How is MQA authenticated?

Audio Gear Talk MQA
1

What we must not forget is what MQA stands for:
Master Quality Authentication
How do we know the providence of the MQA we get.
What does it mean?
How can we be sure it is what the artist/recording- and mixing-engineer intended?
What was the material this so called MQA is based on?
MQA is presented as a solution for reducing bandwidth demands, but the name says otherwise.
What is it the MQA inventors, investors and others want with this format?

To much is fuzzy, undefined and uncertain about MQA.
There is no proof for any of the claims MQA pushers make.
What they have is proprietary technology and any criticism is bullied away.

Come clean or get lostā€¦ be honest about your sh*t.

2 Likes
2

Did you send these questions to MQA? What did they answer you?

1 Like
3

The questions you are asking are under the control of the rights holders, who choose the ā€˜masterā€™ version. It means either the major labels in most cases, or the indie artists or whoever they contract to, to make and distribute their work. The major labels often have 50 or more versions of older titles. So you can ask the question, how would they explain to you the process of choosing the best amongst them. How would you know or be able to compare their various versions to choose a best ā€˜masterā€™. All they can do is maybe, at some point, release a statement or description of their choice process. And even then, you could dispute and decide that itā€™s all too fuzzy if you choose to. I think you have to live with the labelā€™s decisions without being so cynical about their intentions.

3 Likes
4

Not another MQA thread pleaseā€¦ :sweat_smile:

6 Likes
5

Neil Young might disagree with you.

2 Likes
6

The people who dislike MQA should ask themselves what they can and cannot achieve by posting so much about itā€¦ Does MQA really deserve so much attention?

12 Likes
7

The interwebs are a mine of information, this is a Roon forum, not an MQA forum.
Therefore, LMGTFY - Let Me Google That For You

3 Likes
8

Lot of ā€œfoolā€™s goldā€ in that there mine, too.

2 Likes
9

I was writing a serious reply and then when double checking your questions I saw the last line of your post.

So very quick summary on a snipped post:

One of the things MQA was designed with in mind was archiving. A term you may not have come across before is ā€˜The Digital Dark Ageā€™. Easy summary: https://www.youtube.com/watch?v=PSlMzirvsFc
Bit rot and file obsolescence is a big issue. One of the other concepts that falls out of the digital dark age is the digital integrity of a file. If you archive a file for 50 years, swap it to a new hard drive every 7 years - how can you guarantee that the file is the same as it was 50 years ago? How many bits have changed? For most people - nothingā€¦for others this matters. Also, if your archiving millions of files, who is to say a file has been tampered with in the 50/100/150 year storage? A book in the library is easy to spot when its been tampered with. A digital file? Not so if someone knows how to cover their tracks.

Here is a link ā€œDebunkingā€ MQAs authentication Truncating MQA files to 16 bits and the blue light still shines - General Forum - Audiophile Style . However, as with most things MQA, the OP didnā€™t quite fully understand how it works and a later user explains it:

Blockquote This is by design. The MQA identification and authentication data is embedded in bit 8 of a 24-bit PCM stream. Dropping bits 0-7 thus leaves it untouched. The authentication works by computing a Blake2s hash over the top 15 bits (the plain PCM portion) and parts of the control stream in bit 8. This hash is then verified against a cryptographic signature extracted from the control stream. The public key is stored in the decoder. If the signature matches, the blue light goes on. The low 8 bits encoding the high-frequency content are not covered by the authentication.

Source: Audiophile Style. 2021. Truncating MQA files to 16 bits and the blue light still shines. [online] Available at: https://audiophilestyle.com/forums/topic/38608-truncating-mqa-files-to-16-bits-and-the-blue-light-still-shines/?do=findComment&comment=784981 [Accessed 4 June 2021].

N.B. OP of that post later accused me of working for MQA (which I do not).

The authentication only works on the Lossless bit of MQA and not the Lossy bit of MQA. I.e. if you tamper with the ā€œlosslessā€ bit of the file stored in the 16 bit area of the file it will not authenticate. If you tamper with the next 7 bits where the lossy data (audio info above 24kHz is) it will still authenticate.


Source: Stuart, B., 2021. MQA 16-bit and Provenance in the Last Mile. [online] Bobtalks.co.uk. Available at: https://bobtalks.co.uk/a-deeper-look/deeper-look-mqa-16b-in-the-last-mile/ [Accessed 4 June 2021].

So:

  • Future format intentions were there/Archiving intention was there EDIT: Speculation
  • Someone realised they could market this EDIT: Speculation
  • But the aim of ensuring that the user knew if the file had been tampered with at all looks like its false.

With most ā€˜unhackableā€™ things, it does get hacked.

EDIT: I also just saw how old this post is after writing this response. Apologies for resurrecting it. Iā€™ve just been doing a lot of digging into how MQA works after GoldenSounds video.
Mods - feel free to delete the comment if needs be. Apologies.

4 Likes
10

The diagram you post is lifted directly from Bob Stuartā€™s current website in the section called ā€œA Deeper Lookā€ and lacks attribution. The comments you have in block form are a loose approximation to his description of the cryptographic signature.

What gives :slight_smile:

Future format intentions were there/Archiving intention was there
Someone realised they could market this
But the aim of ensuring that the user knew if the file had been tampered with at all looks like its false.

What do these conclusions mean? Considering that the block diagram is a flow chart of the processing also described in MQA patents and designed by intent as part of the codec, I donā€™t understand what it means to say ā€œsomeone realised they could market thisā€? If you read the Stuart/Craven MQA paper in JAES, they say what they think would make a good archive. I donā€™t think they actually thought of the MQA code as an archival format.

11

Iā€™m not sure what your objections are? I donā€™t mean this in a rude way, I just may be miss-reading your objections.

The diagram is linked via URL, so I thought its source should be clear (right click, copy link) - but will edit the post to makes its source clear. I didnā€™t copy the diagram and upload it, I wasnā€™t trying to hide where it came from.

The comments in the block I thought I made clear were from another user in the forum that I linked at Audiophile style. I.e. one person thought he had broken the MQA authentication and another user told him he hadnā€™t. I.e. I wrote: ā€œAnd a later user explains:ā€ before going into the block quote.

If you wish to have proper citations in Chicago Style (16) I can do that, but I thought the sources were clear for an internet forum!! :thinking:

I put the block diagram in to try further illustrate why you can ā€œBit truncateā€ an MQA file and it keeps its authentication. The diagram shows the top 16 bits is the part of the music our ears hear i.e. 24kHz and below. It was an expansion/diagram of what the block text section was saying. Whilst the diagram is mainly concerned with how MQA files ā€œfoldā€ music, I thought it was a useful visual representation. Maybe I should have just sniped the Top 16 bits/Low 8 Bits of the diagram to save confusion.

If you are referring to this paper: AES E-Library Ā» A Hierarchical Approach to Archiving and Distribution they are not using the term archiving in the same manor as I am. Are you referring to a different paper? I donā€™t recall any others talking about archiving in the way I was describing.

However, I will be honest: I am struggling to find my source for MQA being part developed for being an archival format because of its authentication. Over the past few weeks Iā€™ve watched numerous videos, read websites and papers on MQA. I have kept a log of interesting bits to Cite. The long term storage part I remember coming up as I am researching ā€œThe Digital Dark ageā€ for work right now so it stuck in my mind. But I think its still a point that stands, the authentication process helps against bit-rot with upcoming digital dark age (or, thinking more about it - might make it worse as the public key to decrypt it might be lost!).

Iā€™ve already touched on why I put the diagram in. The diagram is nothing to do with marketing.

On the marketing I may have then done a 2 + 2 = 5 approach. My brain went: ā€œBecause they had this cool solution to bit rot - someone thought they could use this tech as a marketing toolā€ and hence we have the name Master Quality Authenticated and different lights on our decoders etc. I am speculating about this. But I do wonder if the authentication lights and ideas came after the original idea of having the file authenticated.

Hope that helps? Its very late were I am, so wonā€™t respond till tomorrow if you have further questions.
Sorry I caused confusion.