MQA is Vaporware

[FredericV]

24 minutes ago, Brinkman Ship said:

If I understand correctly, MQA is lossy in that it discards bits. But as far as I know it is not lossy in the frequency domain.

Sorry to bust your bubble, but's it's lossy and not authentic to the master.
 

Here is 2L.no's 2L-053 demo track in the original studiomaster (blue plot) vs the first unfold (red plot) and second unfold (green plot):
 

The first unfold actually is true to the original spectrum but in a band limited fashion, as it stops at 44.1 Khz nyquist. 2L.no files are all multiples of 44.1 Khz and edited as DXD.

The second unfold which upsamples + dithers the first unfold just creates a totally new response not in the original.

[FredericV]

6 minutes ago, esldude said:

Hey your graph proves MQA is better than the original. There is more response in that second unfold than in the original DXD.  

...
No I don't really mean it.  

 

After seeing this response, my unfold for MQA's acronym is now Master Quack Audio.

[FredericV]

9 minutes ago, mansr said:

Are you even trying to make sense?

MQA's latest shill? Makes sense.

[FredericV]

13 minutes ago, The Computer Audiophile said:

Hi @Brinkman Ship, Welcome to CA. Or, should I say welcome back to CA or welcome to CA under a new name?

 

Based on your posts, your registered email address (mqatruth@<domain>.com), and the fact you're obscuring your physical location pretty well (IP addresses from all over the world, just in the last few hours), I'm quite suspicious of you. It appears you have an agenda and/or something to hide.

He's probably using TOR or some VPN provider with different global exit points. In tor browser bundle, you can change your exit node by just clicking on some button.

Are TOR / VPN ip's allowed to post CA? Some fora filter those IP's, some admins already filter it in the webserver, so it does not reach the forum backend.

[FredericV]

7 minutes ago, The Computer Audiophile said:

Thanks for the follow up. 

 

I'll take you at your word. Even though the signs point to something nefarious, they could equally be benign.

Why does he need to randomize his IP? It only means he has something to hide.

[FredericV]

25 minutes ago, Brinkman Ship said:

Thank you for the summary.

 

So if take at face value all of the information above is correct, why have I not read about any of this in Stereophile or The Absolute Sound, or DAR, or Audiostream?

 

So you did not read my article?

 

https://darko.audio/2017/07/kih-46-mqas-missing-link/

[FredericV]

1 hour ago, Brinkman Ship said:

So if take at face value all of the information above is correct....

 

Gives a new meaning to "face value": MQA is between 15 and 17 bits:
 

 

Uncle Bob also laughs at 24 and 32 bit in the first minutes.

[FredericV]

7 hours ago, Brinkman Ship said:

I still say that in some way MQA offers high value to listeners who have not invested in hi rez digital libraries, expensive NAS units, and hard drives. If MQA albums are starting with hirez masters and there is some bits thrown out, I am not sure that is a deal killer.

This is the classical fanboy / shill argument, also used by the now banned Peter Veth. MQA does not offer this value, but streaming services do.

You could say the exact same about Qobuz: it offers true highres (lossless, not lossy like MQA) and customers don't have to invest in their own library / hardware.

 

[FredericV]

34 minutes ago, ddetaey said:

As I am very suspious about the new member Brinkman Ship, and as I am not native english speaking, I googled the meaning of "brinkmanship"  

This is what wikipedia tells me : Brinkmanship (also brinksmanship) is the practice of trying to achieve an advantageous outcome by pushing dangerous events to the brink of active conflict.

So not only his email adres is specifically made up for participation to this topic on CA, his chosen nickname is also indicating that he is only here to create conflict.

 

Just my appreciation

 

Dirk 

I don't believe in coincidence.

Brinkman Ship is obviously a fake name, as he also uses random IP's. It would not be logical to use random IP's to hide one's true identity and location but keep his real name to post here.

So why did this member use a fake name with a known MQA vendor in it? Brinkmann makes MQA dac's.

 

Whoever Brinkman Ship is, he is most likely affiliated with MQA. Forum rules require that affiliation needs to be put in the signature.

 

 

[FredericV]

1 minute ago, Norton said:

 

In the interests of fairness I  think you too could be clearer as to your industry affiliations.

 

 

My industry affiliation is very clear from my signature. We are one of the many vendors against MQA.

[FredericV]

22 minutes ago, Dr Tone said:

It's pretty obvious, when Bob Stuart bobs and weaves when asked about deblur and all audiphile press avoid the topic in any other way than promotion, that it's not to be discussed.

Bob & weave .... sounds like de-interlacers. Didn't Bob buy Faroudja?
 

 

 

[FredericV]

11 minutes ago, beetlemania said:

 

Outside of gun-for-hire TAS and Atkinson/McGrath (whose files apparently received special attention), most others report  subtle improvement, no difference, slight preference for PCM, or coloration and loss of resolution.

 

A solution to a non-problem.

 

[FredericV]

12 hours ago, Indydan said:

TAS makes me sick. In the last paragraph of this review, the reviewer says DACs that don’t decode MQA are obsolete!

 

True bullshit!

 

http://www.theabsolutesound.com/articles/nad-c-368-hybrid-digital-dacamplifier/

 

Using the same logic as TAS, blusound must be obsolete, as it does not play DSD.

Adding dsdplay to the BS firmware can't be that hard, unless their arm cpu is not fast enough:

https://github.com/SqueezeOnArch/dsdplay

 

[FredericV]

Not sure if this was mentioned, as I could not find this URL on CA in any post:

https://www.whathifi.com/advice/high-resolution-audio-everything-you-need-to-know

Article from 19th January 2018 mentions:
 

Quote

MQA (hi-res): A lossless compression format that packages hi-res files with more emphasis on the timing. Used for Tidal Masters hi-res streaming, but has limited support across products.

MQA is not lossless, but a lossy codec which is the equivalent of 17 bits 96 Khz at best

 

Quote

 

Qobuz Sublime+ - £350/year

Qobuz strikes again here and says its hybrid download-and-streaming tier is '"the best music subscription in the world." This top tier package offers hi-res streaming up to 24-bit/192kHz files (as well as MQA and CD quality tracks) on its desktop and mobile apps. It has a 70,000-strong catalogue of 24-bit/192kHz files, and 40 million other hi-res content (everything above CD quality).

 

 

Qobuz offers the real deal (real lossless hi-res flac), so why mention lossy MQA?

[FredericV]

1 hour ago, The Computer Audiophile said:

Here's the MQA provided image for 16 bit data. It doesn't look pretty.

 

 

Furthermore, blanking 8 LSB bits from a 24 bit MQA file, so that we have 16 bits of entropy padded to 24 bit, still authenticates the file.

[FredericV]

3 minutes ago, Indydan said:

MQA "enthusiast"                        Reason for no longer posting on CA

Peter Veth                                                          BANNED

Peter Veth (as Peter Markus)  BANNED for acting like a dick and being an MQA shill

In the past Peter Veth would announce to his followers in the secret MQA facebook group which pseudonym he would use.

Here he writes "I have challenged Archimago under my Pedro pseudonym":


 

 

So I am not surprised he tries to come back.

 

[FredericV]

4 hours ago, The Computer Audiophile said:

MQA physical releases should not be sold as CDs nor have any Compact Disc logo. The only logo they should have is MQA which notifies the consumer of what s/he is purchasing. 

 

Good catch that the MQA logo isn't even there. It's advertised as CD on the link in that article.

 

 

I remember the days with DRM CD's which had a bad TOC. Any CD transport would read the first TOC.
Any computer would read the latest TOC on the CD, which had been tampered with, so ripping would go crazy.
There was special software such as EAC which could read the original TOC, and this special version was banned also.

With MQA CD, we are back in early 2000.

Philips stated that such discs were not permitted to bear the trademarked Compact Disc Digital Audio logo because they violate the Red Book specification.

https://en.wikipedia.org/wiki/Compact_Disc_and_DVD_copy_protection

Maybe Philips should do some kicking?

[FredericV]

Today I was reading this page:

http://thehbproject.com/en/articles/37/6/MQA---Quality-assured

There are some ridiculous claims:
 

Quote

However, if the file is tampered with in any way, the decoder will notice this and will not light that indicator.

In this CA reply, I basically proved Hans wrong:
 

MQA files can be tampered with, and the the indicator will still light up.

But what bothers me even more is the incorrect representation:

 

Here it is suggested there are 3 unfolds. In reality, there is no third unfold. MQA uses 24 bit files.

Bits 0..7 are used to store the first unfold. These are 8 bits which are below redbook, and when truncating 24 bit MQA files to 16 bits redbook, these bits will be lost. The MQA decoder will still fool it's users believing it can achieve 24/352.8K from these truncated 16 bit files.

Bit 8 stores the MQA control channel and should never be stripped. Bit 9..23 store the baseband part.
Bit 8..23 should never be tampered with, or the MQA light won't shine.

But you can tamper with the data that represents the first unfold, and the MQA decoder will still fool the user this could be 24/352.8

As MQA can only store 24 bits, it can't hide information below -144dB, as 24 bits is not enough to encode the part between -144 and -168. To do that, MQA would need a 32 bit PCM format. But MQA only works with 24 bit files.

Hans seems to have no clue what he is talking about. He tried to dismiss my research where I blanked 8 bits in an MQA file as gossip / fake, while I still managed to light up the blue MQA dot on my dac. I even managed to alter 1/3 of the file with garbage, and the MQA light still shined.

So it's all backwards. Hans is publishing (fake) copy paste news, which after research by people such as Mansr, Archimago and myself, is being debunked layer by layer, just like we research and debunk the claims of other MQA repeaters.

The owner of the Facebook group where we posted this is a Meridian dealer, and he even threatened to delete the post as he believed it is fake news.

So all our debunking must cause some serious cognitive dissonance in the pro-MQA camp. We basically debunk a lot of their belief systems. As someone suggested, maybe we should send flowers to MQA.

[FredericV]

Regarding the discussion if Peter Veth was Peter Markus 2 pages ago in this thread.

PV has been my main opponent some months ago, so I know the usual mistakes he makes. He never learned from my comments which corrected his mistakes, so he makes the same errors over and over again. I actually repeated his mistakes so many times he started to attack my own company, instead of learning anything. This is typical troll / shill behavior.

One of his mistakes is "24/385 resolution" which off course does not make much sense if he would understand anything about sampling and how our most common audio samplerates are all multiples of 44.1 or 48K, which 385K is not. 384K and 352.8K are. Let's ignore old sampling formats such as 32K for now.

So when googling "computeraudiophile 24/385", I find:

http://amedeffect.org/issue/absolutesound/the_politics_of_mqa

 

 

Quote

You are aware that MQA Core Decoding is an end-to-end process, so everyone should be aware of that fact. Especially when the 2L.no testbench MQA files are being used, it is impossible to unfold to the full 24/385 resolution, unless an MQA DAC is being used.

 

Looking at

 

https://www.computeraudiophile.com/profile/32490-peter-markus/

 

he indeed made the same mistake here:
 

Quote

I have seen unfolding it up to 24/385 with some of these 2L and Eudora albums mr Scogins

 

and uses his signature buzzword.

 

Quote

It is end-to-end, suppose the encoding part is more interesting.

He denies being Peter Veth. He also denies being paid by MQA. Yet he administers a secret Facebook MQA group with a lof of hifi professionals. And we have to believe this is all some kind of a hobby that's gotten out of hand?
 

[FredericV]

2 hours ago, Spacehound said:

MQA is a total scam from beginning to end.

 

In the UK, its home country, an analysis such as yours could be sent to the 'Advertising Standards Authority',  they would have it checked out, and  if correct they then would  prevent it from being advertised, promoted, and  with the help of the  'Weights and Measures' department,  sold.

But they are not allowed to instigate such actions themselves. And nobody has yet bothered to report MQA, with evidence. (It's deliberately  a lot of paperwork to discourage spurious reports, and to get them started it's best to have several  independent objectors each  supplying  evidence from separate sources,)

Those who are testing and reverse engineering MQA could provide hard evidence MQA is a scam. But nobody is being paid for this, and we have other jobs to do. Some ask: why do you waste so much time at MQA? This is a fight for our basic freedom, which we had with PCM. With real PCM we could do whatever we want with those files.

MQA is basically taking our rights away and selling them back to us. Real highres being replaced by fake highres with DRM, and no more access to the PCM.

One analogy: MQA for banks.

Hello customer. From now on, we will strip the lowest 8 bits of the binary string representing your balance. Your PCM bank account is automatically upgraded to an MQA bank account. So we steal between 0 and 255 EURO from you. However, if you pay for our upgraded MQA account at 10 euro / month, we will show you the real value of your account with a precision of 1 euro, just like in the old days. As MQA is lossy, we replace anything after the decimal point to some random number, which we call dither. But as dither is only important for technical people, we will no longer show anything behind the decimal point. As we use dither, your value may go up or down with 1 euro depending on the day and time, nothing to worry about, it's just one euro.

If you want to withdraw money or transfer money however, you must always wire multiples of 256 EURO, as we don't give you access to do anything with those 8 LSB bits. We just give you the right to see those bits and enjoy the balance of your account, but we will not allow you to use it for any other purpose.

We own the copyright on MQA bank accounts, so if you cancel your account, we will round down to a multiple of 256 euro, and keep the remainder ourselves.

Enjoy your new MQA bank account.

[FredericV]

On 3/8/2018 at 7:37 PM, Vincent1234 said:

If you're paranoid this is probably true. ? In real life my dCS has no such thing as screwing up regular non-MQA files by applying the specific MQA filtering for those files too. In practice MQA does sound easily better than Redbook, at least when streamed.

My mytek upsamples with MQA's leaky filters unless I completely turn off MQA.

My metrum only does MQA's leaky filter when MQA is being detcted (according to the lead designer) ... I have a version without MQA and I'm not jumping for the MQA module. I warned him MQA can potentially degrade his DAC design for non-MQA content.

R2R dac's  should not upsample for non-MQA content

[FredericV]

6 hours ago, Brinkman Ship said:

..Screen shot from ten minutes ago on the Stereophile site...so...rethink your position.

 

Here are dollars spent by an "MQA partner"...Cary Audio..another company added to my do not buy list...

MQA decoding featured on the top line of the feature list...

 

 

Stereophile probably does not know that this product is now an illegal product software wise.

I am on the MPD mailing list (MPD = free open music server software used by companies such as Aurender, Cary and many others - some selling it as their own by masquerading it through their own custom app), and I saw some discussions about Cary having added MQA to MPD. This is not tolerable, as you cannot add closed source software to open source in the way Cary did.

Max Kellermann who is the lead MPD dev, asked Cary to provide the source code of the binary with MQA, and Cary provided a fake redacted version twice, instead of giving the actual source that compiles into the actual binary as distributed.

So Cary's license was terminated as it violated the GPL, and they continued like nothing happened:

http://mailman.blarg.de/pipermail/mpd-devel/2018-March/000723.html

 

From a copyright perspective, this is nothing different than selling a product with an illegal version of windows copy on it. Would any magazine tolerate such ad?

[FredericV]

2 hours ago, crenca said:

Well, JI's piece comes at the problem from an interesting angle - that the "linking" of lossy compression with "deblurring" makes the SQ claims hard to evaluate.  More accurately it is the IP and NDA's, but ok.  IF this is how Stereophile has to try to save a little face and back off their overly optimistic evaluations then so be it.  He is a bit more forthcoming about the monopolistic threat.

If MQA was just an efficient folding algo, it would have a USP for streaming services, where every mbit counts. No audiophile would accept another lossy format.

So they had to invent some excuse to sell lossy to the audiophiles, the so called end-to-end argument and deblur, which has not been proven. To the contrary, all measurements we have seen by the debunkers & reverse engineers, suggests that MQA is not "authentic" at all, and by my own research, the authentication can be manipulated, and the end user can be fooled with fake resolution indication messages on the screen of his DAC - even with MQA files where a third is thrown away. The customer will never known his files were manipulated, as both manipulated and pure MQA files show the same resolution message and blue dot on the DAC.

[FredericV]

1 hour ago, Confused said:

Some interesting references to MQA here, interesting because it is not being used and 24bit is......

 

https://darko.audio/2018/03/highresaudio-com-announce-hra-streaming-service/

I was recently surfing on HRA and noticed the MQA files are even more expensive than 24/192K and 24/176.4K files.
As MQA is a degraded version with a resolution that approximates 17/96 at best, it's clear that licenses have to be paid.

Good that they are streaming the real highres in the native samplerate, and not some decimated lossy pseudo highres with the DAC lying about the effective samplerate.

It also seems the real highres files are here to stay, and will not be replaced by fake lossy highres versions with DRM.

[FredericV]

29 minutes ago, Indydan said:

It seems OK (for MQA enthusiasts like Lee) for MQA operatives like Peter Veth, to defend MQA using fake identities in forums. But, it's a big problem for these same people that Archimago remains anonymous for personal reasons.

 

Taking several identities, e.g. Pedro at Archimago's blog - how many others?

Denying he was Peter Markus on CA
Discussing his strategies on the secret group, like how he would attack Archimago
Always repeating the same shill arguments
Writing "congrats MQA team" several times

and still denying being paid by MQA for all his efforts