MQA vs HiRez: an apples-to-apples comparison

[PeterSt]

7 minutes ago, mansr said:

Neither file has the MQA rendering data in the LSB. It is not a pure capture.

 

Would it be there when decoded ?

(I wondered this yesterday, for a first time)

[mansr]

2 minutes ago, manisandher said:

Huh???

 

 

Mani.

How did you capture the decoded signal?

[mansr]

Just now, PeterSt said:

Would it be there when decoded ?

(I wondered this yesterday, for a first time)

The MQA decoder puts instructions for the renderer in the LSB of the output. That's how the upsampling filter selection works.

[manisandher]

2 minutes ago, mansr said:

How did you capture the decoded signal?

 

It's in the OP:

 

3 hours ago, manisandher said:

In both cases, the original and the MQA were played back through Roon, and the output captured using Audacity. So everything remained in the digital domain.

 

It's strange that Miska seems to be able to discern a difference from analysis (still not saying whether he's wrong/correct though).

 

Mani.

[manisandher]

9 minutes ago, PeterSt said:

 

It seems it was on a Thursday. The thread ended on Sunday though. Maybe because of being startled I remember that particular day.

 

 

Anyway, the text of one of the posts went like this :

 

A little zoomed out. What we see here, I only saw after a long time as a pattern ... what do you see ? Look at each of the peaks and compare the both takes ... Do you see it ? Somehow there is a consistent "opposite" balance in how the peaks are formed. Just find a peak where one sample bumps out. It doesn't matter where you find it. Now look at its counterpart. See ? the one with one sample jumping out has 3 samples to *let* that sample jump out, where it counterpart has an even number of samples for the width concerned, and it is impossible for one sample to jump out. If you try to reason where the "base" of this emerges, at least it is hard to see, but it must be at the base of the wave somewhere, where the one is wider as the other. This is a most consistent thing throughout the both files ! Will this change sound ? ha, you bet it will.

 

16/44.1 vs. 24/176.4

 

So something like that I have in mind.

 

 

OK. Look forward to your analysis...

 

Mani.

[mansr]

3 minutes ago, manisandher said:

It's in the OP:

3 hours ago, manisandher said:

In both cases, the original and the MQA were played back through Roon, and the output captured using Audacity. So everything remained in the digital domain.

It's not clear how you hooked the Roon output into Audacity. You should also make sure to disable dithering in Audacity when saving the files.

 

3 minutes ago, manisandher said:

It's strange that Miska seems to be able to discern a difference from analysis (still not saying whether he's wrong/correct though).

Not at all. One has been through the MQA processing.

[manisandher]

2 minutes ago, mansr said:

It's not clear how you hooked the Roon output into Audacity.

 

Surely there's only one way... keeping everything in the digital domain, as I did.

 

2 minutes ago, mansr said:

You should also make sure to disable dithering in Audacity when saving the files.

 

There's no conversion going on.

 

2 minutes ago, mansr said:

Not at all. One has been through the MQA processing.

 

OK, so which is the MQA?

 

Mani.

[mansr]

Just now, manisandher said:

Surely there's only one way... keeping everything in the digital domain, as I did.

There are numerous ways it could be done while staying digital.

 

1 minute ago, manisandher said:

There's no conversion going on.

Well, something has messed with the data. If you want an accurate comparison, the first step is to make sure you haven't accidentally altered the files in some way that wouldn't happen on normal playback.

 

1 minute ago, manisandher said:

OK, so which is the MQA?

I don't know (yet).

[manisandher]

21 minutes ago, mansr said:

Well, something has messed with the data. If you want an accurate comparison, the first step is to make sure you haven't accidentally altered the files in some way that wouldn't happen on normal playback.

 

In Roon, I'm selecting 'no MQA support':

 

 

This is exactly how I would config things for normal playback with any non-MQA DAC.

 

Mani.

[mansr]

6 minutes ago, manisandher said:

In Roon, I'm selecting 'no MQA support':

 

 

This is exactly how I would config things for normal playback with any non-MQA DAC.

That still doesn't answer how the data is sent from Roon to Audacity.

[manisandher]

12 minutes ago, mansr said:

That still doesn't answer how the data is sent from Roon to Audacity.

 

WASAPI with all enhancements off.

 

Mani.

[PeterSt]

1 minute ago, manisandher said:

WASAPI with all enhancements off.

 

On a 2nd PC ?

(or else I think Shared Mode (vs Exclusive Mode) would be active and then everything can happen)

[mansr]

4 minutes ago, manisandher said:

WASAPI with all enhancements off.

Still not answering the question.

[wklie]

28 minutes ago, manisandher said:

In Roon, I'm selecting 'no MQA support'

 

I think that's the reason the rendering metadata is absent.  I believe that's the best setting for this test, as the alternative setting may make it too easy for mansr.

 

(For those who don't know: Roon has the ability to separate the music data and the MQA render instructions, DSP the music data, then recombine it with the render instructions.  That's why Roon can perform DSP on MQA while still maintaining MQA music data integrity.)

[PeterSt]

2 minutes ago, mansr said:

Still not answering the question.

 

I know you're having fun, as usual. But give the man some clues could work faster.

[manisandher]

4 minutes ago, mansr said:

Still not answering the question.

 

It's very straightforward.

 

Roon is in Shared Mode (correct Peter). Audacity is capturing via WASAPI. What are you still confused about?

 

Mani.

[mansr]

2 minutes ago, wklie said:

I think that's the reason the rendering metadata is absent.  I believe that's the best setting for this test, as the alternative setting may make it too easy for mansr.

Does Roon apply some additional post-processing if that setting is chosen? That doesn't seem like a good idea if an accurate comparison is the goal.

[manisandher]

1 minute ago, mansr said:

Does Roon apply some additional post-processing if that setting is chosen? That doesn't seem like a good idea if an accurate comparison is the goal.

 

It should make identifying the MQA even easier. And yet you haven't (yet).

 

Mani.

[mansr]

Just now, manisandher said:

It's very straightforward.

That may be, but I still can't read your mind.

 

Just now, manisandher said:

Roon is in Shared Mode (correct Peter). Audacity is capturing via WASAPI. What are you still confused about?

OK, that answers it. Now the question is where the alteration of the decoded MQA data happens. It could be done by Roon, WASAPI, or Audacity.

[manisandher]

1 minute ago, mansr said:

That may be, but I still can't read your mind.

 

OK, that answers it. Now the question is where the alteration of the decoded MQA data happens. It could be done by Roon, WASAPI, or Audacity.

 

Any alteration of the decoded MQA should make the MQA easier to detect.

 

Come on, based on all your criticisms of MQA, it should be trivial for you to identify which is which. So, which is the MQA?

 

Mani.

[mansr]

Just now, manisandher said:

It should make identifying the MQA even easier.

How so? If Roon is doing processing of its own, it may be masking some differences between MQA and the original. If it's doing something only to MQA files, then we're listening as much to Roon as we are to MQA. That wasn't the goal of this test.

 

Just now, manisandher said:

And yet you haven't (yet).

That's because I haven't tried.

[manisandher]

1 minute ago, mansr said:

If Roon is doing processing of its own, it may be masking some differences between MQA and the original. If it's doing something only to MQA files, then we're listening as much to Roon as we are to MQA. That wasn't the goal of this test.

 

Of course it's doing something only to the MQA file that it's not doing to the original 24/88.2. It's called MQA decoding.

 

Mani.

[manisandher]

5 minutes ago, mansr said:

That's because I haven't tried.

 

I look forward to seeing your analysis of how MQA has totally molested the original hirez, and couldn't possibly sound any good.

 

Of course, I don't expect you to even try listening to both and making a decision that way. Using your actual ears to determine whether something sounds good? God forbid.

 

Mani.

[wklie]

12 minutes ago, mansr said:

Does Roon apply some additional post-processing if that setting is chosen?

 

Based on the Roon signal path shown, it performs MQA Core decoding to 24-bit 88.2kHz, does not do any further processing of the music data, then sends it to OS mixer.

[mansr]

5 minutes ago, manisandher said:

Of course it's doing something only to the MQA file that it's not doing to the original 24/88.2. It's called MQA decoding.

I'm talking about what happens after that. The missing renderer instructions mean that something has altered the data after it left the MQA decoder.