MQA: A Review of controversies, concerns, and cautions

[amirm]

The technology has its merits.

Well, no. People are saying it has no merits at all.

 

[svart-hvitt]

Well, no. People are saying it has no merits at all.

I think you are missing the point: People dislike MQA for the sum of the parts; they take a holistiske view. And they wonder too, how efficient the MQA encoding technique really is and if the MQA algorithm makes audible artifacts. Due to lack of transparency, nobody has yet been able to test MQA in a rigorous way. Lack of testing, measurements and transparency in general pisses off most researchers and scientists. I am surprised, disappointed even, that you are not put off by MQA’s lack of transparency.

 

[amirm]

Add to this lack of transparency, bully behaviour and the fact that record labels own MQA stocks.

It is *very* common in the industry to lack transparency in such matters. Commercial contracts are used with strict policies within. Some even have anti-benchmarking provisions.

One of the reasons for lack of transparency is to reduce patent infringement claims. To the extent the patent trolls don't have access to the actual technology, they can't make claims against you.

As for bullying, I see tons and tons of it in the anti-MQA camp. Heck, they have created a mob mentality against MQA.

As for record labels owning MQA stock, unfortunately that is the standard practice to break the chicken and egg problem. You create a new format and try to get it licensed to hardware companies. They ask where is the content? You go to the major labels and they want to throw you out as they have real business to deal with (getting their artist represented better on spotify or itunes). And what is in it for them for MQA to succeed? It is not like they get the royalty from hardware makers. You could solve this problem by writing a huge check in millions of dollars but then all the labels want the same money and no start-up has those kind of funds.

So the "standard" technique is to offer warrants/stock options to the first major label. They in turn produce the content for you. You then get the device guys onboard and that puts pressure on the rest of labels to get on board (gradually).

This is done by countless audio/video start-ups.

The major record labels and studios are basically a cartel with incredible market power. Breaking that power requires tools and one of the few available to start-ups is handing out warrants.

I can tell you though outside of the initial deal, it does not at all change the strategy of the content owner. It is not like they wake up every morning trying to see how they can make their start-up investment actually pay. It is a one-time deal of getting stock options instead of a check and that is that. This is why you don't see the record labels showing up front and center defending MQA. Heck, the head of the labels may have not even heard of MQA!

 

[amirm]

I think you are missing the point: People dislike MQA for the sum of the parts; they take a holistiske view.

Nope. It has become a "thing" to dump on MQA on Internet forums. Getting on board that hate wagon seems to have value to folks. Why else would Chris volunteer to give that talk? It has become a political movement of sorts.

They say what you say. But the spiteful words speak otherwise. Just look at the emotions in this thread which is 1000 times more proper than what goes on in CA forum.

 

[mansr]

Do you have any evidence of this regarding the second unfold?

Yes. I reverse engineered the "rendering" code and found a textbook upsampling algorithm. I can't publish the actual code for copyright reasons, but the bitstream format and filter coefficients are documented at https://code.videolan.org/mansr/mqa.

 

[svart-hvitt]

It is *very* common in the industry to lack transparency in such matters. Commercial contracts are used with strict policies within. Some even have anti-benchmarking provisions.

One of the reasons for lack of transparency is to reduce patent infringement claims. To the extent the patent trolls don't have access to the actual technology, they can't make claims against you.

As for bullying, I see tons and tons of it in the anti-MQA camp. Heck, they have created a mob mentality against MQA.

As for record labels owning MQA stock, unfortunately that is the standard practice to break the chicken and egg problem. You create a new format and try to get it licensed to hardware companies. They ask where is the content? You go to the major labels and they want to throw you out as they have real business to deal with (getting their artist represented better on spotify or itunes). And what is in it for them for MQA to succeed? It is not like they get the royalty from hardware makers. You could solve this problem by writing a huge check in millions of dollars but then all the labels want the same money and no start-up has those kind of funds.

So the "standard" technique is to offer warrants/stock options to the first major label. They in turn produce the content for you. You then get the device guys onboard and that puts pressure on the rest of labels to get on board (gradually).

This is done by countless audio/video start-ups.

The major record labels and studios are basically a cartel with incredible market power. Breaking that power requires tools and one of the few available to start-ups is handing out warrants.

I can tell you though outside of the initial deal, it does not at all change the strategy of the content owner. It is not like they wake up every morning trying to see how they can make their start-up investment actually pay. It is a one-time deal of getting stock options instead of a check and that is that. This is why you don't see the record labels showing up front and center defending MQA. Heck, the head of the labels may have not even heard of MQA!

OK, I guess my point is this:

You know the business world of Microsoft, and you apply this world on music distribution. Many people dislike the world of big monopolies like Microsoft; they prefer competition and free markets to more rigid and less dynamic forms of organization. When a guy from Big Corporate sees MQA, I guess he sees order and business patterns he likes, feels familiar with. Many ordinary but informed people take a more paranaoid view where they fear that Big Corporates win while the small guy loses.

 

[jtwrace]

@amirm you'll like the below along with their USB talk and measurements. Watch it all!

Have you guys seen the Schitt interview and their opinion on MQA? While I'm far from a Schitt fan it's pretty funny.

Start at 38:30 for that part

 

[amirm]

Many people dislike the world of big monopolies like Microsoft; they prefer competition and free markets to more rigid and less dynamic forms of organization.

Well, here we have them attacking a little company, i.e. MQA.

In contrast they have embraced large companies like Apple. Apple sues third-party shops for trying to fix their products for heaven's sake. Wouldn't energy be better spent to go and deal with that than MQA?

 

[amirm]

@amirm you'll like the below along with their USB talk and measurements. Watch it all!

I watched it when it was posted a couple of days ago.

 

[tmtomh]

It is a presentation issue. Again, the market for high-resolution audio is for people who think they need high-resolution audio. Any downsampled 192 kHz to 96 kHz would be scuffed at by such audience. By sampling back to 192 kHz, whatever benefit people think there is out of that higher sample rate is preserved.

There are sites that sell 96 kHz content at lower prices than 192 kHz for example. At least that is what I remember so don't challenge me on that.

You're arguing for the utility of this technology based on a "presentation" issue involving marketing of high-res audio? You might want to heed the warning of an important comment made by this guy called @amirm - you might know him. He says, "It is as if we want to wear the hat of subjectivism while using objectivism to critique it."

 

[svart-hvitt]

Well, here we have them attacking a little company, i.e. MQA.

In contrast they have embraced large companies like Apple. Apple sues third-party shops for trying to fix their products for heaven's sake. Wouldn't energy be better spent to go and deal with that than MQA?

No, you are twisting things again. Most people love small companies. However, if a small company has a technology that openly aspires «world domination» through an encoding technique, the small company may be perceived differently.

Having said that, MQA has the backing of a very big investor who knows a lot about business tactics. So this is not your ordinary «garage shop».

 

[amirm]

You're arguing for the utility of this technology based on a "presentation" issue involving marketing of high-res audio? You might want to heed the warning of an important comment made by this guy called @amirm - you might know him. He says, "It is as if we want to wear the hat of subjectivism while using objectivism to critique it."

Yes, that is what I am saying. I am being pure and pragmatic about my argument. The market for MQA content is NOT objectivists. They don't believe in high sample rate/bit-depths anyway so they are not the customer.

The customer is most likely a subjectivist. He wants to see those specs on his content. We can't wear the hat of objectivism and complain about MQA while at the same time have no use for the mathematically lossless version of high-res either.

We are just an annoying bunch of non-customers for the company. We have to be straight and honest about this. A ton of arguments about MQA reads like "if it were not lossy, we would like and support it." Facts is that "we" want to have nothing to do with high-res content in any form.

 

[tmtomh]

Do you have any evidence of this regarding the second unfold?

I apologize in advance for not providing direct links. But if you listen to interviews (on YouTube, for example) with Bob Stuart himself, he cleverly but very clearly describes this exact process. After the first unfold, he says, "we put the sample rate back to what it was." I believe that is an exact quote. What he's saying is exactly that - they upsample the file so that it then has the original sample rate. But MQA never claims - because it is impossible for them to do - that the original sample rate's data is in any way preserved.

There's also a wealth of technical MQA documentation that specifies that MQA files max out at 96k for - sorry I don't know the right word here - true/actual sample rate for the data in the file itself. Any upsampling beyond 96k is just that - upsampling that is done on the fly, during playback/streaming, simply by doubling the files' native max sample rate of 96k.

There's nothing wrong with this per se - lots of audiophile music playback software offers the option of upsampling during playback before sending to the DAC, for example. But only MQA has given the impression that this upsampling is somehow restoring the original sample rate of the file. MQA has so muddied the waters on this, that folks like me who point it out are asked for evidence of it, rather than the converse of MQA being asked for evidence that they really are "restoring" the full 192kHz sample rate. (No offense at all intended towards you with that last sentence - my view is that MQA is making things confusing for all of us.)

As for the "customized" final digital filtering not being very "custom," I will defer to @mansr who I believe has done some good testing this regard - or if it's not him, he almost certainly knows who it is I'm thinking of, as his command of the technical details is second to none.

 

[maverickronin]

Nope. It has become a "thing" to dump on MQA on Internet forums. Getting on board that hate wagon seems to have value to folks.

Like I said earlier in this thread, it's important to pile on MQA in order to keep these formats open. Lossless/hi-rez music is one of the few areas of media distribution which is DRM free and has open standards and MQA is trying to muscle in on that for the benefit of they copyright cartels.

You can't just keep pointing to even worse behavior in video distribution or something else in order to justify MQA. Yes, most of them are worse but that's been the status quo (in video for example) for 20+ years now. There's better bang for your buck in trying to keep the propitiatory bullshit from gaining a foothold in this market than there is in attempting to dislodge it somewhere where it's firmly entrenched.

 

[Thomas savage]

I’m not on a “hate bandwagon “ , I don’t hate MQA or Bob S just as I don’t hate Heinz for telling me thier ketchup is the best , marketing it to me since a I born only for me to discover the polish one from the foreign section in Tesco is better (more tomato per 100g, spiced better and more satisfying texture) plus half the price .

People get upset when they are taken advantage of , when they fear losing things they feel entitled to and generally these feelings are exasperated by knowing there’s little they can do in the face of large mutually invested cynical commercial interests .

Things then can get political as this situation mirrors the struggles of socialism for some folks, people feel disregarded and on the other ends of people’s ‘greed’ .

All that is understandable, it’s not how I feel but it’s how a few here seem to. I’d not be so ready to belittle those sentiments and motivations as some kind of hate bandwagon. That’s actually quite flippant , maybe a little insulting and will only exasperate the feelings of many and cause further divisions along unhelpful lines far from the core of the debate.

We really do need to get back to objective analysis rather than belittling folks, one can become institutionalised by their life experience and this can turn even the fairest mindeye blind.

 

[amirm]

No, you are twisting things again. Most people love small companies. However, if a small company has a technology that openly aspires «world domination» through an encoding technique, the small company may be perceived differently.

The "world domination" is a campaign talking point against MQA proving what I said. There is no basis for that in fact. Companies like Apple, Google, Amazon, etc. would never, ever give MQA a chance whatsoever to dominate the world. To say nothing of labels not wanting to put MQA in that place whatsoever.

Remember it is trivial to make a competitor to MQA. Should this be a popular thing, open-source community could create a competitor or major companies can.

 

[maverickronin]

The "world domination" is a campaign talking point against MQA proving what I said. There is no basis for that in fact. Companies like Apple, Google, Amazon, etc. would never, ever give MQA a chance whatsoever to dominate the world. To say nothing of labels not wanting to put MQA in that place whatsoever.

The label would just be using MQA as a tool.

 

[amirm]

People get upset when they are taken advantage of , when they fear losing things they feel entitled to and generally these feelings are exasperated by knowing there’s little they can do in the face of large mutually invested cynical commercial interests .

But those people are NOT the customers Thomas. They get "upset" because they see this as a battle and any counterpoint emotionally gets them going.

And commercial interests here as I explained is all in this little company called MQA. It is not like Amazon is pushing this down our throat.

Take Apple iOS and Android. Android is open-source, iOS is not. Will you give up your iPhone because of that?

 

[tmtomh]

Yes, that is what I am saying. I am being pure and pragmatic about my argument. The market for MQA content is NOT objectivists. They don't believe in high sample rate/bit-depths anyway so they are not the customer.

The customer is most likely a subjectivist. He wants to see those specs on his content. We can't wear the hat of objectivism and complain about MQA while at the same time have no use for the mathematically lossless version of high-res either.

We are just an annoying bunch of non-customers for the company. We have to be straight and honest about this. A ton of arguments about MQA reads like "if it were not lossy, we would like and support it." Facts is that "we" want to have nothing to do with high-res content in any form.

@amirm , your bolded statement above actually gets to the heart of the issue. As objectivists, we might not be interested in MQA - but the problem is, MQA most certainly interested is in us - specifically, the music files that are the entire reason that you measure the quality of digital equipment in the first place.

What do I mean by that? Well, if you want to be pragmatic, then let's be pragmatic: What on earth is the purpose of you doing all your hard work and measurements to show us that an Oppo 205 has an amazing 115dB SINAD (apologies if I'm getting the exact details wrong), which is 10dB better than some other equipment, if folks end up having to play MQA files on such equipment, and the MQA files have an effective bit-depth aka noise floor of only, say, 15-17 bits because of how they store their folded high-res content in a lossy fashion, within the first 16 bits' worth of the data?

What you need to keep in mind, again in a pragmatic spirit, is that MQA's business model is predatory, or in their words an "ecosystem." The major record labels reportedly own 21% of MQA, and MQA's own reps have been quoted repeatedly as saying the main rationale is that the labels don't want to keep letting out their "crown jewels" aka their unmolested high-res PCM files.

The goal of MQA is to transform not only the high-res music market, but more broadly the lossless music market in general, so that every content provider needs only one file - the MQA file - which can be played undecoded, or with one unfold, or with both unfolds, based on the bandwidth available and the capacity of the end user's playback device.

This is possible only if lossless digital files are MQA-encoded. That means recording studios use MQA equipment, and mastering houses use MQA encoders and DACs, and DAC/player manufacturers incorporate MQA firmware, and streaming and digital-download services acquire MQA files - and MQA gets licensing fees at every step in the process, which in turn means that the record labels get 21% of that revenue stream for themselves too. And not for nothing, it's also a revenue stream that requires no royalties be carved out for songwriters and artists.

This is in no way, shape, or form a conspiracy theory - this is simply the fact of how MQA works as a technology and business model/"ecosystem," and what MQA's aspirations are as per the company's own statements.

If MQA did not seek to supplant conventional high-res and lossless PCM then you would not be seeing all this controversy. It would be just a gussied-up, 21st century version of HDCD and nothing more.

 

[amirm]

The label would just be using MQA as a tool.

I see no sign of labels doing anything with MQA themselves. MQA itself is doing all the work and spending the money.

Labels these days have no religion/strategy with respect to formats. Anyone who shows up to get content they get it and that is that. What they do with it is their business.