MQA is Vaporware

[feelingears]

53 minutes ago, The Computer Audiophile said:

 

Thanks! I suppose one could read that as simply Bob wishing to capitalize on his data compression tech. HOW they go about it is certainly another thing, but yeah, "exploit" would be a suitable choice of words and intent in the current state of affairs.

 

[The Computer Audiophile]

28 minutes ago, feelingears said:

 

Thanks! I suppose one could read that as simply Bob wishing to capitalize on his data compression tech. HOW they go about it is certainly another thing, but yeah, "exploit" would be a suitable choice of words and intent in the current state of affairs.

 

I hear ya. Exploit just feels a bit different when used by this company. 

[The Computer Audiophile]

Plus, it would be nice to see a company lead off with, "Company ABC was established to solve problem XYZ..." 

[The Computer Audiophile]

If you want to feel even grosser, I present the text from a couple paragraphs down. The demand isn't there of course, and it will be created through favorable relationships. 

 

Where's the part about, our tech solves a major issue to which there is high demand, or even a little demand. 

 

[Rt66indierock]

6 minutes ago, The Computer Audiophile said:

Plus, it would be nice to see a company lead off with, "Company ABC was established to solve problem XYZ..." 

 

Even Theranos was created to solve a problem.

[feelingears]

28 minutes ago, The Computer Audiophile said:

If you want to feel even grosser, I present the text from a couple paragraphs down. The demand isn't there of course, and it will be created through favorable relationships. 

 

Yuck - and I didn't want to feel more gross! And yeah, maybe they've given up obfuscating behind consumer benefits and as you say, problem solving. Just lay it out: we will manufacture a market, with the assistance of willing manufacturers (who, by the way will want something extra to market to their market!), to ensure that royalties are transferred in the most expeditious manner with an assurance of minimal third party participations.

I must say, the only reason I follow this drama is because those opposed stand in somewhat stark contrast to those in favor or those willing to lend an ear to MQA. Through the lens of business, as a consumer and target of said business, how can you possibly support this? (Rhetorical question.) The only drama better, or worse, actually, is crypto... 

 

(Wait, that's it! Put MQA on a blockchain with smart contracts to ensure royalties! That'll ensure provenance! 😉)  Okay okay, back to audio; that's contentious enough.

 

[GregWormald]

"exploit"

It looks as though that has been successful so far, at least for Meridian and the Directors.

[Skirmash]

15 hours ago, The Computer Audiophile said:

If you want to feel even grosser, I present the text from a couple paragraphs down. The demand isn't there of course, and it will be created through favorable relationships. 

 

Where's the part about, our tech solves a major issue to which there is high demand, or even a little demand. 

 

 

The following comment is based upon the assumption the bolded text is directly quoted from an MQA document. I might of missed where you expressly said it was. (EDIT: I now see where you did on re-reading after a cup of coffee!)

 

Despite my personal objection to MQA, I saw the use of the word 'exploit' as no more than just 'monetizing' the IP and that reactions to it seemed a bit on the overly-sensitive side. However, the bolded text above reframes that perspective for me back in line with my contempt for the people pushing MQA in general. The statement "..... ensure that content is experienced in its original quality with an assurance of provenance." has, according to all evidence that I have seen, proven to be marketing hyperbole at best and outright lies at worst.

 

Now that MQA knows the evidence is irrefutable, I view their obfuscation of this inconvenient truth evidence of their desire to truly 'exploit' the audiophile community in the most negative of manners.

 

Ash

 

[GregWormald]

39 minutes ago, Skirmash said:

Now that MQA knows the evidence is irrefutable, I view their obfuscation of this inconvenient truth evidence of their desire to truly 'exploit' the audiophile community in the most negative of manners.

Hey, it's only negative for the exploited. The exploiters will do fairly well out of it. Unfortunately this is the way of the greedy.

[Skirmash]

3 minutes ago, GregWormald said:

Hey, it's only negative for the exploited. The exploiters will do fairly well out of it. Unfortunately this is the way of the greedy.

I agree with the first two parts. Not sure I expect anything other than the latter (greediness) in the current climate. That said, I do expect the people who purport to work on behalf of the consumer's interests to be more altruistic.

 

This is why I am staggered by the unwillingness of certain players in the audio magazine space (who have garnered such stellar reputations on testing products from an engineering perspective) to call out MQA based upon the preponderance of evidence. Self-interest and ego are the only reasons I can see for this failure. Maybe I am missing something or just simply naïve. 

[mevdinc]

On 1/29/2022 at 4:42 PM, mevdinc said:

So sorry to hear Neil's music is no longer available on Spotify, as I've been using the service for over a year now!
Neil is one of my favourite song writers, so many great songs.
Oh well, I have almost all his stuff locally available, it's not a massive loss. But if I were to leave Spotify I'd not be going back to Tidal, that's for sure. There's always Qobuz as a viable alternative.

 

Thanks to Neil Young I stopped using Spotify and returned to Qobuz! So glad to be back to listening to high quality streaming and making much better use of my music system.
 

[Confused]

A recent review from the once great What Hi-Fi:

 

https://www.whathifi.com/reviews/ifi-hip-dac-2

 

Here is an extract:

 

And partly thanks to that chip, there's now full three-fold MQA decoding – a point that will doubtless pique the interest of Tidal HiFi subscribers. 

 

Thanks to the DAC's ‘three unfold’ decoding process, now performed internally (as opposed to just the final unfold in the manner of an MQA ‘renderer’ – the method undertaken by the original hip-dac), MQA files glow green but MQA Studio tracks are now recognised too, glowing blue. Magenta means the file is playing in its original sample rate (MQB).

 

I can't quite believe anyone is still writing this kind of stuff. (although the article is unattributed) 

[Confused]

I also spotted this one:

 

At the annual NY:LON event for music executives, Mike Jbara moderated a discussion on the values-based music economy. Joined by representatives from AIM, Family in Music and Sony Music Entertainment, the panel covered wide-ranging topics including diversity and transparency, as well as how the industry is responding to the issue of climate change.

 

Interesting, I recall delighting in Mr Jbara's moderating (maderating?) skills in a YouTube clip I watched covering an MQA presentation at RMAF back in 2018.

[UkPhil]

2 hours ago, Confused said:

I also spotted this one:

 

At the annual NY:LON event for music executives, Mike Jbara moderated a discussion on the values-based music economy. Joined by representatives from AIM, Family in Music and Sony Music Entertainment, the panel covered wide-ranging topics including diversity and transparency, as well as how the industry is responding to the issue of climate change.

 

Interesting, I recall delighting in Mr Jbara's moderating (maderating?) skills in a YouTube clip I watched covering an MQA presentation at RMAF back in 2018.

Yes it seems they are pushing the climate issue now with MQA seen as less than 10% of Tidals files are MQA and good slug of those are 16/44.1 (sorry 15/44.1) not sure there contribution is going to have much of an effect, have a word with Netflix and Amazon see if they are losing any sleep over it 

[John Dyson]

On 2/1/2022 at 3:13 PM, feelingears said:

 

Yuck - and I didn't want to feel more gross! And yeah, maybe they've given up obfuscating behind consumer benefits and as you say, problem solving. Just lay it out: we will manufacture a market, with the assistance of willing manufacturers (who, by the way will want something extra to market to their market!), to ensure that royalties are transferred in the most expeditious manner with an assurance of minimal third party participations.

I must say, the only reason I follow this drama is because those opposed stand in somewhat stark contrast to those in favor or those willing to lend an ear to MQA. Through the lens of business, as a consumer and target of said business, how can you possibly support this? (Rhetorical question.) The only drama better, or worse, actually, is crypto... 

 

(Wait, that's it! Put MQA on a blockchain with smart contracts to ensure royalties! That'll ensure provenance! 😉)  Okay okay, back to audio; that's contentious enough.

 

One frustrating thing is about all the MQA space in the market is being wasted on a big nothing-burger.   I have been trying to find new material to test/listen to.  It seems that older recordings,  recently available/distributed, the quality is becoming generally WORSE and not better -- at best, just as bad as a few yrs ago.  This is NOT about new recordings, but focusing stuff made since the middle 1950s when HiFi really started becoming HiFi, perhaps to the very early 1990s.    There have been few recordings on a certain distributor's site that havent been mangled with signal processing.  Out of the specific genre, I have found perhaps 2-3 acceptable recordings out of about 20.   I am NOT including the signal processing vis-a-vis my personal project, but instead even more, nastier, signal processing damaging the recordings.   One might excuse recent pop material which doesn't have or need great/natural recording quality.  I am not looking at recent pop AT ALL.  My focus is NOT recent pop material, but Jazz and 1900's American 'classical'. Just trying hard to find new, good copies.   More and more often, It seems that there is too much processing going on against the recordings, sometimes against the older ones whose hiss doesn't need to be made worse.  For the best listening experience, recordings should  simplify back closer to the original mix.  (Certain boutique recordings might be okay -- limited budget though.)

Why am I mentioning this on this forum?   There is so much mind power being used to create& market nonsense like MQA, where perhaps a more constructive and useful kind of signal processing would be a good thing.   For example, it seems like more and more that 'remastering' actually means 'compressed to h*ll and back'.   Since the world seems to want 'loudness wars' or 'demi-loudness wars' material, hows about some genius come up with a more gentle and natural sounding form of compression?   THAT would be beneficial.   Of course, I am assuming that the destributors strongly want to distribute 'compressed' and/or 'louder' material to begin with.   (There ARE things that can make compression a little less ugly.)   It would be best, of course, to simply 'master' while maintaing the most of the quality original recording.  Without heinous damage, all you can get is a few dB of greater loudness.  Of course, loudness wars stuff can give you >+6dB or more loudness, but why is it needed?   We have so much SNR nowadays, the loudness is simply a waste.

 

For me, it looks like it is time to look again through the used CDs in the various shops around town.  Clean recordings appear to be more and more difficult to find.

 

Why can't the rather bright person who invented MQA  have developed something that actually HELPS music?   MQA is sophisticated enough that the designer obviously knows something about signal processing -- what a waste.

 

MQA is just a big waste and if anything, destructive. Continued pushback is important.

 

 

 

[ssh]

"We have so much SNR nowadays, the loudness is simply a waste."

 

Thank you.

[Iving]

1 hour ago, John Dyson said:

It seems that older recordings,  recently available/distributed, the quality is becoming generally WORSE ... focusing stuff made since the middle 1950s when HiFi really started becoming HiFi, perhaps to the very early 1990s ... mangled with signal processing ... nastier, signal processing damaging the recordings ... Just trying hard to find new, good copies ... For the best listening experience, recordings should  simplify back closer to the original mix.

 

Different CD instances of the same pre-1980 music leave me scratching my head - not just to discern the most listenable amongst them (weeding out "loud" instances easily) but also wondering which might be least messed-about-with as a digital master. Which is "closer" to the master tape. I may say to myself "the one I like" - or "the one that reminds me most of the analogue product". But who knows.

 

1 hour ago, John Dyson said:

There is so much mind power being used to create& market nonsense like MQA, ...  It would be best, of course, to simply 'master' while maintaing the most of the quality original recording. 

 

Agreed! I reject mQa's assurance of provenance as a way to tackle this problem. Fundamental reasons already expressed all over this thread. Anyway - doubt there is sufficient market demand to produce what audiophiles of our kind want wrt back catalogue material.  Similar point applies to this thread ["HDR Audio"].

 

1 hour ago, John Dyson said:

For me, it looks like it is time to look again through the used CDs in the various shops around town.  Clean recordings appear to be more and more difficult to find.

 

+1

Pre 1995

[FredericV]

The life of an mQa influencer: just repeat old articles.
There's zero new research being posted. Just keep digging in the old stuff.



And looking at the title of the original article "Zen & the Art of A/D Conversion", he missed the point. Why are there no mQa enabled A/D converters except for the Mytek?

 

Quote

Brooklyn ADC is the first commercially available ADC to feature built in MQA Kernel (TM) to be used for mastering to MQA streaming format.

If this format was so revolutionary, all studio ADC vendors would want to add this feature. He clearly forgot to mention that this did not happen. Almost 3.5 years later, mQa still has no portfolio of mQa enabled ADC's.

From the same article:
 

Quote

The music we enjoy never fully occupies a rectangular amplitude/frequency channel, which is one of the ways in which we can improve on the sampling process.

This is once again about that flawed encoding triangle, which applies to most music, but not all music.
Based on Bob's own research, the above claim is false, just look at the musicbox:

[Cebolla]

Also the diagram posted has nothing to do with the highlighted 'no such thing as an MQA filter', 'end to end', 'splines', etc nonsense and can only serve to add to the obfuscation with a 'pretty' (well he & the current 17 approvers think it is) picture:

 

The irony that it only shows the structure of a full MQA decoding DAC, complete with the very real 'no such thing' MQA filter applying MQA Renderer is obviously beyond him.

[Indydan]

Is Peter Veth still shilling for MQA?! The words sad and pathetic come to mind...

[feelingears]

@John Dyson Maybe the new PGGB algo discussed/introduced here at A.S. is worth investigating for you? I don't have the DAC or compute power for it yet, but it's on my mind for the future of recordings such as you wrote about. I think I'll prefer that baked-in re-calculating solution to one that requires extensive tweaking of settings for real-time filters and transcoding. A different price to pay, I suppose.

[John Dyson]

20 minutes ago, feelingears said:

@John Dyson Maybe the new PGGB algo discussed/introduced here at A.S. is worth investigating for you? I don't have the DAC or compute power for it yet, but it's on my mind for the future of recordings such as you wrote about. I think I'll prefer that baked-in re-calculating solution to one that requires extensive tweaking of settings for real-time filters and transcoding. A different price to pay, I suppose.

Thanks for the pointer...

 

I did a check into what they are doing.   It looks like mostly an optimum or nearly so EQ scheme.  The 'decoder' that I am working on has VAST amounts of dynamic gain that undoes compression.   It consists of many, many fixed filters and 8 gain control devices.  The decoder actually *removes* noise and slightly enhances peaks.   A fixed, passive scheme can certainly sometimes help also. *as mentioned elsewhere -- BOTH have their purposes.)

 

The amount of work to do the passive EQ schemes can be done on a single core of a CPU, and with some left over.   The decoder is lucky to run in realtime on a 4 core CPU running full-out.   These are different things, each one has it's application.   For example, the decoder has probably 100 2nd order EQs, and perhaps 100-200 1st order EQs, with 8 virtual pseudo DolbyA units each doing 10-15dB of gain control.   The goal of the decode project  is VERY difficult to achieve and perhaps the singular most difficult piece of software that I have ever been involved with.

 

Back about 5yrs ago, I was offered a chance to emulate a Telcom C4 noise reduction unit after doing a plausible (now near perfect) DolbyA emulation, but those kinds of projects are not a challenge to me.  (DolbySR would have been a challenge, but is too complex -- on the order of the decoder in detailed complexity, with very little benefit.)   This damned decoder has been too much of a challenge, but it is getting closer and closer to success.   Frustratingly, measuring the program objectively without specs, without architecture, etc -- has caused a lot of embarassment.  Even my minimal objective measures to determine behavior are only approximate.  (The measures must be driven by program material -- cant do it with a sine wave like a conventional frequency response measure.)  Measuring frequency response/flatness is impossible using conventional techniques.

 

The only reason why the decoder project hasn't been given up is simply because I have infinite patience, infinite persistence, and don't give up when starting a project.   This is the reason why I got 'interesting' projects back at AT&T Bellr Labs -- I really try hard!!!

Success is imminent, but even success wont' solve the problems with today's recordings, only those up to about 1992 or so.

 

 

Thanks again!!!

 

John

 

[feelingears]

7 hours ago, John Dyson said:

The 'decoder' that I am working on has VAST amounts of dynamic gain that undoes compression.   It consists of many, many fixed filters and 8 gain control devices.  The decoder actually *removes* noise and slightly enhances peaks.

 

 

If this worked, as in sounded subjectively "real" in  most all circumstances, it would be amazing. I hear the ever so slight dynamic deltas between the addition of, say, a bit of silver in my cabling, but what I really want of course is for the dynamic range to bring the energy of the whole band/orchestra/arrangement coming in to punch me...well, more than it does now, which regrettably is "not much" as you note. Good luck and thanks for your persistence!

[John Dyson]

6 hours ago, feelingears said:

 

If this worked, as in sounded subjectively "real" in  most all circumstances, it would be amazing. I hear the ever so slight dynamic deltas between the addition of, say, a bit of silver in my cabling, but what I really want of course is for the dynamic range to bring the energy of the whole band/orchestra/arrangement coming in to punch me...well, more than it does now, which regrettably is "not much" as you note. Good luck and thanks for your persistence!

Take a listen to the current snippet demos...  Perhaps a version or two (perhaps 1 month) away from being professional/commercial quality.   Don't listen for an extreme difference, but the difference between the original material is subtle, but very real (less hiss, more clean transients, tighter sound.)   Currently, there is a bass problem where it sounds slightly 'stunted' (slight resonant bass at 20Hz, can be a bit irritating), and the very highs are a little rough.  Solutions are known, just there are sometimes errors even in recent releases...   The next release, with these problems fixed, will be in approx 1-1.5wks, and will be the final 'RELEASE5Final'.   The project is very dependent on the reviewers/testers -- was clueless for a couple of years that my hearing has not been reliable, therefore caused a LOT of embarassment and some frustration, even by those who had been interested.  It changed by >10dB at 10kHz and even below 100Hz throughout the day.  The only thing that stable was between 200Hz and 2kHz (now taking steroids, which are really helping.   Still, my hearing is not trustworthy. )

 

The design is based on standard values (+-3dB/+-6dB/+-1.5dB, standard freqs 25Hz,50Hz,75Hz,125Hz,... 3kHz, 6kHz,9kHz...27kHz,30kHz, standard Q: 0.577,0.8409, 1.19, 1.414, 2 - not 0.7071 ?!??!), very few 'tweak' values.  Most tweaks (about 5 or 6 out of approx 500 EQ settings) are related to DolbyA type characteristics and the small differences between the DA decoder emulators and true HW DolbyA units.  

 

Here is the pointer to the demos (about 70 snippets) -- getting closer!!!  I haven't announced these yet, but was planning to do so today anyway.

 

Look for flacsnippet-RELEASE5F-V5.9.9G+8-0: (downloads are best, the dropbox online player sucks -- just be forewarned.)

https://www.dropbox.com/sh/i6jccfopoi93s05/AAAZYvdR5co3-d1OM7v0BxWja?dl=0

 

Further discussion should probably be private (PM) or on the 'FA' thread -- I think that we all have great respect for this MQA thread,  we shouldn't pollute it any further.

 

[dericchan1]

Decided to give mqa another spin. Causal listening to this album while playing with my phone. Could not stand it after the third song - Barcelona and had to switch back to redbook 16/44. Yes it was that bad Freddie sounded like some random guy I didnt recognize his voice anymore