• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

MQA Deep Dive - I published music on tidal to test MQA

Status
Not open for further replies.

KeithPhantom

Addicted to Fun and Learning
Forum Donor
Joined
May 8, 2020
Messages
641
Likes
658
My question is: how accurate are the ultrasonics “unfolded” by MQA? Are they similar at the very least to the ones in the original recording? Also, what is the actual benefit of having those ultrasonics if nobody can hear them in the first place???
 
OP
GoldenOne

GoldenOne

Not Active
Joined
Jun 25, 2019
Messages
201
Likes
1,469
If I can give one advice (that applies for a lot of posts in this thread) : not generalize a found case.
Not all the Hifi are 16bit MQA with flagging removed, there are still lots of real 16bit FLAC


And to avoid any piracy argument on tests later, better download the free 2L tracks only, or record Tidal stream instead of dowloading.
I tested and can record part or full song in full digital, I even saw that I got all info as recording a folded MQA stream (by disabling the MQA Core on Roon for example) gives the MQA flag inside and my DAC lights up Blue or Green if I play it.

On a different point, back to the supposed prolem created by the MQA encoder that can't process correctly the files you submitted, I find it a bit strange since it was not removed automatically but once you contacted them.
I can also confirm that there were/are on Tidal server other files with test tones, and that have not be removed or only removed by their creator.
I've yet to find any instance where the hifi version of an mqa track was not simply the mqa file without flagging.

If you're aware of any I'd be happy to check. 2lno or other purchase sites are a different situation (or at least I would sincerely hope they don't have the same issue.....)
 

Grooved

Addicted to Fun and Learning
Joined
Feb 26, 2021
Messages
679
Likes
441
I'm debating. Whats missing in this discussion is an objective test of the noise added to 44.1 by downfolding realistic but aggressive ultrasonics. I spent years in the lab testing lossy codecs, and @amirm is right, high level ultrasonic tone bursts or square waves will tell you nothing about the ultrasonic noise that is actually folded into the audioband with real music. This may not be a huge issue to audibility and may just look like excessive dither. We don't know.

Tidal also has the best android eq through uapp, and the best quality lossy encoder so I will probably keep it for those.

But it royally sucks that Tidal is selectively but increasingly ripping us off by delivering 13 bit files as lossless.
We can do further analysis, but here are the difference from the same song, one with very low levels... and his FLAC version is a real 16bit FLAC
It's "Bayou" by Thomas Strønen

FLAC 16-44.1 Qobuz
View attachment 125305

FLAC 16-44.1 Tidal (real FLAC for this one, not a 16bit MQA)
View attachment 125306

FLAC 24-48 folded Tidal
View attachment 125307

FLAC 24-96 Qobuz
View attachment 125308

MQA 24-48 Tidal Unfolded via Roon to 24-96
View attachment 125309

I've yet to find any instance where the hifi version of an mqa track was not simply the mqa file without flagging.

If you're aware of any I'd be happy to check. 2lno or other purchase sites are a different situation (or at least I would sincerely hope they don't have the same issue.....)
Yes, just look at the one I tested above ;-)
It's a different situation for 2L, but there's a small problem acutally : their MQA tracks are on Qobuz as MQA while nothing says it on the Qobuz interface.
 

pjug

Major Contributor
Forum Donor
Joined
Feb 2, 2019
Messages
1,775
Likes
1,561
I just signed up for a Tidal trial. First "Master" I checked is from a 48K source (a track from new Valerie June album). I am guessing that at this point the majority of Tidal Masters have no real content above 24KHz. I'm not up to speed on the Tidal labeling; does "Master" always mean MQA?
 
Last edited:

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,423
Likes
7,940
Location
Brussels, Belgium
I've yet to find any instance where the hifi version of an mqa track was not simply the mqa file without flagging.

If you're aware of any I'd be happy to check. 2lno or other purchase sites are a different situation (or at least I would sincerely hope they don't have the same issue.....)

Are you serious? I mean I haven't used Tidal in 6 months so things could have changed.
 

mieswall

Member
Joined
Nov 10, 2019
Messages
63
Likes
106
With the mass replacement of redbook on Tidal, surely a lot of these MQA are made from 44.1K or 48K originals. So then the spectra have a gap between the music and the fake ultrasonics, like this (from https://audiophilestyle.com/ca/reviews/mqa-a-review-of-controversies-concerns-and-cautions-r701/)
View attachment 125479
When you find a fact as crude as the one showing this graph, if you are interested in learning something instead of just bashing MQA, what would be interesting is to discuss why do you think that gap is happening, that happens to be exactly at the Nyquist frequency of Redbook (other than the presumption of most people here that this MQA process is done by absolutely incompetent engineers, of course)...
My guess: MQA claims that it is trying to correct the aliasing problems found in sources when dealing with digital data (instead of analog sources), applying some filter techniques. Perhaps if they can't fix it because of the nature of the source presented, the algorithm assumes that is better just to erase a small 1/3 octave supra-aural band instead of leaving those aliasing artifacts in. Is it that bad? from a religious "losslessness" credo it is; if you are instead interested in the aural results, perhaps it is not.

Another surprising element in these tests, that after over a thousand posts nobody has dared to discuss: it is evident that is you are measuring "losslessness" of a source file against a MQA-process, that by definition replaces the original noise floor with real information dithered as to still appear as noise to a normal DAC, you will get gigantic differences: that noise floor which you are expecting to find as an exact copy doesn't exist anymore in MQA. And yet, not a single person has commented why is this. I think, just perhaps it is not because MQA engineers are stupid, but instead because they are doing this on purpose: they are trying to achieve a different result (gain space for real information in what's garbage in the source file).

In the upper bits of the audio spectrum, above the threshold of maximum expected amplitudes of music, it is happening the same thing: Instead of recording voids, MQA is recovering that space to store real data. As the practical bit depth used to register the content in this sampling is smaller, when you feed the algortithm with test tones, like square waves, totally discrepant with this assumptions, the reconstruction of that content (that square wave) cannot be done. The resulting wave in GoldenEar test shows exactly that. If any, the only thing this test it is proving is that MQA is doing what their papers say they do.

If any of the people commenting here dare to read just the most general papers fo the MQA process, (most available for free elsewhere), they would be aware of the above things. Then we would be discussing if those assumptions are correct to not, but all people here insist in analyzing this issue as if MQA were just a flawed new kind of pkzip algorithm...

As a crude analogy, I t's like these test are comparing a file copy of a bunch of data done from an old, highly used hard disk, to a fresh new one. Files copied are "lossless" (that's what MQA claims, I agree, rather carelessly); then Archimago and GoldenEar are not looking if those files are identical (the music) but comparing each sector of each disk against the other. The first one, with all file fragmented and rest of old files deleted (the noise floor) is obviously different form the contiguous files encountered in the new disk. This new disk also doesn't contain all the garbage of those hidden erased files from the old one. Thus, it is "lossy"!, I will close my Tidal account!! They are lying to me!!! I may even build a case for seeing them!!!!
 
OP
GoldenOne

GoldenOne

Not Active
Joined
Jun 25, 2019
Messages
201
Likes
1,469
Are you serious? I mean I haven't used Tidal in 6 months so things could have changed.
Yep :(

This wasn't always the case. When I did a deltawave comparison of the master and hifi version of the same track a few months back they were different. But now every one I've tried has been the same
 

Grooved

Addicted to Fun and Learning
Joined
Feb 26, 2021
Messages
679
Likes
441
I just signed up for a Tidal trial. First "Master" I checked is from a 48K source. I am guessing that at this point the majority of Tidal Masters have no real content above 24KHz.
There are a lot of 44.1/48 and 88.2/96, which can be fully unfold with any MQA DAC or even a non-MQA DAC if Core decoding is done by the software (Tidal app, Roon...).
There are also 192 and up to 352.8 tracks (not found any of 384), which need rendering but since it's mainly a upsampling, it makes less improving from 96 to this upsampling than the improvement you get from the first unfold
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,423
Likes
7,940
Location
Brussels, Belgium
Yep :(

This wasn't always the case. When I did a deltawave comparison of the master and hifi version of the same track a few months back they were different. But now every one I've tried has been the same

I seriously don't understand why would they ever do this. they still have to host two different files for the same track (as Hi-Fi is without the tagging).

there is not a single reason i can think of why this would be a good idea.
 

Tks

Major Contributor
Joined
Apr 1, 2019
Messages
3,221
Likes
5,494
@GoldenOne Has MQA been in contact other than the first communication?

I seriously don't understand why would they ever do this. they still have to host two different files for the same track (as Hi-Fi is without the tagging).

there is not a single reason i can think of why this would be a good idea.

Seems like they really want to push MQA stats for some reason (maybe something to show to investors or future investors on how a music DRM scheme is still quite viable). Also while hosting two files is currently the case, they could eventually just hold one and be done (as shown from the disappearing of lossless). This also means they never have to bother with the headache of supporting a truly other sort of file. Both can be MQA, and unfold differently (it's not like this information would be available to us unless people pull the move OP did, and/or Niel Young basically calling them out, saying they're just upsampling). Having one reference file to draw from, and perhaps segment the pricing tiers, while also maintaining viability for the DRM-like scheme as something the market will bare, seems to be the longer term play.

Like, if they can get so many people to still maintain a defense of MQA after this ordeal (with Amir also attempting to take as neutral of a position as he can muster in spite of the obvious conflict), you can only imagine how much signalling this is that they don't have much to fear in the long term it seems. Like the fact that any lossless is being removed is a sort of thing I cannot comprehend. Not so much from the perspective of not understanding a streaming service offering a lossy format for streaming (that's fine), but offering a lossy format that purports better than lossless (with all their "original" and "master authenticated" etc..). And on top of that, charging more for it than other lossy streaming platforms.

If Tidal would simply offer only MQA as they intend, and ceased with the false advertising claims. Have at it, I would have zero problems there. But they're wanting to have cake and eat it.
 
Last edited:

DDF

Addicted to Fun and Learning
Joined
Dec 31, 2018
Messages
617
Likes
1,355
We can do further analysis, but here are the difference from the same song, one with very low levels... and his FLAC version is a real 16bit FLAC
It's "Bayou" by Thomas Strønen

FLAC 16-44.1 Qobuz
View attachment 125305

FLAC 16-44.1 Tidal (real FLAC for this one, not a 16bit MQA)
View attachment 125306

FLAC 24-48 folded Tidal
View attachment 125307

FLAC 24-96 Qobuz
View attachment 125308

MQA 24-48 Tidal Unfolded via Roon to 24-96
View attachment 125309


Yes, just look at the one I tested above ;-)
It's a different situation for 2L, but there's a small problem acutally : their MQA tracks are on Qobuz as MQA while nothing says it on the Qobuz interface.

Thanks, but these dont adress the most important question: what is the noise level created by down folded ultrasonics? Nearing 50 pages, and best estimate available is just from knowing mqas technique of chucking the bottom 3 bits: -78 dBFS. This is bad enough but likely inaudible if volume normalization is turned on. Given a choice, would prefer to avoid it, of course.
 

Grooved

Addicted to Fun and Learning
Joined
Feb 26, 2021
Messages
679
Likes
441
Yep :(

This wasn't always the case. When I did a deltawave comparison of the master and hifi version of the same track a few months back they were different. But now every one I've tried has been the same
Even if the FLAC file has been replaced by a MQA file, it will be by a 16bit one. If this MQA file is the same than the Master one, then it means the Master one is alo a 16bit MQA, which happens. But it can also be a 24Bit one.

The problem is that Tidal use the Master logo for all MQA files, not for saying it comes from Hi-Res Master.
And sometimes, a same album has different version, even different in MQA !
Testing the Warner catalog, I used "Blood Sugar Sex Magik" because they deleted the version coming from the "original master" and now have only the "2014 remastered" version. And they deleted FLAC version, and it's not one version 16bit/44.1 MQA , and one 24bit/96 MQA
 
Last edited:

pjug

Major Contributor
Forum Donor
Joined
Feb 2, 2019
Messages
1,775
Likes
1,561
There are a lot of 44.1/48 and 88.2/96, which can be fully unfold with any MQA DAC or even a non-MQA DAC if Core decoding is done by the software (Tidal app, Roon...).
There are also 192 and up to 352.8 tracks (not found any of 384), which need rendering but since it's mainly a upsampling, it makes less improving from 96 to this upsampling than the improvement you get from the first unfold
I edited my post. It did unfold as an MQA. But had the telltale gap showing that the source used to make the MQA was 48K.
 
OP
GoldenOne

GoldenOne

Not Active
Joined
Jun 25, 2019
Messages
201
Likes
1,469
@GoldenOne Has MQA been in contact other than the first communication?
They have not. I have offered to do tweaked testing based on some of the criticisms in their original response but I've not heard anything.

I have had contact from a few manufacturers though with some.....interesting information regarding how mqa has forced manufacturers to alter their products before they'd be allowed to implement mqa.

I need to get a couple things verified but there may be a followup video at somepoint.
 

DDF

Addicted to Fun and Learning
Joined
Dec 31, 2018
Messages
617
Likes
1,355
I seriously don't understand why would they ever do this. they still have to host two different files for the same track (as Hi-Fi is without the tagging).

there is not a single reason i can think of why this would be a good idea.

It lowers their operational costs.

This cost reduction changes the product's very nature hoisted on (previously) unsuspecting consumers. In a less sleazy world than audio, this would drive a new product name, or part number etc.
 

Grooved

Addicted to Fun and Learning
Joined
Feb 26, 2021
Messages
679
Likes
441
Thanks, but these dont adress the most important question: what is the noise level created by down folded ultrasonics? Nearing 50 pages, and best estimate available is just from knowing mqas technique of chucking the bottom 3 bits: -78 dBFS. This is bad enough but likely inaudible if volume normalization is turned on. Given a choice, would prefer to avoid it, of course.
Right, it's just the first track I tested, and just down a fast waveform statitics. It's also not the bets track to do a test I think.

The "nearly 50 pages" is more due to the fact that this thread is now regrouping too many subject (analysis of MQA processing, what claims from MQA are right or wrong, what claims from Tidal are right or wrong,...). Should have keep it seperated.
 

jensgk

Active Member
Forum Donor
Joined
Mar 21, 2020
Messages
256
Likes
564
Location
Denmark
It's a different situation for 2L, but there's a small problem acutally : their MQA tracks are on Qobuz as MQA while nothing says it on the Qobuz interface.

This, I think is another problem with MQA. Before MQA we could (mostly) assume that a flac file contained lossless audio, but now, because they chose to use flac a container, we cannot anymore :-(
 

abdo123

Master Contributor
Forum Donor
Joined
Nov 15, 2020
Messages
7,423
Likes
7,940
Location
Brussels, Belgium
It lowers their operational costs.

This cost reduction changes the product's very nature hoisted on (previously) unsuspecting consumers. In a less sleazy world than audio, this would drive a new product name, or part number etc.

what operational costs? all of this can be automated anyway.
 

AdamG

Proving your point makes it “Science”.
Moderator
Forum Donor
Joined
Jan 3, 2021
Messages
4,636
Likes
14,918
Location
Reality
Right, it's just the first track I tested, and just down a fast waveform statitics. It's also not the bets track to do a test I think.

The "nearly 50 pages" is more due to the fact that this thread is now regrouping too many subject (analysis of MQA processing, what claims from MQA are right or wrong, what claims from Tidal are right or wrong,...). Should have keep it seperated.
You are welcome to open a new thread with a specific drill down topic. Keeping it on track and subject focused is the problem. I am willing to give it a go if the Member who makes the new thread helps keep the conversation on point. ;)
 

RichB

Major Contributor
Forum Donor
Joined
May 24, 2019
Messages
1,946
Likes
2,611
Location
Massachusetts
I don’t think that FLAC can encapsulate a high resolution audio stream that’s backward compatible with a standard one (CD). That’s what I read into his comment.

PS: unless it’s MQA, that is.

No, it is not MQA when you examine the real-world requirements.

Streaming vendors must provide more than one resolution, not only for the price-point plans but to accommodate mobile and connections via lower bandwidth networks.

Here is the QOBUZ app settings for streaming via WI-FI and Mobile network:

QOBUZStreamingOptions.jpg


So there can be up to 4 different versions of a title (depending on plan).
Apparently they have programmers who can develop mobile, browser, desktop apps and protection that are proprietary.

Providers have a choice to create pre-processed versions for each streaming option that uses space.
They could also store difference (delta) files to save space.
I suspect that disk space is cheaper than real-time processing so I suspect they make all the versions they need for each title.

Since MQA does not provide for up to 4 streaming options from the same source, it has not provided a single source solution for streaming services. For all we know, there is 3'rd party developers providing the required optimizations to save space as well.

- Rich
 
Last edited:
Status
Not open for further replies.
Top Bottom