MQA Listening Impressions

[Sonicularity]

Technically, I don't see how the MQA format could be considered more accurate.  MQA may be different, which could translate to something better or worse for some listeners in various situations, but MQA does not appear to be a more accurate format.  What is the OP's intent for this thread?  There was little meaningful criteria provided.  

 

 

[manisandher]

5 hours ago, Sonicularity said:

What is the OP's intent for this thread?

 

To gather MQA listening impressions

 

Mani.

[mcgillroy]

6 hours ago, Sonicularity said:

...but MQA does not appear to be a more accurate format.  What is the OP's intent for this thread?

 

 

 

Astroturfing.

[manisandher]

On 11/22/2017 at 4:34 PM, manisandher said:

I've since been playing a bunch of MQA albums from Tidal. Invariably, I prefer the sound of these MQA albums over the CD rips and/or downloads of the same albums I already have. They tend to have a clarity that my rips/downloads don't have.

 

Here's the sort of thing I mean. Here are two 24/192 captures from the analogue output of my DAC (Phasure NOS1 G3, with files upsampled to 705.6 and 768 kHz respectively in XXHighEnd before being sent to DAC), playing a CD rip and an MQA file from Tidal.

 

MFSL CD rip:

https://drive.google.com/open?id=1r1UEnu8rfAhMrO5_vObI-OiuDek-cbJs

 

MQA file from Tidal:

https://drive.google.com/open?id=1HJHivSQVXWktlGFMX8-jvR0FyceGHDp4

 

I actually quite like the sound of the MFSL rip. But there's no doubt that the MQA file has extra clarity, bite and punch... to my ears.

 

Mani.

[Miska]

11 minutes ago, manisandher said:

Here's the sort of thing I mean. Here are two 24/192 captures from the analogue output of my DAC (Phasure NOS1 G3, with files upsampled to 705.6 and 768 kHz respectively in XXHighEnd before being sent to DAC), playing a CD rip and an MQA file from Tidal.

 

MFSL CD rip:

https://drive.google.com/open?id=1r1UEnu8rfAhMrO5_vObI-OiuDek-cbJs

 

MQA file from Tidal:

https://drive.google.com/open?id=1HJHivSQVXWktlGFMX8-jvR0FyceGHDp4

 

I actually quite like the sound of the MFSL rip. But there's no doubt that the MQA file has extra clarity, bite and punch... to my ears.

 

Mani.

 

Different master? The other one sourced from hires version compared to RedBook? If you are not comparing same versions, it doesn't tell anything about MQA.

 

If you want to compare RedBook and MQA, you need to play MQA without unfolding.

 

[matthias]

Much better comparison would be to compare the HighRes version from Qobuz vs. the MQA version from Tidal.

 

Matt

[bhobba]

I have read this but people don't seem to 'get'  MQA.  I will get to my listening impression soon, which is reasonably extensive, but I think something needs to be understood first.

 

We are getting to the era where people will not have their own libraries of stuff they listen to - maybe just a few hard to get stuff you cant stream - but the vast majority by most people, including me, will listen by steaming services like Tidal.  MQA allows higher quality streaming - its that simple.  I have yet to find anyone that doesn't prefer MQA after the first unfold - which except for a few cases that is is simply 48/24 and does not have the first fold is lossless 96k.  It may be 18 or so bit dithered (and subtractive dither is used which is better than normal dither) - but it doesn't really matter - after dithering its effectively about 20 bits or better which is very close to the thermal noise limit of microphones.  I know of only one person that can tell the difference between 20 bit and actual 24 bit - via a blind test.  There is no doubt such people are not common - I certainly can't.  As you would expect hearing into a recordings noise floor is difficult - you may think its impossible - but it isn't - we can hear below a noise floor - but doing it is hard.  Anyway the bottom line is after the first unfolding it's not that controversial - MQA is better than 44.1/16 and obtainable via streaming services like Tidal.

 

The controversy is after that - some prefer the full MQA unfolding - which is simply reversing the spline down-sampling the 96k was created with while others prefer different up-sampling - its purely a personal thing.

 

My listening impression of pure MQA comes via the Explorer 2 and the Bridge on the Direct Stream.  To me its slightly thin and polished - but I like it - it tends to sound sort of understated more real.  Some others don't like it as much - they like a thicker presentation.

 

I have heard it through the Aurilac - it's a bit different from the correct full unfolding - but not necessarily worse - again its a personal thing  - I might prefer it to the Direct Stream on some material - but it's hard to say.

 

My favorite for what its worth is 192k up-sampled by HQPlayer to 384k with one exception -  I have a genuine DXD master.   I have the Xivero entropy encoder which is like a truth machine.  Most stuff peters out before 30k but usually, not always, but usually goes down to 24 bits.  This DXD master actually goes all the way to 176K at 24 bits - its the real deal.  BTW the 2l DXD stuff doesn't pass muster - it cuts out well before 176k and is not genuine 24 bits.  As an experiment I down-sampled the DXD master to 88/24 then used HQPlayer to upsample it to DXD.   Sorry - no cigar - the original was easily better.

 

My view - DXD masters done properly are better but require a lot of bandwidth - although there are lossless encoders better than FLAC such as Optim-Frog (and others I have figured out better still but that is another story) and until downloads become common at that speed MQA is pretty much the best we can do right now - but that will improve.   Also IMHO  we need better recording techniques to realize it.    But they will come as well - eventually.

 

I will be doing a lot more comparisons over the coming weeks/months and obviously can say more later.

 

Thanks

Bill

[firedog]

25 minutes ago, bhobba said:

I have yet to find anyone that doesn't prefer MQA after the first unfold - which except for a few cases that is is simply 48/24 and does not have the first fold is lossless 96k.

You just found one.

My reaction to MQA is that is not really an improvement. Some albums sound better, some worse, and some about the same. So  I just don't see the point and what we gain. Certainly it isn't needed for streaming. MQA in terms of bandwidth.

And in any case, we don't know when we compare MQA and non MQA  files if we are comparing the same masters . So the comparisons themselves may be invalid, which makes the whole exercise and evaluation sort of pointless. 

[bhobba]

10 minutes ago, firedog said:

You just found one.

 

Was that just the first unfolding?  Because what you are saying is 96k sounds worse  than 48k - that is a very strange thing.  It can sound the same because the frequency doesn't extend much past 24k - but worse - I am really scratching my head.

 

Sure it may just be different masters - but otherwise it beats me - and the listening I have done with others.

 

Thanks

Bill

[Miska]

1 hour ago, bhobba said:

It may be 18 or so bit dithered (and subtractive dither is used which is better than normal dither) - but it doesn't really matter - after dithering its effectively about 20 bits or better which is very close to the thermal noise limit of microphones.  I know of only one person that can tell the difference between 20 bit and actual 24 bit - via a blind test.

 

More like 14 - 16 bit. And it is quite far from what you are saying when you actually have some ultrasonic content to encode...

 

1 hour ago, bhobba said:

My view - DXD masters done properly are better but require a lot of bandwidth

 

To me, even just 96/24 master is better than MQA version of the same. Even if reduced to 96/16 which makes it smaller/less bandwidth while providing more dynamic range than the MQA version...

 

[Miska]

4 hours ago, manisandher said:

 

Here's the sort of thing I mean. Here are two 24/192 captures from the analogue output of my DAC (Phasure NOS1 G3, with files upsampled to 705.6 and 768 kHz respectively in XXHighEnd before being sent to DAC), playing a CD rip and an MQA file from Tidal.

 

MFSL CD rip:

https://drive.google.com/open?id=1r1UEnu8rfAhMrO5_vObI-OiuDek-cbJs

 

MQA file from Tidal:

https://drive.google.com/open?id=1HJHivSQVXWktlGFMX8-jvR0FyceGHDp4

 

I actually quite like the sound of the MFSL rip. But there's no doubt that the MQA file has extra clarity, bite and punch... to my ears.

 

Mani.

 

Both seem to have high level images around multiples of the sampling rate and cannot see unfolding happening at all. The MQA version is 48 kHz while RedBook is of course 44.1k. Distance (in frequency) to the first image is higher on the MQA version and a slightly lower in level too. Second image is again a bit further up just because of of higher sampling rate. MQA version seems a bit hotter too with some extra clipping. So the MQA version is a bit more compressed.

 

MFSL version:

 

MQA version:

 

[firedog]

2 hours ago, bhobba said:

 

Was that just the first unfolding?  Because what you are saying is 96k sounds worse  than 48k - that is a very strange thing.  It can sound the same because the frequency doesn't extend much past 24k - but worse - I am really scratching my head.

 

Sure it may just be different masters - but otherwise it beats me - and the listening I have done with others.

 

Thanks

Bill

Th

Nope, with an MQA DAC. And I don't think it is so strange. There often is very little to distinguish between even Redbook and hires. Very recording dependent.

[monteverdi]

33 minutes ago, firedog said:

Th

Nope, with an MQA DAC. And I don't think it is so strange. There often is very little to distinguish between even Redbook and hires. Very recording dependent.

Same was very true for SACD with a few exception. That was mostly a waste of effort and money and and had the side effect  to inhibit developments of DSP to improve speaker design (as MQA likely will too)

[manisandher]

1 hour ago, Miska said:

 

Both seem to have high level images around multiples of the sampling rate and cannot see unfolding happening at all. The MQA version is 48 kHz while RedBook is of course 44.1k. Distance (in frequency) to the first image is higher on the MQA version and a slightly lower in level too. Second image is again a bit further up just because of of higher sampling rate. MQA version seems a bit hotter too with some extra clipping. So the MQA version is a bit more compressed.

 

Yes, exactly as expected, seeing as I used XXHighEnd's 'Arc Prediction' filter to interpolate up to 705.6/768 kHz. I could repeat the captures using HQPlayer's poly-sinc-xtr-mp filter if you'd prefer... The sonic difference between the files will still be there.

 

In any event, this is the 'MQA Listening Impressions' thread'. So, which do you prefer the sound of, and why?

 

Mani.

[Archimago]

5 hours ago, manisandher said:

 

Here's the sort of thing I mean. Here are two 24/192 captures from the analogue output of my DAC (Phasure NOS1 G3, with files upsampled to 705.6 and 768 kHz respectively in XXHighEnd before being sent to DAC), playing a CD rip and an MQA file from Tidal.

 

MFSL CD rip:

https://drive.google.com/open?id=1r1UEnu8rfAhMrO5_vObI-OiuDek-cbJs

 

MQA file from Tidal:

https://drive.google.com/open?id=1HJHivSQVXWktlGFMX8-jvR0FyceGHDp4

 

I actually quite like the sound of the MFSL rip. But there's no doubt that the MQA file has extra clarity, bite and punch... to my ears.

 

Mani.

 

This is an example of muddying the waters!

 

How can you compare an MFSL mastering with typical dynamic compressed  consumer mastering and make any generalization about MQA!?

 

As Miska nicely shows us, this is like comparing a Red Delicious and a McIntosh... Both apples but different intrinsically.

 

Like I said before, given the vagueness and complexities of MQA, let's make sure we control the variables before making clajms... Either pro or against MQA.

[manisandher]

5 minutes ago, Archimago said:

How can you compare an MFSL mastering with typical dynamic compressed  consumer mastering and make any generalization about MQA!?

 

So, the MFSL should sound substantially better than the MQA, right? Does it?

 

Mani.

[Archimago]

13 minutes ago, manisandher said:

 

So, the MFSL should sound substantially better than the MQA, right? Does it?

 

Mani.

 

That's subjective isn't it?

 

Some might prefer it... But whether it does or not has little if anything to do with which mastering TIDAL chose to stream nor necessarily the effect from MQA processing...

 

The bottom line is this is not how one compared and discusses the merits of MQA.

[Miska]

13 minutes ago, manisandher said:

 

Yes, exactly as expected, seeing as I used XXHighEnd's 'Arc Prediction' filter to interpolate up to 705.6/768 kHz. I could repeat the captures using HQPlayer's poly-sinc-xtr-mp filter if you'd prefer... The sonic difference between the files will still be there.

 

In any event, this is the 'MQA Listening Impressions' thread'. So, which do you prefer the sound of, and why?

 

Mani.

 

MFSL sounds more natural, something like you'd get off an analog tape. Not really anything disturbing. A bit messy, like a bit older masterings of this genre, like 90's style or so. Sounds like analog tape.

 

MQA version certainly sounds louder and I notice that some algorithm is pumping and splashy (a bit like Dolby C noise reduction on old tape decks). Some EQ to make it sound artificially bit brighter and more on-your-face. Sounds grainier and highs are fussy. Sounds like a cheap sound card or mobile phone, "plastic Tupperware-sound".

 

I could listen the MFSL version longer, my ears already hurt after listening one track of the MQA version at same volume. Neither one is good though.

 

I tested by first converting both to 44.1/32 WAV using poly-sinc filter to remove the images. Then upsampled that to DSD256 using poly-sinc-short-mp filter and ASDM7 modulator. Played through RME ADI-2 Pro (in Direct DSD mode) to Fostex HP-A8C headphone amp and Sennheiser HD-800 headphones.

 

[Ron Scubadiver]

2 minutes ago, Archimago said:

 

That's subjective isn't it?

 

Some might prefer it... But whether it does or not has little if anything to do with which mastering TIDAL chose to stream nor necessarily the effect from MQA processing...

 

The bottom line is this is not how one compared and discusses the merits of MQA.

After reading and enjoying many pages on your blog I am inclined to agree with you on MQA. 

[kumakuma]

Your captures look strange when run through MusicScope.

 

[manisandher]

11 minutes ago, Miska said:

 

MFSL sounds more natural, something like you'd get off an analog tape. Not really anything disturbing. A bit messy, like a bit older masterings of this genre, like 90's style or so. Sounds like analog tape.

 

MQA version certainly sounds louder and I notice that some algorithm is pumping and splashy (a bit like Dolby C noise reduction on old tape decks). Some EQ to make it sound artificially bit brighter and more on-your-face. Sounds grainier and highs are fussy. Sounds like a cheap sound card or mobile phone, "plastic Tupperware-sound".

 

I could listen the MFSL version longer, my ears already hurt after listening one track of the MQA version at same volume. Neither one is good though.

 

Thanks. This is what this thread is all about.

 

Mani.

[manisandher]

9 minutes ago, kumakuma said:

Your captures look strange when run through MusicScope.

 

This is the 'MQA Listening Impressions' thread. You can keep the spectra for other threads.

 

Mani.

[manisandher]

On 11/24/2017 at 9:46 PM, watercourse said:

There are inconsistencies in the results of the current MQA catalog, e.g. some albums simply sound inferior to non-MQA'ed albums, that I feel deserve some explanation. Is it poor quality control? Who does the processing, and how are results checked? Will there be some sort of quality standard that is achieved at some point in the future?

 

Having listened to quite a lot of MQA recently, I'm inclined to agree with this. Where I've managed to get hold of a non-MQA and MQA file definitely from the same master, I don't always prefer the MQA version. I'll post an example shortly.

 

Mani.

[Miska]

9 minutes ago, manisandher said:

 

Thanks. This is what this thread is all about.

 

Mani.

 

This mastering I have is closer to the MQA version (album version), not as annoying as the MQA version:

https://www.highresaudio.com/en/album/view/tfohhv/rush-sector-two-remastered

 

I think that is better comparison point against the MQA version. At least I would myself rather listen this one than the MQA. Try it out!

 

It is not really hires, there's the sharp cut just above 20 kHz in spectrum. I made a spectrogram plot of it, but I don't post it now on your request.

[manisandher]

5 minutes ago, Miska said:

This mastering I have is closer to the MQA version (album version), not as annoying as the MQA version:

https://www.highresaudio.com/en/album/view/tfohhv/rush-sector-two-remastered

 

Haha...

 

I`m sorry!

Dear HIGHRESAUDIO Visitor,

due to territorial constraints and also different releases dates in each country you currently can`t purchase this album. We are updating our release dates twice a week. So, please feel free to check from time-to-time, if the album is available for your country.

We suggest, that you bookmark the album and use our Short List function.

Thank you for your understanding and patience.

Yours sincerely, HIGHRESAUDIO

 

Mani.