I saw a number of points/questions written about things that I'd see referenced in MQA's material (including how they treat the various ranges being discussed on this page). So I figure, why not go straight to their videos on YT and see what they have to say, so we're talking apples to apples, here, and in a format that is easily readable and provides direct quotes.
Transcribing some of Bob Stuart's answers from a recent YouTube video posted May 1, 2021:
MQA (Master Quality Authenticated) explained (sponsored): youtube.com/watch?v=8L8Vo8_gKQQ
(Bob Stuart)
06:03+:
7:29+
8:21+
14:02+
MQA – An Introduction: youtube.com/watch?v=3i69U69pqu0
(Spencer Chrislu)
What is MQA?
What is the science behind MQA?
What is blurring?
What makes MQA such a small file?
What is authentication?
MQA Music Origami: youtube.com/watch?v=BrgjycGhoSM
(Bob Stuart)
MQA - Ringing and Filters: youtube.com/watch?v=drv9ESli5yI
(Bob Stuart)
From MQA's previous version of "Is MQA Lossless?" https://web.archive.org/web/20210515030602/https://www.mqa.co.uk/newsroom/qa/is-mqa-lossless
It has been since updated to:
With the elaboration:
Transcribing some of Bob Stuart's answers from a recent YouTube video posted May 1, 2021:
MQA (Master Quality Authenticated) explained (sponsored): youtube.com/watch?v=8L8Vo8_gKQQ
(Bob Stuart)
06:03+:
This is one of the reasons why many people are confused about MQA, because we're not just doing one thing. We're not a codec, we're not high-res delivery, we're actually thinking about the whole problem from the studio all the way to your home. How do we capture it? How do we make it right? How do we make sure, for the benefit of the artists, that it's actually arriving? And then how do we allow you, as a user, to play it back everywhere you want to hear it? On a computer, on a phone, in your car, on a hi-fi, and so on.
7:29+
People will talk about the amount of data, the size of a file, and data rate -- and the traditional way with high-resolution is just to make the file bigger and bigger and bigger, and put more space in it, right -- but a lot of that is wasted, and the real skill and the real trick of what we've been doing is to find out where the music is in that file.
So you've got 10 times the amount you need -- in the middle, this is audio -- and then we give it more attention than the conventional method, so we'll take more effort to get it right, but not the things that that you can never hear, or that don't make any difference. So this is looking for the information inside the data. Those two things are different.
8:21+
Lucy Hedges: So what are the main advantages and differences that MQA delivers to the listener?
Bob Stuart: Clarity. Clarity, transparency, easy access to the sound. So when you listen to sound, whether it's over headphones or loudspeakers, your brain is working to interpret what's coming in, and paints a picture in front of you. You don't hear noises, you hear a guitar or a piano or a singer, right? The easier it is for our brain to do that, the better, and it's easier to do if the sounds only contain natural sounds.
In the natural world, our ears are extremely acute, and by the natural world, I mean the sounds of the real world: wind and rain, rustling leaves, snapping twigs, music, people speaking, birds singing -- these are sounds that are actually, curiously, mathematically are very similar, but if we if we introduce certain kinds of... let's call them "errors" in the system, you can get unnatural things, and your brain starts fighting you, wondering what's going on.
So you may think "This sounds good, but it's very edgy," or "very bright," or "I can't hear the room," so one of the things that MQA gives you with this clarity, is you may hear an instrument, but you might hear all the space around it. And you might hear the decays. And the great thing is it pulls you into the emotion more quickly, so great sound gives you connection with the artist, connection with the music. You may not even notice it. As we know, for example, from our partners in Tidal, that the people who are keen on many people listening to MQA are not all audiophiles, right? They're music lovers, they're people who think, "Well, I don't think about this, but it sounds great," and that's what you get. So clarity, openness, are the main things.
14:02+
One of the things that MQA has helped the whole industry to do is to make sure that we can get the data from the studio, actually, to your DAC. Because, lossless -- you think lossless would solve that, but it doesn't, because lossless is just how to move the data, but you've got to make the data, and you've got to convert it. And D2A converters are not lossless either, so this is all quite subtle, but the main point is we've kind of pushed the game forward, as well as raising awareness and interest in high resolution.
MQA – An Introduction: youtube.com/watch?v=3i69U69pqu0
(Spencer Chrislu)
What is MQA?
MQA is actually a revolutionary new British technology. It is a way of rethinking digital audio, and a way of rethinking how we capture digital audio. We hear everything in analog, and it's that conversion from analog to digital, and then back again from digital to analog, that are actually introducing these subtle yet important distortions, especially in the time domain. MQA removes those and actually gets us back to the sound of where we are in the recording studio, and that's what MQA is about: It's about recapturing that magical moment that happens in studio, and it does it tremendously.
What is the science behind MQA?
The fundamentals actually come from the neuroscience, and new research that is showing us that we are much more sensitive to sound in the time domain than sound in the frequency domain, which is a natural reversal of how we designed digital audio systems. In an audio world, resolution is really about discerning two things that happen very close together in time. And if you don't have those things right in the time domain, they smear and they get congested against one another. When you remove those distortions and actually restore it back to where it was, to what you hear in the studio, the resolution is amazing.
What is blurring?
So even when we hear things naturally outside, there is blurring. We know this when we go to see fireworks, for instance: If you're very close to it, it sounds very crisp, and pops, and when you're far away from it, all you hear is the boom. So there are frequencies that tail away no matter where we are, and there's this sort of "blurring." In the digital audio, we have exchanged that sort of analog natural-blurring effect for something that actually is very unnatural. The way that audio distorts in a conventional digital audio system... it takes that very sharp transient, actually spreads it out over time, but the worst part is this "ringing" that happens before and after. We don't hear things in nature that ramp up, have an impulse, and go like this (curves hand downward, unclear / out-of-frame). We hear things that have an impulse, an attack, a sustain, and a decay. So when these things were introduced, people were hearing things in digital audio that they had never heard before in an analog system. When MQA goes in and actually removes this pre-ring, and actually restores these transients to what they're supposed to be, the result is absolutely remarkable. This is the goal of MQA, is actually to create no more blur in the system, than a few meters of air in the analog world. And when we look at it, and we measure it, that is exactly what it does.
What makes MQA such a small file?
One of the other things in the digital world, is in the beginning we recorded everything in the studio, and we played it back at home at CD rates. The way we tried to address these things was going to a higher sample rate, so we doubled that number from 44,000 to 88,000, and there from 88,000 to 96,000, and then we doubled that again to 192,000. The effect was good, and it was real, and it was audible, but it was just sort of diminishing returns, and what we got in exchange were these very large files. "Music origami" is taking these large files, and actually taking all of the information in these large files, and folding it like you would an origami into something as small as CD. This is the file that actually gets stored and put onto your streaming services, and when it hits an MQA decoder on the end of the system, it unfolds it to its fullest high resolution, to that resolution that we heard here in the studio. So now you can take the highest resolution with you on your phone, and on your tablet, and through a streaming service, and on your home high-fi.
What is authentication?
What MQA does as well, is it can take a real signature -- a digital signature from the artist and the producer and whoever owns that record -- and sign it, and encrypt it, and lock it inside the file. And what it does again when it hits an MQA decoder is it'll light a light that tells the music fan that they're hearing the exact analog audio that their favorite artist created in the studio. And it tells the artist that that art that they've worked on for so long is getting to all of their music fans completely intact and unadulterated by anything that happened in the chain that went from them to their music fans. That's very powerful, that again a music fan will always know when that light lights up, they're getting the exact master -- they're hearing exactly what the artist heard in the studio.
MQA Music Origami: youtube.com/watch?v=BrgjycGhoSM
(Bob Stuart)
Here's how "Music origami" works: If we look at what's called a Shannon diagram, which is a picture that shows us the level of music versus frequency, you can see that there's a region that we call A on the diagram - it goes up to the normally-accepted high-frequency we hear of 20 KHz. This is like the sound that goes onto a CD.
There's another region which is above that, where we still see that the music is above the noise floor -- and the noise floor of a recording is a very important component here, because this is the sound in the recording before the musicians start. There's a sound below which we can't go. Between 20 KHz and 50 KHz, there's a small component which is very critical: That part we label B, and the thing about B is it contributes the microstructure of the sounds in the region A. We know if we take that part away, like if we filter down to a CD quality, people could hear the difference. There's scientific tests that prove this. It damages the sound, however you do it. It's very important to preserve that part.
And then there's another region which is C on the diagram, where we have really just noise. This region is important because it's the bit that allows the D2A converter to run at higher speed, but it has no music information in it. The music information is contained in A and B. Nevertheless, we don't want to take anything out of the file. We don't want to remove any of these components. So the way "origami" works is we take region C and we fold it first and bury it underneath B. We do that using a process that we call "encapsulation." And we're able to bury it below the noise, because we know that no one can ever hear this noise. This noise is inaudible, totally inaudible.
The next step, using [an] advanced digital resampling process, we fold it again and we bury the region B in the noise -- way down in the noise, under A. By doing this, we've taken C, put it under B, and then -- a lossless process, completely reversibly, lossless process -- we've buried B under A. By doing that, we end up with a file which contains three times the amount of information.
Now that we have this together, we've got a file which is 44 or 48 KHz. Inside it, it's got region A, and if we play that without a decoder, that's what we hear: we hear the better-than-CD. Buried below the noise -- way below the noise, buried as noise -- is the information from region B and region C. The decoder will unwrap this perfectly, give you exactly what was heard in the studio. Once it's unwrapped, then we put the sample rate back to where it was, we put the bitrate back to where it was, and you get the original sound restored.
MQA - Ringing and Filters: youtube.com/watch?v=drv9ESli5yI
(Bob Stuart)
From the earliest days of digital audio, we used sampling methods and reconstruction methods, and something happens, you know, when you capture the sound -- the A-to-D conversion which used "brick wall" filters. Sampling theorem mathematically leads you in that direction. The problem with the brick-wall filters is although they're perfect in frequency, they smear time.
What this means is that if a sound comes along, if a transient comes along, the output starts to build up to that transient, and then it decays from that transient. This is very very unnatural. There's nothing in the natural world it behaves like that. We snap the twig, you hear the crack, and then you hear the decay. If you hit a piano, you hear the start of the note, and it dies all the way.
So with MQA, we use the right filters, or we fix these filters at both ends. Because if we have the timing wrong, we might smear them together, and you just think "That's kind of a weird instrument." If we separate it, tease them apart... you can hear it, you can enjoy it -- and you don't even necessarily know you're doing that -- it's just that it sounds natural, and you smile, and that's one of the great things about MQA, is that it sounds natural.
From MQA's previous version of "Is MQA Lossless?" https://web.archive.org/web/20210515030602/https://www.mqa.co.uk/newsroom/qa/is-mqa-lossless
Q. Is MQA Lossless?
A. Yes.
MQA comes in a lossless (FLAC) file from the music label, so you get exactly what the creators intended.
But a lossless file is just a digital container, a box for data, and what really matters is the content!
Inside the file, MQA is very different: the audio data is higher resolution; it's cleverly packed and designed to preserve and confirm that you get full Master Quality, wherever you listen.
MQA delivers clearer sound: our encoders remove the audible ‘digital blur’ that builds up in studio production. The decoder authenticates the file, to guarantee that nobody changed it, and it maintains that pristine clarity, so you can hear the original wherever you are.
MQA is more efficient: it puts the full sound into the container without wasting or losing data.
Is it better than lossless? Yes, that's the sort of progress you should expect from the world-class team who developed lossless compression in the first place (30 years ago).
CD and so-called high-resolution PCM are 30–40-year-old technologies; they are better than MP3 that throws away audible content, but still a long way from what is heard when the recording is made.
MQA is revolutionary. MQA is the only significant leap forward in sound distribution for two decades – and it takes us way beyond the old-fashioned ideas of lossy and lossless.
It has been since updated to:
Q. Is MQA lossless?
A. It's better than lossless!
With the elaboration:
Is it better than lossless? Yes, MQA goes beyond the file and considers the entire signal path to give you a clearer sound. That's the sort of progress you should expect from the world-class team who developed lossless compression in the first place (30 years ago).