MQA deep dive (and why it's a load hogwash)

[james.miller]

Veery interesting couple of videos posted by GoldenSound on youtube taking a look at MQA, and what happens when you start asking too many questions...

Grab a coffee (or tea, you monsters ), sit back and listen to it 'unfold' heh, pun intended.

tl/dr: smoke and mirrors baby!

 

[Housey]

You have an entire industry built on subjective opinion pretty much so this is no surprise. Claims of "doing it this way is better because we think so" is common. MQA has never really sold it's value to me, but I won't pretend to be an expert (or want to be).

 

[jpaul]

like subjective codecs aac, mp3, sbc(car bluetooth) ? ... earlier thread discussion on the video ,
the 96K unfold sounds good to me on home system , it's a pity that current car AA/CP systems incompatible.

https://www.overclockers.co.uk/foru...heir-lossless-quality.18926496/#post-34726556

interesting , in the world of psychoacoustical encoding aac/mp3 ... , i'm not sure you can objectively measure what sounds best to the human ear;
would science pick aac over mp3 ? equally netflix optmising encoding of the parts of the image that are human faces improves people perceptions

i had coincidentally (other tidal thread in hifi) read this blog on a human experiment setup http://archimago.blogspot.com/2017/09/mqa-core-vs-hi-res-blind-test-part-ii.html#more
which is ambivolent on whether mqa is better;
I am, however looking for a quobuz high def audio trial when my tidal trial finishes.

 

[Mr_Sukebe]

I have to say that I'm surprised at the level of "noise" about MQA.
Some thoughts:
- Yes I've tried it, for 6 months on Tidal. Have to say, I personally rather rated it, but clearly not enough to renew my subscription. I was software decoding within Roon
- Lossy or not - I did think that this was identified that the format was "lossless" under 20khz, and "lossy" above that. What I've seen written is MQA people arguing it's lossless and the detractors claiming it's not. Clearly, it's a bit of both, depending upon the frequency
- It's reason for being (ie. reduced bandwidth) really doesn't exist anymore. There was a point a few years ago, when data bandwidth was move expensive
- The "size" of the data. Again, both sides over arguing. Yes, it takes roughly double the space of 16bit FLAC, but it also takes MUCH less space that 24 bit music. After that, it's a question of whether you think it sounds better than 16 bit or not. As mentioned, I did think that it did
- The cost of the licensing. Quoted by the detractors as a HUGE issue, but never actually defined. Is the cost really an issue? Until someone can confirm, we're just not going to know.
What I find interesting is how this is supposedly such a massive issue. After all, it was OK for Dolby to license and sell everything from Dolby B noise reduction on tape, through DD, TrueHD and now Dolby Atmos, and I can't believe they're free. So why is it such a bit deal for MQA. Put it another way, did Sony make the SACD format free, or was licensed?
- Not possible to recreate the original file. Of course it isn't, as it's partially lossless, so why mention it? Again, this style of presentation exists in other formats, whether that is SACD, Pioneer with it's "legato link" and similar.

In short, as a format, personally I don't see the issue with it. Sounds good to me and the rest of the "noise" from detractors appear to be hyperbole and creating noise out of nothing particularly important.
For all that, and to re-iterate, seems a bit pointless now that we can happily stream 24 bit music.

 

[jpaul]

For all that, and to re-iterate, seems a bit pointless now that we can happily stream 24 bit music.

have you saught out a lossless 24bit supplier - which one ?

for mobile connection like cars, where you are at the mercy of 4g, bandwidth is still an issue -I think you can still discern the MQA benefit in that environment;
so that market, would still justify good lossy codec development.

 

[Mr_Sukebe]

have you saught out a lossless 24bit supplier - which one ?

for mobile connection like cars, where you are at the mercy of 4g, bandwidth is still an issue -I think you can still discern the MQA benefit in that environment;
so that market, would still justify good lossy codec development.

I tried Qobuz for a while and also amassed some locally stored material from mainly HDTracks.

 

[james.miller]

- Not possible to recreate the original file. Of course it isn't, as it's partially lossless, so why mention it? Again, this style of presentation exists in other formats, whether that is SACD, Pioneer with it's "legato link" and similar.

Because when MQA was released, they claimed it was lossless. They've only very recently backtracked on this, doing so on the quiet then denying they ever made that claim. Now, it's "better than lossless" Is MQA Lossless?

Sounds good to me and the rest of the "noise" from detractors appear to be hyperbole and creating noise out of nothing particularly important.

I dont understand this. GoldenSound saw and heard problems with the converted MQA files that were not present in the originals and MQA the company couldnt and still havent been able to explain any of it. This is a big deal, why do you think it's hyperbole?

 

[jpaul]

As Tidal replied the codec wasn't designed to encode the artificial, step impulses( if I remember right) he trojan horsed into his music submissions.

Everyone knew it couldn't be really lossless, Flac's mathematically proven, I thought, as the minimum bandwidth, so if you have less, it's lossy.

 

[james.miller]

As Tidal replied the codec wasn't designed to encode the artificial, step impulses( if I remember right) he trojan horsed into his music submissions.

He submitted another track for testing but Tidal removed everything he uploaded before he got a change to test it. He also reached out to Tidal and asked them to encode the 192khz flac version of the audio for the second video which is all voice and nothing else. Tidal has so far not done so.

 

[Mr_Sukebe]

He submitted another track for testing but Tidal removed everything he uploaded before he got a change to test it. He also reached out to Tidal and asked them to encode the 192khz flac version of the audio for the second video which is all voice and nothing else. Tidal has so far not done so.

I'll say it again, IMO BOTH sides have "overclaimed" to the point that I simply can't take either side particularly seriously.
As such, I tried it and made my own mind up.
What I'm not going to try to do is convince anyone else (either way).

 

[james.miller]

I dont understand why or how you think GoldenSounds has overclaimed. What claims of his do you take issue with?

GoldenSounds didnt set out to make any claims, he intended to test MQA's claims, and what he found didnt support them. He isnt the only one of course, and he's linked to a statements from a number of companies who also do not support MQA claims, such as Schitt audio and Linn.

 

[Mr_Sukebe]

I dont understand why or how you think GoldenSounds has overclaimed. What claims of his do you take issue with?

Did I mention anyone specific? Don't think so.
I really don't understand why a small number of people feel so motivated to be so noisy about it.
Don't like it, don't use it. There's plenty of other services besides Tidal.
And after all, it's not that I'm paying for it myself, so clearly I have no skin in the game.

The key point is that groups including Schitt and Linn do have skin in the game, which is why I don't trust their comments any more than MQAs.

 

[james.miller]

Did I mention anyone specific? Don't think so.

Ok, I don't understand your hostility. Since this thread is about GoldenEars and his investigation in to MQA, who else would you be talking about when you said 'both sides overclaimed'? I asked you which claims you took issue with and.. You refuse to tell me, instead I get a sarcastic 'Did I mention anyone specific? Don't think so', which suggests you don't take issue with anything GoldenEars has said in the two videos because well, what's the alternative? You either do or you don't. If you have 'no skin in it' then why the hostility? Bizarre.

The key point is that groups including Schitt and Linn do have skin in the game, which is why I don't trust their comments any more than MQAs.

Actually i'd argue the opposite. What have Schitt got to gain by not supporting MQA? and what about PS Audio? they literally said the only reason they did implement MQA decoding in their DACs is because the customers asked for it. They don't like it either. Are you discounting them also because they're in the industry? Soo...what or who would it take for you to drop the defence and listen?

 

[Mr_Sukebe]

James>

I'll refer you back to my initial post, such that I don't have to repeat what I've already stated.

As for the article. Sorry, but I just see that as another "influencer". They exist on both sides and as mentioned, I don't really fully trust any of them.

Ref Schitt and Linn, by taking the stance that they have, they can:
- Avoid the cost to update/redesign their existing kit to incorporate MQA tech
- They deliberately position themselves as having a "stance", which as can be seen, has resulted in awareness of their brand increasing. As the old saying goes "all marketing is good marketing"
So both of the above have very solid reasons for avoiding the tech.

BTW, did you notice that Linn have pulled the plug on their "view" on the subject:
https://www.linn.co.uk/blog/mqa-is-bad-for-music


I'll re-iterate one point from earlier. I don't fully trust EITHER side. It very much has the potential for being another audio "snake-oil" solution. Having said that, how many times have you seen people argue (often in MQA threads) that hires music (24/96) brings no benefits, and I like that too.
Hence why I spent some time to conduct my own testing on MQA.

In short, try it. If you like it, use it. If you don't, ignore it.

 

[jpaul]

BTW, did you notice that Linn have pulled the plug on their "view" on the subject:
https://www.linn.co.uk/blog/mqa-is-bad-for-music

thanks hadn't read their take
https://web.archive.org/web/20200923135223/https://www.linn.co.uk/blog/mqa-is-bad-for-music
hadn't realised apple lossless was opened up, not sure if that includes airplay,
they don't seem to mention AptX either, which must be categorised similarly, but maybe Linn reject(ed) that too.

 

[james.miller]

I'll refer you back to my initial post, such that I don't have to repeat what I've already stated.

Whatever.

BTW, did you notice that Linn have pulled the plug on their "view" on the subject:
https://www.linn.co.uk/blog/mqa-is-bad-for-music

No, they didnt.
https://www.linn.co.uk/uk/blog/mqa-is-bad-for-music-heres-why

This pretty much sums up your views from what you've posted here:
- You like MQA, even though you have no reference to determine whether MQA was actually better than the studio masters or not.
- You think anybody who has something negative to say about MQA is 'overclaiming' but are unwilling to talk specifics in any capacity. You will not tell me what claims or by who, leaving very little to discuss with you, instead you just keep repeating that you wont...keep repeating yourself. Defensive / hostile without reason.
- You claimed that MQA was a mixture of lossy/lossless (techniques). which is fine. You mentioned it being lossy over 20khz or thereabouts however this is ignoring the fact that MQ stated multiple times that the folding process was lossless. Now that's been disproven, MQA have changed their website from reading 'lossless' to 'better than lossless'. I said this and you have not acknowledged it. TL/DR? MQA made unsubstantiated claims, got caught, backtracked and you are treating it as though it's always been lossy....which it has, but that's not what MQA claimed. You are ignoring this and talking about 'the other side overclaiming', trying to direct the conversation towards discrediting those people who want to speak up instead.


That, is mind boggling.

 

[Mr_Sukebe]

James>
To respond:

1 - You like MQA, even though you have no reference to determine whether MQA was actually better than the studio masters or not.
A - Yes, it's called having an opinion. No it wasn't a DBT, rather a comparison of tracks that I already have on CD, vs their equivalent in MQA. Could they be different? Yes, but again, it's an opinion.

2 - You think anybody who has something negative to say about MQA is 'overclaiming' but are unwilling to talk specifics in any capacity. You will not tell me what claims or by who, leaving very little to discuss with you, instead you just keep repeating that you wont...keep repeating yourself. Defensive / hostile without reason.

A - What you did was post a link to an article who might aswell be labelled as an "influencer". I've seen and been involved in MQA discussions since it was first inbound and frankly I'm tired of the debate. All I see are the same comments regurgitated against it, some of which has some genuine foundations, and some of which are clearly just "noise" and "speculation" as a certain person on the PF forums seems to love doing.

You also seem to have ignored my comment that there's clearly bs on both sides. We all know that some data is thrown away, and the remainder has some MQA specific filtering applied. So quite clearly it's not going to be possible to recover that back to it's original 24 source file. So it's bs from MQA about it being lossless. Having said that, apparently it IS lossless below 20khz, so it's not fully correct to say that it's a fully lossy file.

I'll say it, I do NOT fully trust the arguments from either side, I think that there's simply too much money at play for either side to be considered unbiased.

What I find mind boggling is your seeming insistence that all of the rest of us should automatically agree with your link to another's thread.
Sorry, but that's just not going to happen. I'll make up my own mind thanks based upon the level of testing that I'm happy with.
I assume that when you bought your audio gear, that it was most probably based upon listening comparisons, just as mine was, as we both know that audio "specifications" can sometimes been rather misleading. So why not apply the same logic to MQA?

 

[james.miller]

1 - You like MQA, even though you have no reference to determine whether MQA was actually better than the studio masters or not.
A - Yes, it's called having an opinion. No it wasn't a DBT, rather a comparison of tracks that I already have on CD, vs their equivalent in MQA. Could they be different? Yes, but again, it's an opinion.

Lack of objectivity. And you know full well comparing cd audio to MQA tells you nothing beyond which format of that particular track you prefer. It doesnt tell you anything about the masters, it doesnt tell you if the MQA track is even as good as the studio master, let alone 'better than lossless'.

A - What you did was post a link to an article who might aswell be labelled as an "influencer". I've seen and been involved in MQA discussions since it was first inbound and frankly I'm tired of the debate. All I see are the same comments regurgitated against it, some of which has some genuine foundations, and some of which are clearly just "noise" and "speculation" as a certain person on the PF forums seems to love doing.

I've already asked you which points GoldenEars made that you objected to and you refused to answer. Don't then tell me it's noise and speculation if you aren't willing to be challenged. Nothing you have said refutes anything reported in the videos.

You also seem to have ignored my comment that there's clearly bs on both sides.

Have I? You really think that? Despite me asking you in every reply to explain the bs from this side, you're now claiming I am ignoring you?

So it's bs from MQA about it being lossless.

Oh, now you acknowledge the lies. very good - we're getting somewhere.

Having said that, apparently it IS lossless below 20khz, so it's not fully correct to say that it's a fully lossy file.

Which is amusing isn't it? because again that is not what Bob Stuart said.

reminder: Bob Stuart explains how the 24-48khz band is folded under the 0-24khz band (losslessly!)

skip to 1m49s.

Audio origami

What I find mind boggling is your seeming insistence that all of the rest of us should automatically agree with your link to another's thread.

Given your complete refusal to talk about any point he made, that's completely ridiculous. You like MQA beyond objective reasoning and wont be questioned on it. I get it. move along then.

 

[Mr_Sukebe]

I actually spent 20 mins watching most of the first of those videos.

Maybe I've missed something here.
MQA have described at least at a high level their process. Here's a link that you might find interesting:
https://darko.audio/2016/06/an-inconvenient-truth-mqa-sounds-better/

And some of the key content:

According to Bob Stuart, the third and most important strand of the MQA trifecta takes place before the music even reaches the end user’s DAC, the download store or streaming service.
MQA isn’t a codec, it’s a process. And that process begins at the studio. The MQA algorithm corrects time domain inaccuracies (pre- and post-ringing) caused by the original A/D converter. This is either done at the studio, soon after the recording is made or, for existing recordings, with one of the handful of MQA-equipped computers dotted around the globe.
The implication being that MQA can potentially correct the temporal smearing of almost every recording in existence. Success here largely depends on whether the original A/D converter’s make/model is known. If it is known, precise settings can be applied, otherwise a best guess filter is applied.
Bob Stuart explains: “What is happening here is that the encoder (using system metadata and/or AI) resolves artefacts that are obviously different in each song according to the equipment and processes used. When these distracting distortions are ameliorated then the decoder can reconstruct the analog in a complementary way.”
“Removing ‘pollution’ is not lossless” [my emphasis].

To paraphrase the above, during the MQA encoding process, and BEFORE the data is compressed, some MQA voodoo witchcraft is applied, meaning that quite clearly, if you stuff a 16bit FLAC file in one end of he sausage machine, that you will NOT get the same out at the other end, as the MQA encoding is very much CHANGING the file.
Whilst the "claimed benefits" may well be utter BS, that very much explains why the video that you posted was a complete and utter waste of time, as it's pretty explicit that MQA encoding is NOT just lossy data compression, before that, the data is being changed.

 

[FrankJH]

Thought I would come in here to read this as I have recently purchased an iFi Zen dac and some frighteningly expensive headphones (Sennheiser 660S) and a small selection of my favourite albums from HD tracks (although the best quality available for some are still only cd quality a little frustratingly) - Im interested in the MQA debate as I also have Tidal Hifi subscription.

Personally , as Mr Sukebe said, this is highly subjective - there isnt one single answer. Language translation in one sense is completely factual, its either correct or not, but its also down to local variables and may not be completely literal. The same for music and the different encoders or codecs used.

Im really looking forward to experimenting with my vast array of mp3's and much smaller Hi res (or not so) collection and seeing where I end up. Either way Im sure Tidal will keep me entertained and I look forward to hopefully discovering some new artists.