MQA is Vaporware

[Fitzcaraldo215]

49 minutes ago, kumakuma said:

 

Full Decoder:

 

https://helpdesk.bluesound.com/discussions/viewtopic.php?f=4&t=4059

How about that. My theory looks kaput.  I though I might have had a way of reconciling the differing accounts of MQA behavior.  

 

I really think the next step would be for @mansr to step up and ask Bluesound to explain why their device does not behave like Stuart says it should.  Possibly tricky, I know, because he might not own the piece he was working with.  

 

I realize many here are not going to change their minds regardless of the outcome, so what is the point, you might ask.  I think there is always a point to getting closer to the truth.  

[kumakuma]

21 minutes ago, Fitzcaraldo215 said:

I really think the next step would be for @mansr to step up and ask Bluesound to explain why their device does not behave like Stuart says it should.

 

Seems more a question for Stuart as the same thing has been observed on multiple MQA-certified DACs.

 

Or perhaps your new theory is that at least three different companies working independently from each other failed to implement MQA properly and that MQA itself missed it in all three cases during the certification process.

 

At this point, though, Occam's razor would seem to apply...

[rickca]

15 minutes ago, Fitzcaraldo215 said:

I really think the next step would be for @mansr to step up and ask Bluesound to explain why their device does not behave like Stuart says it should.

Problem is that Bluesound has most likely signed an MQA non-disclosure agreement and can't talk about it.  Welcome to Catch-22.

[Fitzcaraldo215]

Just now, kumakuma said:

 

Seems more like a question for Stuart as the same thing has been observed on multiple MQA-certified DACs.

 

Or perhaps your new theory is that at least three different companies working independently failed to implement MQA properly and MQA itself missed it in all three cases during the certification process.

 

At a certain point, Occam's razor

No, I have no new theory.

 

 Sure, if someone could get Stuart to sit still and pay attention to this, that would be great.  But, honestly, that ain't gonna happen, looking at all the vague information as it has trickled out.  It is hard to make sense of it unless it is pulled together and organized.   But, going through Bluesound might be the best pathway to get this to the MQA folks for an explanation.  

 

 

[AJ Soundfield]

23 hours ago, AJ Soundfield said:

Plus I don't see any credentials. When in doubt (i.e. the technical stuff is way over my head), I trust higher authority.

Told ya 

[Ralf11]

I agree with Fitz's comments on Bob S.

 

also, what about 3 times folding of a part of the spectrum?  room for that?

[mansr]

2 minutes ago, Ralf11 said:

what about 3 times folding of a part of the spectrum?  room for that?

Nope.

[Ralf11]

hang on - I did not specify how infinitesimally small a portion of the spectrum would be selected ;]

[Fitzcaraldo215]

1 hour ago, Ralf11 said:

I agree with Fitz's comments on Bob S.

 

also, what about 3 times folding of a part of the spectrum?  room for that?

Thanks.

 

 Mansr says no, but Stuart says yes, repeatedly, over and over, about the entire frequency spectrum up to a 384k sampling, not just a piece of it.  Stuart's secret sauce is proprietary, and it is not even in patent applications.  But, we know part of the secret is a dramatically reduced ultrasonic dynamic range, as he has said over and over. Check out that MQA Origami video, with its A, B and C portions of the frequency spectrum.  It is not great, but it is a start.

 

It is a matter of who you wish to believe, mansr or Stuart.  Don't believe me.   I am willing to listen to the evidence wherever it goes, but mansr has not convinced me.  As I said, mansr's claims are cheap, Stuart's are huge and existential if he is defrauding us. As I also said, this issue is one of provable, measurable substance, unlike typical audio marketing claims, which are unverifiable.  

 

 

[Ralf11]

I am willing to believe everybody, as I rarely venture beyond Redbook.  It's a desert out there...

[kumakuma]

54 minutes ago, Fitzcaraldo215 said:

As I also said, this issue is one of provable, measurable substance, unlike typical audio marketing claims, which are unverifiable.  

 

Yes, but as this thread clearly demonstrates, there will always be individuals who believe the marketing claims over the measurements...

[AJ Soundfield]

1 hour ago, Ralf11 said:

I am willing to believe everybody, as I rarely venture beyond Redbook.

Well, there's no such thing, since 44.1k filters aren't real and what's left is totally smeared

[Rt66indierock]

2 hours ago, Fitzcaraldo215 said:

Mansr says no, but Stuart says yes, repeatedly, over and over, about the entire frequency spectrum up to a 384k sampling, not just a piece of it.  Stuart's secret sauce is proprietary, and it is not even in patent applications.  But, we know part of the secret is a dramatically reduced ultrasonic dynamic range, as he has said over and over. Check out that MQA Origami video, with its A, B and C portions of the frequency spectrum.  It is not great, but it is a start.

 

It is a matter of who you wish to believe, mansr or Stuart.  Don't believe me.   I am willing to listen to the evidence wherever it goes, but mansr has not convinced me.  As I said, mansr's claims are cheap, Stuart's are huge and existential if he is defrauding us. As I also said, this issue is one of provable, measurable substance, unlike typical audio marketing claims, which are unverifiable.  

 

 

 

As always watch the video. Up to 20kHz nothing is folded. This is region A. Region B is above 20kHZ to 50KHz. This is where a small component of the music is that impacts region A. At about 1:30 in Bob says that there is no music above 50kHz. Region C is in the video from above 50kHZ to 96kHz. Region C is important because it allows the D/A converter to run faster but there is no music up there just noise.

 

Region C is folded under region B then region B and C are folded under region A. Since nothing is happening above the noise floor in Region A no unfolding occurs. Regions B and C are unfolded. Two regions are unfolded when the MQA file is decoded.

 

This is consistent with what Bob said at the Los Angeles Audio Show seminar. I agree with him that there is no music above 50kHz. His statements are similar to the reasoning and research for the first digital recording systems in studios that went to 50kHz in the late seventies.

 

If there is no music above 50kHz then what is up there is of little significance if its purpose is to allow the D/A converter to run faster.  

[mansr]

3 minutes ago, Rt66indierock said:

Region C is folded under region B then region B and C are folded under region A. Since nothing is happening above the noise floor in Region A no unfolding occurs. Regions B and C are unfolded. Two regions are unfolded when the MQA file is decoded.

There is only one (un)fold. Your region C is simply discarded. The renderer replaces it with aliases of A+B and dither noise.

[Rt66indierock]

23 minutes ago, mansr said:

There is only one (un)fold. Your region C is simply discarded. The renderer replaces it with aliases of A+B and dither noise.

 

I'm only relaying what the MQA video said and what Bob said at the LAAS. That I agree with him that there is no music above 50kHz and what is above 50kHz is of little significance. 

 

You can say what the render does today, but I can't due to my professions confidentiality rules and NDAs of the people who I've talked with. I accepted a couple of invitations that will not be covered by either in July. But in the MQA Ltd Group Strategic Report approved by the board April 6, 2017 notes that approved decoder availability is an issue. You noted the decoder is pretty generic that leaves the renderer as an issue.

[AJ Soundfield]

10 hours ago, kumakuma said:

 

Yes, but as this thread clearly demonstrates, there will always be individuals who believe the marketing claims over the measurements...

[AJ Soundfield]

7 hours ago, Rt66indierock said:

 

I'm only relaying what the MQA video said and what Bob said at the LAAS. That I agree with him that there is no music above 50kHz and what is above 50kHz is of little significance. 

 

Bob had told us, "No A/B demo can be done effectively in 45 seconds, or against prejudice, or on the basis of one trial. Frankly, it makes it a non-event.

"According to ECG measurements around content with and without components above 20kHz, the brain needs 100–200 seconds to process information before it can effectively move from A to B. We at MQA infer this is primarily a response to more or less temporal smearing, and therefore probably also applies to comparisons between grades of higher rate content." (footnote 1)

 

Footnote 1: In subsequent discussion by email, Bob referred us to four papers, one of which can be found here. The others are "High-resolution music with inaudible high-frequency components produces a lagged effect on human electroencephalographic activities," by Ryuma Kuribayashi, Ryuta Yamamoto and Hiroshi Nittono (Clinical neuroscience, 2014); "Multidisciplinary study on the hypersonic effect," by Tsutomu Oohashi, Emi Nishina, Manabu Honda (International Congress Series 1226), and "Inaudible High-Frequency Sounds Affect Brain Activity: Hypersonic Effect" by Tsutomu Oohashi, Emi Nishina, Manabu Honda, Yoshiharu Yonekura, Yoshitaka Fuwamoto, Norie Kawai, Tadao Maekawa, Satoshi Nakamura, Hidenao Fukuyama, and Hiroshi Shibasaki (The American Physiological Society, 2000).

 

Ummm, do you know what all those papers he referenced say about >50kHz content? 

Did he really say that?

[asdf1000]

51 minutes ago, AJ Soundfield said:

 

Ummm, do you know what all those papers he referenced say about >50kHz content? 

Did he really say that?

 

I only have access to that one linked and don't understand much of it anyway, but saw this: "high-resolution audio with inaudible high-frequency components induces a relaxed attentional state without conscious awareness."

 

So MQA agrees with parts of what is referenced but disagrees with other parts?

 

[AJ Soundfield]

12 minutes ago, Em2016 said:

 

I only have access to that one linked and don't understand much of it anyway, but saw this: "high-resolution audio with inaudible high-frequency components induces a relaxed attentional state without conscious awareness."

They all say >20k is inaudible, but may (still speculative) cause a more "relaxed" state.

IOW, absolutely nothing like MQA sound quality claims.

 

Quote

So MQA agrees with parts of what is referenced but disagrees with other parts?

That would depend on degree of desperation.

With regards specifically to the purported >50k statement made, here in a nutshell

http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0095464

Quote

As for the positive hypersonic effect, HFCs at around 80−88 kHz induce the maximum activity of Alpha-2 EEG. Such frequencies are within the ultra-high frequency domain, which is far beyond and not contiguous to the 20 kHz upper limit of the human audible range. The authors had not anticipated that human brain activity would sharply respond to such ultra-high HFCs. Furthermore, the application of even higher HFCs, such as 96–112 kHz or even over 112 kHz, which are extremely faint in power, also increased Alpha-2 EEG no less than did HFCs of 40–48 kHz and 48–56 kHz. Such data imply the existence of unknown human sensitivity to high frequency air vibrations, which may further contribute to discussion in the basic neuroscience field in light of the discovery of the hypersonic effect

[asdf1000]

29 minutes ago, AJ Soundfield said:

They all say >20k is inaudible, but may (still speculative) cause a more "relaxed" state.

 

Thanks. Does this mean it's impossible to say who is right or wrong about any of this stuff (importance/effect of >20k), at present?

 

[Rt66indierock]

1 hour ago, AJ Soundfield said:

 

Ummm, do you know what all those papers he referenced say about >50kHz content? 

Did he really say that?

 

Yes he did at about 1:30 into the MQA is Oragami video.

[AJ Soundfield]

5 minutes ago, Em2016 said:

Does this mean it's impossible to say who is right about any of this stuff (importance/effect of >20k), at present?

 

No. MQA/BStuart (and cheering section) claim it is audible and every study he cites says its inaudible. Nothing ambiguous there.

Now if one is unconcerned with audibility and want to alleviate the "tension" and other drama caused by the sight of Redbook, then the cited studies say you should apply band filtering >20k ~ 32k to avoid a "negative" effect, then apply a lot of power >32k, especially around 84kHz, to a rather super supertweeter.

Of course, if one is in good mental health, one could conclude this Hyperbolic Effect business is all but a tempest in a teapot and hit play on that 16/44 player. YMMV. 

[asdf1000]

16 minutes ago, AJ Soundfield said:

Of course, if one is in good mental health, one could conclude this Hyperbolic Effect business is all but a tempest in a teapot and hit play on that 16/44 player

 

Eep! I hope this doesn't turn into a general (non-MQA related)  "24/96 is vaporware" thread.

 

[AJ Soundfield]

37 minutes ago, Em2016 said:

"24/96 is vaporware" 

Eh? That has been the near minimum standard of encoding for 2 decades or so. It's quite real..and sensible.

Now, it's also real that 70+ yr old audiophiles with say large panel electrostats like MLs, will claim to be able to hear >20k content (despite all evidence to the contrary) and be "limited" by 16 bits as offered by >16/44 playback files.

So then the questions are, ok what is your rooms broadband noise floor, let's see what happens when you apply >16bits of dynamic range + noise floor to those stats, specifying exactly what recordings used, that presumably have >16bits dynamic range. Maybe post on Youtube, the arcing, explosion and smoke.

[Sal1950]

2 hours ago, AJ Soundfield said:

They all say >20k is inaudible, but may (still speculative) cause a more "relaxed" state

 

Mark Waldrep lost his seat on the CEA  (now CTA)  for not supporting their definition of HDA. In short they supported putting the HDA logo on any file created with a data rate of higher than Redbook. Thus allowing Edison's original "Mary had a little lamb" to be transferred to 24/96 and called High Definition Audio along with any other audio ever made regardless of provenance.  Mark believed that the HDA logo should be restricted to original material created and released at higher than Redbook.  When asking why the CEA insisted to slap a HDA logo tag on any file over 16/44 he was told privately "because it's good for commerce".

No wonder today there is still such debate over the audibility of HDA