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

[esldude]

4 minutes ago, Arpiben said:

 

Using Sonic Visualizer & Audacity I do not reproduce the spectogram singularities of your MQA vs Original HiRes.

 

The MQA file needs to be moved left 3 samples to line up with the HiRes.  Or 3 samples cut off the front of MQA.  

[Arpiben]

Spectogram of the difference between Original Hi-Res  and Sample_C (MQA).

[pkane2001]

5 minutes ago, Arpiben said:

Spectogram of the difference between Original Hi-Res  and Sample_C (MQA).

 

There are some differences betweeen the files — in addition to a fractional offset there is a sample ‘clock’ drift. For my plot, both of these were corrected before subtracting, and the spectrogram taken of the result.

[esldude]

Here is a result in Audacity of moving the file 3 samples.  It doesn't show the curving blue line, and colors are different.  But broadly the same result.  You see the dips in response at the same frequencies and some idle tones in the same places. 

 

[Arpiben]

9 minutes ago, pkane2001 said:

 

There are some differences betweeen the files — in addition to a fractional offset there is a sample ‘clock’ drift. For my plot, both of these were corrected before subtracting, and the spectrogram taken of the result.

 

I corrected the 636234/5 samples at the beginning and cut the excess at the end.

I didn't apply any drift correction. 

[pkane2001]

2 minutes ago, Arpiben said:

 

I corrected the 636234/5 samples at the beginning and cut the excess at the end.

I didn't apply any drift correction. 

 

When I get back to the computer, I’ll rerun the spectrogram with the drift correction disabled. Let’s see what that looks like.

[pkane2001]

2 hours ago, Arpiben said:

 

I corrected the 636234/5 samples at the beginning and cut the excess at the end.

I didn't apply any drift correction. 

 

So the sample clock drift correction is what's causing the artifact. Here's the difference spectrogram with drift correction turned off, but still aligned to a fraction of a sample:

 

 

[pkane2001]

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.

 

Please see the large peak at 24KHz and an additional +/-4KHz ringing around it (blue is original Hires, pink is MQA):

 

What's interesting is that after MQA processing, some of the ringing has been reduced (same plot, but MQA is now on top):

 

 

[austinpop]

12 hours ago, PeterSt said:

 

Rajiv,

Let someone rename them for you, while taking notes of what was done. Then it is your turn to listen ...

After your writing out of what you think of it, all the A's B's and C's should be changed back accordingly.

 

Using this suggestion, I did some anonymized listening listening tonight. Once done, I mapped my impressions back to the original A-B-C.

 

I'll talk in terms of two snippets of the file:

1. From 0:00 to 0:20, when the guitar starts the piece off:
   • I found A to sound flatter, with the guitar not as dimensional. The plucks did not have the same texture and detail
   • B and C sounded very similar. Both sounded fuller than A, with more detail
2. From 1:10 to 1:30, when the cymbals kick in:
   • Again A seemed to lack a bit of detail
   • B was more detailed, where the fine details of the drumsticks striking the cymbals were more evident
   • C matched B in detail, and additionally seemed to be tonally calmer, and more natural sounding.

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. I will say that the differences in this case were all small. It was nothing like a component upgrade.

[PeterSt]

Odd that people maybe have consensus over B and C sounding very similar;

For me B is the one which is the most different. But I started out with that one, maybe that makes a perceptual difference ?

 

Still, most of this (for me) will be about the fake Hires I somehow can't bear while the first MQA I can't bear yet has to come. Well, sort of on the latter but if I have heard 200 MQA's by now maybe 20 failed on these ears (and I count in all the Led Zeps which are all "off" IMO).

 

So B and C sound similar to many ? that would be logical of course, assumed that both are in the "hires domain". An as logical conclusion could be that if you upsample Redbook (A) the proper way, it starts to sound similar to MQA. Mentioned "proper" is subjective of course, but obviously it is what I am trying to do ...

 

Leaky filter me.

[PeterSt]

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

or maybe :

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. 

[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.

[PeterSt]

9 minutes ago, manisandher said:

I'm done with all hires.

 

Ah, not so fast ...

 

 

Can anyone make something of this beautiful piece of art ?

@pkane2001, what can your software make of this ?

@esldude, your subjective view, please ?

[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.

[sandyk]

22 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. I will say that the differences in this case were all small. It was nothing like a component upgrade.

Hi Rajiv

I found the differences to be very obvious, although I made no attempt to decide which file was which, just which sounded best to me.

 I was however listening via coax SPDIF from an internal Asus Xonar D2X soundcard into a highly modified X-DAC V3 and Class A Headphone amp into AT W1000 headphones.

Regards

Alex

 

[church_mouse]

Oh Mani, what have you done?  Certainly, this is now one very confused Stratfordian.

 

I spent my Sunday morning listening to the samples and my conclusion was A and C similar, but I preferred A, with B definitely different but not in good way.  The lower end on B was definitely fuller, but the start and stop off the notes seemed to have gone, the thwack on the string missing, and the top end seemed slightly off.

 

I now discover that, as with your previous challenge, I liked the Hi-Res least!

 

Are my ageing ears so attuned to Redbook that nothing else satisfies?

 

Grasping for a silver-lining; I guess I can look with cool detachment as vendors seek to tempt me to update my library with with MQA or Hi-Res versions, confident/resigned to the fact that I will probably prefer the Redbook.

[esldude]

10 minutes ago, PeterSt said:

 

Ah, not so fast ...

 

 

Can anyone make something of this beautiful piece of art ?

@pkane2001, what can your software make of this ?

@esldude, your subjective view, please ?

Why does this go up to 60 khz? Otherwise the banding of dips in the noise floor are what I've seen already.   So what are you making of it?  I've seen ADCs with filtering like that.  Or it could be a resampling artifact. 

[PeterSt]

24 minutes ago, esldude said:

Why does this go up to 60 khz?

 

I just show no more. It goes to 96KHz ...

[PeterSt]

9 minutes ago, church_mouse said:

Are my ageing ears so attuned to Redbook that nothing else satisfies?

 

But what does your DAC to to it ? I mean, you could have an NOS DAC with nothing else, or with upsampling/filtering in software, or ...

Or an OS DAC with nothing else or with upsampling/filtering in software.

Or you make DSD of it.

Or ...

 

All these things matter. For example, supposed that Hires was fake indeed and you'd take the 44.1 version of it, and next you'd use a NOS DAC with coincidentally the same upsampling means as was applied to the fake Hires, there wouldn't be a difference, right ?

 

---------------

 

What people might easily forget is the importance of how the hardware runs. Mani can possibly remember how I started out with the DAC and how a prerequisite was that in all circumstances the hardware had to run 100% the same because else we would be dependent on buffer sizes, different current draw for different sampling rates, etc. etc. Because, without notice this was not the case. Most of us (beginning computer audio) guys used a FireFace800 and the higher the Hires the higher the buffer settings had to be. This made comparison apples and oranges ....

And believe it or not, already these 10 years or so ago I set for the "as honest as possible" comparison between Hires and Redbook, so I could be working on Redbook being the best and I explicitly created that DAC which always runs at the same speed. From one came the other and now software had to do upsampling ... (and the DAC had to do nothing, hence Non Oversampling).

 

Example of the importance : People with some knowledge on audio chips will easily tell you that any PCM1704U-K (24 bit PCM R2R) which runs at 705600 (16x Redbook) will not behave at its best because possibly too much to its limits (the datasheet stating 96000 as the max which is abusively anyway, but still ...). But I wouldn't care if all I output runs always at this speed, so things become comparable.

 

This is also how it is so plain wrong to output to 2x even if the MQA is 44.1 (or 48) natively only. See the previous "Mani" thread. How would I even be able to compare with Redbook if FIRST some stupid upsamples it himself (say Roon inspired) so next I wouldn't be able to make my own upsampling of it ? That would not allow to compare Redbook to MQA ...

 

It goes on and on. But the strategy is / has to be very deep or else it is apples and oranges everywhere.

[PeterSt]

24 minutes ago, esldude said:

Why does this go up to 60 khz?

 

My original response said "I just show now more".  The w was a typo which makes all even more confusing. Sorry for that.

So I just show no more.

[Le Concombre Masqué]

2 hours ago, manisandher said:

 

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.

Peter think I go too far with the hypothesis that the obscure, poor as of musical interest, made for MQA/audiophilia Chopin was deliberately flawed before MQA stage for demonstration purpose and now I take the other way position : this ECM hires is not flawed. Maybe I'm too naive but I don't buy the fake hires hypothesis ; please read this interview of the engineer : http://www.josephbranciforte.com/james_farber.html or at least this paragraph : How important is mastering to your process?
That's really part of the process for me. I have my favorite mastering engineers, and I attend mastering sessions 99% of the time. I won't use tape until I know who's going to master it, because it can really be botched. It's got to be someone who really can deal with it like Greg Calbi or Mark Wilder. With tape, we get to convert to digital at mastering with the highest quality A/Ds, which are better than the ones that are available in the studio. We really get to choose the wires, the circuitry, the playback chain, and set the playback level of the bass tone so it's tweaked just right, and really maximize that 2-track playback. And then you've got a high-resolution mix that you've got to get down to 16/44.1. What piece of gear do you use for that? And what dither circuit sounds good? So, every step of the way there are these subtle choices that define the sound. There's so many layers of this stuff that you just develop preferences over time. At one point you do a lot of comparison and then eventually you settle on things that you know will work. I know if I'm mastering at X place, I know what their best choices are – just like I know the best mics at a certain studio.

 

If I googled correctly it was recorded in Avatar's studio C : http://avatarstudios.net/rooms/studio-c/ ; would make sense the Digidesign Pro Tools HD3 Accel was used yielding a native 24/96

 

 

It finally boils down to personal preferences, and how one's suspension of disbelief is turned on/off. I understand how the fat round edged RB can seduce here, I myself had a slight preference for MQA (C) at first listen before I dug how low dry and fast the bass goes in B. And that characteristic matches my system (active, "asservi" (servo -controlled?) for control of bass membranes in speed and acceleration)

 

But in the end, do we really want to encourage flawed design, drm etc with the hope that Masters will be actually Quality Authenticated ?

 

If Peter is right, and we've been abused with fake hires in DVDA , SACD, HDTracks run by the Chesky guys (well I think it's now proven/admitted/corrected that it happened), do you really believe that there would be any more efforts to go for the actual masters with MQA in the long run?

 

 

 

 

[PeterSt]

28 minutes ago, Le Concombre Masqué said:

If Peter is right, and we've been abused with fake hires in DVDA , SACD,

 

Small correction : I don't think anyone is saying that DVD or DVDA and SACD were filled up with faked Hires - especially I don't say that;

The stereo mixes derived from it, all so often are (faked) - no failed.

 

I read through your link earlier today, especially watching for the decency in it (and I do such thing just because you point it out). But I couldn't really prove to myself that it was sufficiently decent to not have it flawed somewhere along a line. There's also too few time "hires" mentioned in it. Exactly one time and it was not even really a subject.

... Wasn't it ECM with the Watermarking ? maybe we see that ... (I don't think we'd see it like this, so just saying)

 

Please keep in mind that there's a difference between faking and not well-executed. So mind you, this (ECM) one does not look like being upsampled to me (although I too saw the 48KHz spuriae) . But what we see can't exist. It even could be Hires which we also can't see because it is masked by the anomalies (at the very same spots normal higher frequency would be). And a ride cymbal not going beyond 600Hz is also not going to help (which btw I doubt somewhat regarding the square sound of it the more lightly you tick (hit) it, but alas).

Anyway the high frequencies do not decay here while they should. If the spectrograph would extend to 192KHz (for visible area) we would still see energy (which can't exist).

Explain how the cymbals at the end - which Mani did not show - only arrive at this realistic (!) level :

 

 

I know, those cymbals were not loud either, but at least those were cymbals and at a fairly normal level - and not a bit of background ticking on a ride cymbal.

So what I mean to say is : this is not well executed somewhere and although it implies Hires just by the looks of it, it is not Hires for that matter because Hires looks different.

 

But let's say that if we can not agree over this, we should agree that Hires sounds worse than Redbook. Is that a better consensus then ?

Nah ...

 

To be fair with you, yesterday I said "through these eyes" (it does not look like Hires). I know myself and I'll say it in such fashion when I can't be 100% sure. If I am 100% sure I would talk slightly differently ...

 

 

[trappy]

Watermarking is mostly DG. I haven’t come across it on ECM.

[pkane2001]

4 hours ago, PeterSt said:

 

Ah, not so fast ...

 

 

Can anyone make something of this beautiful piece of art ?

@pkane2001, what can your software make of this ?

@esldude, your subjective view, please ?

 

Send me the audio file, and I’ll run it through Deltawave. My software doesn’t analyze screen shots from other software