Jump to content
IGNORED

Testing MQA: Is it worse than FLAC?


Recommended Posts

I just signed up so I could comment on this. 

 

Thanks for taking the time to write up your theory and the tests undertaken. Interesting, insightful read. I recall the first time I heard about MQA and I spent hours reading their literature and other people analysis, what a farce yet genius way the extract money from the punters. 

 

I look forward to future updates when your time permits. Thanks for sharing this. Best wishes, Dan

Link to post
Share on other sites

I think your ignoring the question that might be most important to some:

 

Does an MQA file played through a dac with full MQA capabilities sound better than the 16/44.1, 24/96, 24/192, etc.. alternatives?

 

Who cares if it isn’t truly lossless if it actually sounds a little better?

 

I’m starting to see posts that claim MQA cd’s sound better than SACD’s. It certainly is interesting.

 

Link to post
Share on other sites

I wonder, does expectation work both ways? i.e. those who think MQA is fake news will think it sounds bad?

 

Don’t have any MQA gear, cd’s, or a Tidal subscription personally, but I have a hard time dismissing everyone who says they like how it sounds as flat earthers just because someone analyzed some wave forms and made some observations about them.

Link to post
Share on other sites
1 hour ago, saturdayboy said:

I think your ignoring the question that might be most important to some:

 

Does an MQA file played through a dac with full MQA capabilities sound better than the 16/44.1, 24/96, 24/192, etc.. alternatives?

 

Who cares if it isn’t truly lossless if it actually sounds a little better?

 

I’m starting to see posts that claim MQA cd’s sound better than SACD’s. It certainly is interesting.

 

Unfortunately its a question that can't be properly addressed.
Though I did address in the post WHY it can't be addressed.

MQA does not allow any device with 'full unfolding' capabilities to output a digital stream. It is ONLY allowed for analog only output devices. 
This means that we can never get a proper test of the native file vs fully unfolded MQA.
This is also assuming that you can find an MQA track that has a 192khz native hires release.

We can ONLY test by recording and comparing the results of analog output.
I will be doing some null tests recording from the analog output of a fully MQA capable DAC using the ADI-2 Pro ADC (performance info for which can be seen here: http://archimago.blogspot.com/2020/09/measurements-rme-adi-2-pro-fs-r-black.html ) 
This will not be able to tell us if the fully unfolded file is 'bit perfect' as it is recording from the analog output. But it will be able to show if there are any obvious discrepancies between native and MQA.

I'll additionally compare the audible band only content of the fully unfolded MQA file to the native FLAC audible band content to see if there are differences.

Its not the most ideal test, but unfortunately MQA has deliberately made it impossible to do an ideal test.

In regards to MQA being better than SACD, its an apples to oranges comparison.
SACD is 1 bit and so the quality of playback will be much more dependent on the source file, how it was produced and modulated down to 1 bit etc. (Much in the same way as you can use HQPlayer DSD modulators to get some pretty fantastic sound out of otherwise mediocre DACs by effectively 'replacing' the delta sigma modulator)

Whereas MQA is still PCM, and so will be more dependent on which DAC you're using.

Subjectively, to me, MQA sounds similar to minimum phase upsampling. And other third party testing suggests that it's quite likely a lot of MQA content IS just minimum phase upsampling. 
What's the best subjectively will always be up for debate unfortunately. So I won't get into that other than being open about the fact that I don't like it though mostly because of the attached business practices, if it were a choice it would actually be nice to have for certain situations where data or storage is limited. But for anyone at home with a good internet connection or cheap hard drive to put their music on, MQA is a solution to a problem that doesn't exist, being forced upon you and something good (native hires audio) getting taken away in the process.

https://youtube.com/goldensound

Roon -> HQPlayer -> SMS200 Ultra/SPS500 -> Holo Audio May (Wildism Edition) -> Holo Audio Serene (Wildism Edition) -> Benchmark AHB2 -> Hifiman Susvara

Link to post
Share on other sites

Really thoughtful post, although I don’t agree that it’s unfortunate that people will always debate what is subjectively best. We do that with almost everything, why should it be any different with digital file formats/conversion?

Link to post
Share on other sites
7 hours ago, saturdayboy said:

This actually involved some listening, which I know is discouraged when MQA bashing is involved.

 

https://audiophilereview.com/cd-dac-digital/a-comparison-of-sacd-vs-mqa-in-physical-format/

It's a  bunch of blah blah which has zero meaning. He didn't even mention if he knows the MQA and SACD were derived from the same master. Good chance they aren't. If so, you are comparing masters at least as much as formats. 

You aren't a veteran here Lots of us have listened to MQA and even blind tested it. Not sure your attitude is appropriate.

Main listening (small home office):

Main setup: Surge protector +_iFi  AC iPurifiers >Isol-8 Mini sub Axis Power Conditioning+Isolation>QuietPC Low Noise Server>Roon (Audiolense DRC)>Stack Audio Link II>Kii Control>Kii Three >GIK Room Treatments.

Secondary Listening: Server with Audiolense RC>RPi4 or analog>Matrix Element i Streamer/DAC (XLR)+Schiit Freya>Kii Three .

Bedroom: SBTouch to Cambridge Soundworks Desktop Setup.
Living Room/Kitchen: RPi 3B+ running RoPieee to a pair of Morel Hogtalare. 

All absolute statements about audio are false :)

Link to post
Share on other sites
2 hours ago, fiske said:

I assume the closest testbench is the one from 2L.
http://www.2l.no/hires/index.html?

 

I'm not into MQA and have no intention of checking it out.

Not necessarily. Those 2L files were appparently made using an MQA "white glove" treatment - that's nothing like the batch MQA processing that is used for 99.9% of the cases. 

Main listening (small home office):

Main setup: Surge protector +_iFi  AC iPurifiers >Isol-8 Mini sub Axis Power Conditioning+Isolation>QuietPC Low Noise Server>Roon (Audiolense DRC)>Stack Audio Link II>Kii Control>Kii Three >GIK Room Treatments.

Secondary Listening: Server with Audiolense RC>RPi4 or analog>Matrix Element i Streamer/DAC (XLR)+Schiit Freya>Kii Three .

Bedroom: SBTouch to Cambridge Soundworks Desktop Setup.
Living Room/Kitchen: RPi 3B+ running RoPieee to a pair of Morel Hogtalare. 

All absolute statements about audio are false :)

Link to post
Share on other sites
5 hours ago, firedog said:

Expectation works both ways. Caveat: it a subconscious process. By definition, you don't know what your expectation bias is. Saying "I expected it to sound better/worse" (take your pick) means nothing.  That's why we have blind testing.

The expectation can be conscious or subconscious. Blind testing eliminates the conscious elements. 

Link to post
Share on other sites
2 minutes ago, saturdayboy said:

The expectation can be conscious or subconscious. Blind testing eliminates the conscious elements. 

Nope.  Look it up. Conscious expectations and subconscious expectations are different. They can also be shown by testing to be opposite; that is you think your bias is one way, but it's shown to be another. Blind testing, done properly, eliminates bias of both types. That's the point. 

Main listening (small home office):

Main setup: Surge protector +_iFi  AC iPurifiers >Isol-8 Mini sub Axis Power Conditioning+Isolation>QuietPC Low Noise Server>Roon (Audiolense DRC)>Stack Audio Link II>Kii Control>Kii Three >GIK Room Treatments.

Secondary Listening: Server with Audiolense RC>RPi4 or analog>Matrix Element i Streamer/DAC (XLR)+Schiit Freya>Kii Three .

Bedroom: SBTouch to Cambridge Soundworks Desktop Setup.
Living Room/Kitchen: RPi 3B+ running RoPieee to a pair of Morel Hogtalare. 

All absolute statements about audio are false :)

Link to post
Share on other sites
16 minutes ago, firedog said:

Nope.  Look it up. Conscious expectations and subconscious expectations are different. They can also be shown by testing to be opposite; that is you think your bias is one way, but it's shown to be another. Blind testing, done properly, eliminates bias of both types. That's the point. 

From the wiki

 

“It may include conscious or unconscious influences on subject behavior including creation of demand characteristics that influence subjects, and altered or selective recording of experimental results themselves.[2]

 

 

Link to post
Share on other sites
5 hours ago, firedog said:

It's a  bunch of blah blah which has zero meaning. He didn't even mention if he knows the MQA and SACD were derived from the same master. Good chance they aren't. If so, you are comparing masters at least as much as formats. 

You aren't a veteran here Lots of us have listened to MQA and even blind tested it. Not sure your attitude is appropriate.

Ok. What about people who are of the opinion that the MQA versions streaming on Tidal are better than the non MQA? I’m pretty sure that more often than not those are the same masters.


The point is are all those people who think it sounds better wrong, and you’re right?

Is personal taste irrelevant in listening?
 

How many times have you challenged someone to perform double blind testing when they tell that one pizza joint is better than another, or they prefer the smell of one rose over another?

Link to post
Share on other sites
52 minutes ago, saturdayboy said:

The expectation can be conscious or subconscious. Blind testing eliminates the conscious elements. 

 

Sorry not always. Conscious is what MQA does when they demo their files. What I mean is, they actually tell you what to expect to hear compared to a non-MQA CD and that is conscious bias. Your brain then fills in what the MQA peeps tell you. 

 

Unconscious bias is based on a person's likes and dislikes and this happens in the background w/o the person knowing about it.

 

Double blind tests, which are random, basically do away with both. 

Current:  JRiver 26 on Win 10 PC (AMD Ryzen 5 2600 with 32 GB RAM) or Daphile on an AMD A10-5700 with 8 GB RAM

DAC - TEAC UD-501 DAC 

Amplification - Audio Research SP-16 > Pioneer M-22 - Bow Technologies Wazoo Integrated (great amp - silly name)

Speakers: Wharfedale Linton Heritage - KEF LS50 - Revel M126Be - others......

Cables: Tara Labs RCS Reference speaker cables and DiMarzio Interconnects

Link to post
Share on other sites
1 hour ago, saturdayboy said:

Ok. What about people who are of the opinion that the MQA versions streaming on Tidal are better than the non MQA? I’m pretty sure that more often than not those are the same masters.


The point is are all those people who think it sounds better wrong, and you’re right?

Is personal taste irrelevant in listening?
 

How many times have you challenged someone to perform double blind testing when they tell that one pizza joint is better than another, or they prefer the smell of one rose over another?

You have no idea which master is which on Tidal. So same problem.

It's not a question of being wrong or right. It's saying "I hear this in sighted conditions, therefore MQA is shown to be 'better'". 

 

Your final question is irrelevant and has nothing to do with the issue. I haven't said people can't like what they like. Personal taste exists.

 

But that's not the situation with MQA. Saying it as a format sounds better than: CD, SACD, hi-res, etc is a different kind of statement. If you are going to make that kind of statement you need to do a proper comparison. Otherwise it's a comparison and conclusion based on false premises. 

 

BTW, MQA has gone to great lengths to make sure that proper comparisons don't happen. Such as refusing Mark Waldrep's request that they use his own masters to make an MQA'd version so he could directly compare and analyze them. His is just one example.

 

The only properly blinded test of MQA we know of (McGill) didn't show any listener preference for MQA. It also wasn't a full blown test comparison, but it's the best we've got so far. 

Main listening (small home office):

Main setup: Surge protector +_iFi  AC iPurifiers >Isol-8 Mini sub Axis Power Conditioning+Isolation>QuietPC Low Noise Server>Roon (Audiolense DRC)>Stack Audio Link II>Kii Control>Kii Three >GIK Room Treatments.

Secondary Listening: Server with Audiolense RC>RPi4 or analog>Matrix Element i Streamer/DAC (XLR)+Schiit Freya>Kii Three .

Bedroom: SBTouch to Cambridge Soundworks Desktop Setup.
Living Room/Kitchen: RPi 3B+ running RoPieee to a pair of Morel Hogtalare. 

All absolute statements about audio are false :)

Link to post
Share on other sites
1 hour ago, saturdayboy said:

From the wiki

 

“It may include conscious or unconscious influences on subject behavior including creation of demand characteristics that influence subjects, and altered or selective recording of experimental results themselves.[2]

 

 

Again wrong. You looked up only one kind of bias. There are multiple types. I suggest you give up this part of the argument as you are only repeatedly showing you don't understand the basics. 

Main listening (small home office):

Main setup: Surge protector +_iFi  AC iPurifiers >Isol-8 Mini sub Axis Power Conditioning+Isolation>QuietPC Low Noise Server>Roon (Audiolense DRC)>Stack Audio Link II>Kii Control>Kii Three >GIK Room Treatments.

Secondary Listening: Server with Audiolense RC>RPi4 or analog>Matrix Element i Streamer/DAC (XLR)+Schiit Freya>Kii Three .

Bedroom: SBTouch to Cambridge Soundworks Desktop Setup.
Living Room/Kitchen: RPi 3B+ running RoPieee to a pair of Morel Hogtalare. 

All absolute statements about audio are false :)

Link to post
Share on other sites
20 minutes ago, firedog said:

Again wrong. You looked up only one kind of bias. There are multiple types. I suggest you give up this part of the argument as you are only repeatedly showing you don't understand the basics. 

What I looked up was “expectation bias”, which I thought was what was being discussed. If you search “expectation bias” on Wikipedia you’ll find my quote was taken from that page.

 

Are you an observational psychologist? You seem to claim expert level knowledge in all things relating to perception and bias.

Link to post
Share on other sites
26 minutes ago, saturdayboy said:

What I looked up was “expectation bias”, which I thought was what was being discussed. If you search “expectation bias” on Wikipedia you’ll find my quote was taken from that page.

 

Are you an observational psychologist? You seem to claim expert level knowledge in all things relating to perception and bias.

And you got an article about observer-expectancy bias, which isn't really what we are talking about in this setting. I didn't claim to be an expert. Just more expert than you. Quoting not particularly relevant passages from Wikipedia doesn't do a lot for your argument. I suggest you talk about something audio related instead. 

Main listening (small home office):

Main setup: Surge protector +_iFi  AC iPurifiers >Isol-8 Mini sub Axis Power Conditioning+Isolation>QuietPC Low Noise Server>Roon (Audiolense DRC)>Stack Audio Link II>Kii Control>Kii Three >GIK Room Treatments.

Secondary Listening: Server with Audiolense RC>RPi4 or analog>Matrix Element i Streamer/DAC (XLR)+Schiit Freya>Kii Three .

Bedroom: SBTouch to Cambridge Soundworks Desktop Setup.
Living Room/Kitchen: RPi 3B+ running RoPieee to a pair of Morel Hogtalare. 

All absolute statements about audio are false :)

Link to post
Share on other sites
1 hour ago, firedog said:

You have no idea which master is which on Tidal. So same problem.

It's not a question of being wrong or right. It's saying "I hear this in sighted conditions, therefore MQA is shown to be 'better'". 

 

Your final question is irrelevant and has nothing to do with the issue. I haven't said people can't like what they like. Personal taste exists.

 

But that's not the situation with MQA. Saying it as a format sounds better than: CD, SACD, hi-res, etc is a different kind of statement. If you are going to make that kind of statement you need to do a proper comparison. Otherwise it's a comparison and conclusion based on false premises. 

 

BTW, MQA has gone to great lengths to make sure that proper comparisons don't happen. Such as refusing Mark Waldrep's request that they use his own masters to make an MQA'd version so he could directly compare and analyze them. His is just one example.

 

The only properly blinded test of MQA we know of (McGill) didn't show any listener preference for MQA. It also wasn't a full blown test comparison, but it's the best we've got so far. 

The McGill study compared 16/44.1 MQA files to their uncompressed 24/96 counterparts, who really expected MQA to win that battle?

 

Isn’t 16/44.1 MQA vs regular 16/44.1 the actual fair fight? 

Link to post
Share on other sites
2 hours ago, saturdayboy said:

Ok. What about people who are of the opinion that the MQA versions streaming on Tidal are better than the non MQA? I’m pretty sure that more often than not those are the same masters.


The point is are all those people who think it sounds better wrong, and you’re right?

Is personal taste irrelevant in listening?
 

How many times have you challenged someone to perform double blind testing when they tell that one pizza joint is better than another, or they prefer the smell of one rose over another?

 

People also use Shun Mook discs to improve the sound of their audio system, even though proof that the discs work is non-existent. I guess it is all in their brain, that there is an improvement.

 

If they like MQA, fine. I have no issues with people saying they like it. I do have an issue with the idea, to pay more for a lossy file with a questionable master (as we don't know what master they use for encoding). The other issue is the fact the the audiophile press has done NO TESTING on MQA, to prove what they are saying about this format. They test everything else, why take BS/MQA's word for it?

Current:  JRiver 26 on Win 10 PC (AMD Ryzen 5 2600 with 32 GB RAM) or Daphile on an AMD A10-5700 with 8 GB RAM

DAC - TEAC UD-501 DAC 

Amplification - Audio Research SP-16 > Pioneer M-22 - Bow Technologies Wazoo Integrated (great amp - silly name)

Speakers: Wharfedale Linton Heritage - KEF LS50 - Revel M126Be - others......

Cables: Tara Labs RCS Reference speaker cables and DiMarzio Interconnects

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now


×
×
  • Create New...