much rather consider Richter Pollini Arrau Argerich...) I wonder if the Chopin's recording at the center of test II has/was made with/ any other purpose than biased demonstration to start with

 

Not everyone preferred Sample A (the MQA file), though a surprising number did. My personal preference was B (the hires).

 

Interestingly, this FINAL test is proving a tad more consistent than the first two. I still haven't released the answer yet, so there's still time for others to take part.

 

Mani.

[manisandher]

@John_Atkinson, do you fancy giving this 'test' a go?

 

Mani.

[manisandher]

2 hours ago, austinpop said:

What genre of music is this? The last 2 were not genres I listen to. 

 

It's a fusion of Arabic music and jazz... I think. I picked it because it has a lot of energy throughout the spectrum.

 

Mani.

[manisandher]

1 hour ago, Ralf11 said:

Bias is inherent in how all organisms work.  Even very simple neural nets "fill in" or predict "information" - after an instant's thought i suppose I should limit this to non-plants but I'd bet money that systems of auxins do the same.

 

OK, so which of A, B, or C are you biased towards, and why?

 

Mani.

[manisandher]

@PeterSt, thanks for the detailed description of what you're hearing. I'm assuming your ears are no longer stuffed?

 

4 hours ago, PeterSt said:

This leaves A as the Hires.

 

I think you mean "A as redbook", to be consistent with your detailed description.

 

Mani.

[manisandher]

10 minutes ago, Arpiben said:

As a side note, I am having more differences with null test than expected. Need to double check &/or wait for results' disclosure.

 

If you're doing analysis, you'll identify the redbook track immediately. Nulling the other two, you should get what I showed in the OP. (Let me know if you're not.) I'll link the original hires (for a limited time only) when I reveal the results, probably some time tomorrow. You'll then be able to null the captures with the original.

 

Mani.

[manisandher]

18 hours ago, audiventory said:

(MQA is lossy, as far as I know).

 

If (and it's a massive 'if') MQA does what they claim it does, I think 'corrected' is a more accurate term than 'lossy'.

 

18 hours ago, audiventory said:

Because conclusion is technically obvious: lossless is better.

 

The hires should therefore sound "obviously" better than the MQA. In which case it should be possible to identify which is the MQA. Care to tell us which of the samples sounds "obviously" inferior to you? (Understanding of course that we can't generalise your subjective thoughts to everyone else.)

 

Mani.

[manisandher]

3 hours ago, PeterSt said:

The one subject to this thread is also doubtful to these eyes. Straight to 48KHz for this track throughout ?? How ? where ? what exactly ?

 

 

Here's my analysis of the original hires (with no gain applied):

 

 

What troubles you about this?

 

(If I apply 60dB gain, I get something closer to yours. But is this in any way meaningful?)

 

Mani.

[manisandher]

Thanks to the 9 people who posted their thoughts.

 

I've given a score of '2' for your preference, and a score of '1' for your second choice. Also, you'll see some ticks to the right-hand-side. A black tick shows that you explicitly identified all 3 tracks correctly. A grey tick shows that you probably identified all 3 tracks correctly, but didn't explicitly say so.

 

And here are the results:

 

 

(With my scores added, the totals become 6, 10 and 10).

 

I think we can conclude that the truth between hires and MQA does indeed lie somewhere in the middle!

 

Mani.

[manisandher]

On 10/27/2018 at 1:10 AM, austinpop said:

 

Could you also post a Tidal (or Qobuz) link to the track? 

 

The track was Anouar Brahem - Blue Maqams - Opening Day (MQA on Tidal: https://tidal.com/track/79927158).

 

Mani.

[manisandher]

4 minutes ago, Rexp said:

Good game Mani, thx for organising!

 

You're welcome.

 

4 minutes ago, Rexp said:

My take away is neither hi res nor MQA is the answer to providing (much needed) better sound than redbook. 

 

?

 

Mani.

[manisandher]

15 minutes ago, PeterSt said:

Also, I looked into this this morning because if I had to be correct on the how I came to my judgment on the Hires (sounds the worst) then first I had to be correct on it being fake, right ?

 

The label is ECM, which I felt wouldn't engage in such things. It's a real shame if indeed it is 'fake' hires. But perhaps it was out of their control?

 

Mani.

[manisandher]

@Arpiben and anyone else who's interested, here's the original hires file (for a limited time only) to help with any analysis you'd like to do:

 

https://drive.google.com/open?id=14Xau4aJ5ms391itiAYbEQLt6R3SAaDwt

 

Mani.

[manisandher]

5 minutes ago, esldude said:

I correctly identified hires.

 

Uh no, you didn't - you thought A was the hires:

On 10/26/2018 at 7:10 AM, esldude said:

If I were going with conventional ideas, Hi-res, MQA and CD.

 

You said this about the actual hires (that you thought was MQA):

On 10/26/2018 at 7:10 AM, esldude said:

B sounds a bit artificially airy, and soft imaging.

 

You said this about the actual MQA (that you thought was CD):

On 10/26/2018 at 7:10 AM, esldude said:

C is missing some low end, and has unnatural hardness on the upper mids.

 

I agree with your descriptions in both cases though.

 

5 minutes ago, esldude said:

I correctly commented that mqa was more different than the others.

 

Yes, you did.

 

Mani.

[manisandher]

2 hours ago, pkane2001 said:

By the way, I neglected to post the cepstrum plot for the comparison of MQA and Hires. It does point fairly strongly to a Hires file that was manipulated from a 48KHz original.

 

This is soooooo disappointing. You pay good money for a hires album from a respectable label (ECM), purchased from a respectable site (Qobuz), and you get 'fake' hires.

 

Congratulations to @esldude for being the only one to choose non-manipulated redbook as his first choice.

 

I'm done with all hires. Still interested in exploring more 44.1/48 MQA though.

 

Mani.

[manisandher]

16 minutes ago, austinpop said:

Based on this, I liked C best, then B, then A. Looking at Mani's reveal of what these are, I seem to have liked MQA > Hi-Res > CD.

 

Thanks Rajiv. Travelling for work right now, so can't update my little spreadsheet, but will do so when I get back.

 

Mani.

[manisandher]

6 minutes ago, PeterSt said:

@manisandher, did you already put forward your most subjective judgment ?

 

Not explicitly, but I did say:

 

16 hours ago, manisandher said:

(With my scores added, the totals become 6, 10 and 10).

 

My choice was C > A > B.

 

For me, it was a close call between C and A. The cymbals on C were a bit too hot for my liking though. I could definitely live with A, but it didn't have the same 'life' as C.

 

B sounded muffled in comparison to C and A - like a blanket had been placed over the speakers. There was no tunefulness in any of the instruments, especially the piano.

 

14 minutes ago, PeterSt said:

Can you be honest on how you came to select this particular track (or underlaying album) ? the answer "it looked to be from the same master" is also an answer. 

 

Lots of energy from LF to HF. ECM label. Looked like same master. I like the music (for my sins).

 

Sorry, no time to write more right now...

 

Mani.

[manisandher]

7 hours ago, Ryan Berry said:

C. Definitely MQA.  Sounds like someone pressed the bass-boost/loudness button.  Individual players loose their "realness."  

 

I agree, the highs come across as more accentuated in C vs. B. But I'm certainly not hearing any bass-boost in C vs. B. Quite the opposite - the bass sounds tauter and more tuneful in C.

 

To my ears, it seems that MQA (C) is 'correcting' the mess made in the creation of the hires (B), and therefore sounding more like the redbook (A).

 

I'm still totally done with all hires. It seems to me that if you can't get properly done hires from ECM/Qobuz, then you're not likely to get it from anywhere. And no wonder MQA sounds strange to some people, if most of it has been derived from 'fake' hires.

 

Mani.

[manisandher]

2 hours ago, manisandher said:

I'm still totally done with all hires. It seems to me that if you can't get properly done hires from ECM/Qobuz, then you're not likely to get it from anywhere. And no wonder MQA sounds strange to some people, if most of it has been derived from 'fake' hires.

 

1 hour ago, PeterSt said:

Anyway, in my view no reason to ditch MQA. And oh, you can ditch it all right, but I wouldn't do it for this reason. This was just an unlucky pick.

 

Make that 2 unlucky picks then. I think there was an identical issue with the hires I chose in the original apples-to-apples thread (well-reputed label and downloaded from HDTracks):

 

https://drive.google.com/open?id=1_nIJnB5_WspDOMa_-So4LTB-d11b3Qdv

 

I'm not at home right now to check, but IIRC, this has lots of energy going right up to 44.1 kHz (88.2 sample rate), which looks suspicious. I did actually email the label to ask them where all the HF energy was coming from, but never got a reply.

 

Mani.