MQA Blind Test Summary & Summary Comments...

[Archimago]

Hey all,
Glad to be done the MQA Core vs. Hi-Res blind test series and basically end off with a summary of sorts:

Part I: Procedure
Part II: Core Results
Part III: Subgroup Analysis
Part IV: Subjective Impressions

MQA: "Final" Thoughts... On Politics and Paradigms.

Many thanks to all the folks here who participated in the testing and submission of survey results!

 

It has been fun "work" getting this together with your help and I think over the years we can recognize that the only way to change the nature of this "hobby" is to try it ourselves. Sometimes there is a need to create a narrative that can be coherent and based on scientific and engineering principles rather than claims and vague impressions of the industry that might or might not be on the same path as consumers. I know that these thoughts may be uncomfortable for some as it intrudes into the "politics" aspect of the hobby... But I trust that if we can be gracious, communicate what we're doing and be transparent in intents, the ensuing discussions, controversies, and debates lead us to progress as we all explore the concept of value together.

 

Whether it's in the "formats" we might prefer or the gear we buy...

Cheers!  

 

[plissken]

Thanks for doing this.

[Archimago]

A pleasure, man...

 

Have a great weekend.

[jabbr]

1 hour ago, Archimago said:

MQA: "Final" Thoughts... On Politics and Paradigms.

Outstanding work.

[FredericV]

What a comprehensive and interesting article. I'm now reading this part, which is worth mentioning:

Quote

 

So, here we are in 2017, with our digital-audio systems designed around tried-and-true paradigms of information theory and psychoacoustics. MQA is forging ahead with its proprietary, partially lossy compressed 24/44 or 24/48 PCM stream with a little bit extra to tell the DAC how it should dither and with which of 16 weak, poorly antialiasing, upsampling filters it should use embedded in the lowest few bits of data. The PCM stream is then typically losslessly compressed using an open-source FLAC encoder for delivery. It can then be streamed by TIDAL or downloaded as a file through typical Internet mechanisms. 

On playback, the audio data is software decoded either on your computer or in your DAC with MQA-compatible firmware back into a reconstituted PCM hi-res stream at 24/88 or 24/96. From there, dither and/or noise shaping can be applied, then the embedded upsampling filter choice is used for final conversion to analogue.

 

 

[left channel]

Well argued, thank you. But I clicked the "electric vehicle" link and found myself looking at a little electric scooter. I doubt that's the EV your wife wants to fight you for every morning.

 

Edit: looks like that's just VigLink playing tricks with us. That service is just evil sometimes. Another example: the "crisis" link alternately takes us to some Dr. Jay's Crisis tween jeans instead of the DVD of a Kennedy documentary you'd probably prefer.

[Mordikai]

I really appreciate the effort. I do wish you could somehow embed the mystery science 3000 guys in the videos though.

[witchdoctor]

I think the takeaway for me was the concerns about the SQ of MQA are being overblown by the haters. At the end of the day MQA has the same SQ as hirez PCM... today. With the labels devoting their considrable resources to MQA it will likely be the goto format for hirez in the future. I don't see the labels chucking the format and suddenly start streaming in 96/24 PCM. As more new music gets mastered in MQA the SQ will only improve.

[firedog]

28 minutes ago, witchdoctor said:

I think the takeaway for me was the concerns about the SQ of MQA are being overblown by the haters. At the end of the day MQA has the same SQ as hirez PCM... today. With the labels devoting their considrable resources to MQA it will likely be the goto format for hirez in the future. I don't see the labels chucking the format and suddenly start streaming in 96/24 PCM. As more new music gets mastered in MQA the SQ will only improve.

 

That makes no sense. Just wishful thinking on your part. 

 

"SQ will only improve? " How, why? Based on what? Some radical improvement in the MQA process? Doubtful. 

 

"Their considerable resources to MQA": But the question is, "How is MQA going to generate an additional revenue stream for them over mp3 and PCM to make up for those expenditures?"

 

So far it generates nothing. And it adds both cash costs and more stages to the process. Unless they see it as a true long term revenue producer, they will dump it, just as they have other formats.

 

So far the cash market for even Redbook based streaming is tiny and has made no real market impact.  Audiophiles on Tidal don't count as market impact - they don't count as a market that means much at all. Only an afterthought. No industry wide infrastructure will be setup and maintained for the small amount of extra cash generated by them. 

 

The real answer to this question may well be what you don't want to hear: manipulating the DRM aspects of MQA to try and "create" a  restricted access product that people will pay additional cash for. Don't say I didn't warn you.

 

BTW, I don't think it will work, unless they remove non-MQA files from the market. And probably not even then. Why? Because we have zero evidence that the broader market, or even a solid chunk of it, is willing to pay serious money for what MQA is offering: as you said, "the same SQ as hi-rez PCM". 

 

 

 

 

[witchdoctor]

Firedog you seem unable to accept the fact that ALL of the major labels are converting their catalogs to MQA and will be mastering in MQA going forward. There are plenty of boutique labels that will continue to offer PCM and DSD for now. Good luck with that that. You can store the files next to your VCR and cassette players.

[beetlemania]

38 minutes ago, witchdoctor said:

Firedog you seem unable to accept the fact that ALL of the major labels are converting their catalogs to MQA and will be mastering in MQA going forward. There are plenty of boutique labels that will continue to offer PCM and DSD for now. Good luck with that that. You can store the files next to your VCR and cassette players.

 

As Firedog wrote, the ONLY reason for the labels to move to MQA is for the DRM.

 

"You can store the files next to your VCR and cassette players."

lol - you know how these files work, right?

[witchdoctor]

40 minutes ago, beetlemania said:

 

As Firedog wrote, the ONLY reason for the labels to move to MQA is for the DRM.

 

"You can store the files next to your VCR and cassette players."

lol - you know how these files work, right?

Stop being in denial and I would advise you to build up your vinyl collection. You'll be happier in the long run.
I'm happier listening to 7000+ MQA albums currently on Tidal with more being dropped every week for an extra $10 a month.

[GUTB]

The testing methodology is bieng extremely disingenuous. The selection of 2L tracks, the participation of mostly non-MQA DACs and the bizarre need to up-sample and filter non-MQA samples shows only that you want to confound differences and achieve a null result.

[witchdoctor]

Hey GUTB, please publish your own test results so we can all benefit, thanks.

[GUTB]

Do you have an MQA DAC? Go listen to some MQA and non-MQA versions of the same track through Tidal. Why do you need validation through a highly manipulated blind test?

[witchdoctor]

2 minutes ago, GUTB said:

Do you have an MQA DAC? Go listen to some MQA and non-MQA versions of the same track through Tidal. Why do you need validation through a highly manipulated blind test?

I did my own blind test, I validated it myself and use the bluesound node. Where are your test results?

[firedog]

1 hour ago, witchdoctor said:

Firedog you seem unable to accept the fact that ALL of the major labels are converting their catalogs to MQA and will be mastering in MQA going forward. There are plenty of boutique labels that will continue to offer PCM and DSD for now. Good luck with that that. You can store the files next to your VCR and cassette players.

They have said they are doing it. Do you realize how many albums are out there to convert? And are all new albums being done in MQA? I don't think so.

What you don't seem to get is that all of those so called commitments are dependent on income streams being generated by MQA. If in a year or two those big income streams don't seem to be appearing, they will cut their losses and dump it.

You still haven't told us how MQA is going to make money for the labels. What's the realistic path to revenue generation that you see?

[barrows]

I really would prefer it if MQA fails.  While it does have the potential to offer some improvements in SQ, IF ONE REQUIRES A LOW RES DATA STREAM IN THE FIRST PLACE (due to perhaps limited bandwidth cloud based services), it does nothing to improve sound quality versus a real hi resolution 24/192 (or DSD, but that is another story) file played back with very good digital filters.

All one has to do to avoid the "time blur" introduced by the ADC is to use an apodizing digital filter, although such "time blur" is virtually inconsequential if the original file was recorded at 24/192 (or higher).  If one wants to avoid pre-ringing artifacts, one can use a minimum phase filter.  With real hi res, and these types of filters one will get better (technically) sound quality than MQA (as the artifacts inherent with MQA's lossy scheme will not be present).

The fear for me, being an audiophile, is that with wide spread acceptance, MQA could mean the near death of real, unaltered, hi res audio, both PCM and DSD, and that would be a real shame.

For the average consumer, perhaps getting all their music from a cloud based streaming service, MQA offers an advantage, if it was widely accepted, but do we really expect that MQA will ever become ubiquitous enough to be standard for all streaming services (spotify, etc)?

In a sense maybe MQA could replace MP3 for average consumers, that would be OK with me if it did not limit the access for audiophiles to true hi res music files.  Maybe we then end up with the general public listening to MQA, and audiophiles listening to true unaltered hi res files, on DACs which produce the advanced sound quality (using their choice of digital filters) they are looking for.

Additionally, the politics, specifically not allowing full MQA "unfolding" via software players, requiring everyone interested in MQA to purchase new hardware (clearly a tactic designed to make hardware manufacturers get on board and pay up, with the promise of selling a bunch more now MQA capable DACs) is not consumer friendly at all.  And of course all the marketing claims, including the name itself are pure BS of the highest order: "authenticated", really, by whom that is, Bob Stuart?

[witchdoctor]

5 minutes ago, firedog said:

They have said they are doing it. Do you realize how many albums are out there to convert? And are all new albums being done in MQA? I don't think so.

What you don't seem to get is that all of those so called commitments are dependent on income streams being generated by MQA. If in a year or two those big income streams don't seem to be appearing, they will cut their losses and dump it.

You still haven't told us how MQA is going to make money for the labels. What's the realistic path to revenue generation that you see?

The path to revenue is for the mass market being willing to pay a premium for hi rez streaming. They sure pay a premium for UHD bluerays and 4K TV's, lets see what happens in 2018 with audio but I agree that they need to wait and see at this point.

[barrows]

9 minutes ago, GUTB said:

Do you have an MQA DAC? Go listen to some MQA and non-MQA versions of the same track through Tidal. Why do you need validation through a highly manipulated blind test?

This is a rigged test.  Think about it, what are you comparing here?

 

to fairly compare MQA vs non-MQA first you need the same master at 24/192, are you getting that from Tidal?  Secondly you need to compare the real hi res file played back on the best DAC of your choice (as we are comparing digital filters here, and some are way better than others) or using a playback software like HQPlayer, or Audirvana which allows one to "create" digital filters similar to MQA (apodizing and minimum phase), vs the MQA DAC.

 

To really understand MQA, and what it is doing it is necessary to have a bit of an understanding of digital filtering, both on the A-D and D-A sides.  If one is interested in learning about digital filtering, i suggest starting here:

 

There is no "magic" in MQA which cannot be reproduced by good digital filters.  Do not be fooled by the "marketing speak" and terms like "time blur", etc.  None of this is new, and it all has been well understood for years.

 

[barrows]

3 hours ago, firedog said:

So far the cash market for even Redbook based streaming is tiny and has made no real market impact. 

I know this is not what you are referring to, but please consider the impact the streaming services are having on musicians, it is very, very significant.

[witchdoctor]

1 hour ago, barrows said:

This is a rigged test.  Think about it, what are you comparing here?

 

to fairly compare MQA vs non-MQA first you need the same master at 24/192, are you getting that from Tidal?  Secondly you need to compare the real hi res file played back on the best DAC of your choice (as we are comparing digital filters here, and some are way better than others) or using a playback software like HQPlayer, or Audirvana which allows one to "create" digital filters similar to MQA (apodizing and minimum phase), vs the MQA DAC.

 

To really understand MQA, and what it is doing it is necessary to have a bit of an understanding of digital filtering, both on the A-D and D-A sides.  If one is interested in learning about digital filtering, i suggest starting here:

 

There is no "magic" in MQA which cannot be reproduced by good digital filters.  Do not be fooled by the "marketing speak" and terms like "time blur", etc.  None of this is new, and it all has been well understood for years.

 

Good luck with your digital filters in cars, your phone, your desktop etc. If it was that easy everyone would do it. MQA=one hirez format that will play everywhere in legacy systems or MQA enabled.

[GUTB]

It’s clearly not just digital filters — what on Earth are you talking about? There would be no need to spend the extra bandwidth on meta data stored in the 44/24 streams. There would be no need for DAC engineers to deal with the Meridian lab to configure thier implementation.

[barrows]

1 minute ago, GUTB said:

It’s clearly not just digital filters — what on Earth are you talking about? There would be no need to spend the extra bandwidth on meta data stored in the 44/24 streams. There would be no need for DAC engineers to deal with the Meridian lab to configure thier implementation.

What are you talking about?  Any proposed sound quality improvement via MQA is strictly, and only, due to the digital filters in use.  This is an engineering fact.

It sounds like you may have been duped into believing "magical" claims and marketing speak.

[GUTB]

1 hour ago, barrows said:

What are you talking about?  Any proposed sound quality improvement via MQA is strictly, and only, due to the digital filters in use.  This is an engineering fact.

It sounds like you may have been duped into believing "magical" claims and marketing speak.

 

I'm sorry, what? Are you an MQA engineer working for Meridian? A DAC designer who implemented MQA?

 

Do you at least own a MQA DAC?