Jump to content
IGNORED

MQA is Vaporware


Recommended Posts

46 minutes ago, mansr said:

There is no such "unfold." There is only upsampling, and a poor one at that.

Marketing.

So, if we look at 2L recordings available in MQA, here is a sampler:

 

http://www.2l.no/hires/

 

Most of these were originally recorded in DXD = 352.8k sampling resolution.  Are you saying that in making the MQA version, 2L downsampled the original 352k to 88k and to 44k, then UPSAMPLED that 88k version to 176 and 352k?  That, as opposed to enfolding 352, 176, and 88k all DOWNSAMPLED straight from the 352k original?  

 

And, of course, you have concrete proof of this?

Link to comment
2 hours ago, mansr said:

They downsampled to 88k sample rate. The resulting 0-44k frequency content was then split into two bands. 

 

Just as an outsider looking in at things simplistically, the idea of splitting into bands sounds clever for reducing file size of Hi-Res streaming. As in, the whole folding of the region B (and region C, for completeness although there's not much happening in there) into region A to get a smaller file size looked clever to me. At least from the marketing origami video lol.

 

I just wish there was there no lossy compression involved anywhere, at all. And it was more open for all to use.

 

But open usually means free and I guess they spent a lot of time and resources that they need to recoup a lot of those costs and want to profit of it longer term.

 

Unless there's no way to do this folding and unfolding without lossy compression somewhere in the process?

 

Link to comment
27 minutes ago, Em2016 said:

 

Just as an outsider looking in at things simplistically, the idea of splitting into bands sounds clever for reducing file size of Hi-Res streaming. As in, the whole folding of the region B (and region C, for completeness although there's not much happening in there) into region A to get a smaller file size looked clever to me. At least from the marketing origami video lol.

 

I just wish there was there no lossy compression involved anywhere, at all. And it was more open for all to use.

 

But open usually means free and I guess they spent a lot of time and resources that they need to recoup a lot of those costs and want to profit of it longer term.

 

Unless there's no way to do this folding and unfolding without lossy compression somewhere in the process?

 

 

There is a lossless tier defined in the spec, but I suspect it would not save much (if any) over existing compression methods.  

 

The big savings in file size for MQA is the fact they do not encode all the "empty space" (i.e., most of the content) above the uncompressed 0-22k band, as mansr mentions above.

John Walker - IT Executive

Headphone - SonicTransporter i9 running Roon Server > Netgear Orbi > Blue Jeans Cable Ethernet > mRendu Roon endpoint > Topping D90 > Topping A90d > Dan Clark Expanse / HiFiMan H6SE v2 / HiFiman Arya Stealth

Home Theater / Music -SonicTransporter i9 running Roon Server > Netgear Orbi > Blue Jeans Cable HDMI > Denon X3700h > Anthem Amp for front channels > Revel F208-based 5.2.4 Atmos speaker system

Link to comment
9 minutes ago, jhwalker said:

There is a lossless tier defined in the spec, but I suspect it would not save much (if any) over existing compression methods.

 

Thanks.

 

I like reading Rt66indierock's updates and I saw one where MQA have said the real push needs to come from the labels themselves going forward, so we'll see over the next 12-18 months how keen the labels are on pushing this (marketing).

 

Mind you, there's no cost increase with a Tidal subscription, it's part of their 'HiFi' service for no extra cost - I guess that may change potentially when the library gets to a reasonable size. The labels may require streaming services to charge a premium for MQA streaming, when they feel it's reasonable to do so with a large library? My understanding is at present it's the labels that dictate the cost of streaming services. I remember reading that Apple wanted to charge something much cheaper but the labels rejected it.

 

Streaming numbers are continuously growing (if you believe Spotify and Apple Music subscriber numbers) and both of those don't even have a lossless service yet released. I know why MQA themselves are keen to push this as quickly as possible but if streaming numbers are naturally growing nicely, I don't know what the rush would be from labels, unless the labels get convinced that this will lead to an exponential up-kick in subscriber numbers, over the already increasing numbers? Maybe they have access to data that we don't.

 

Nobody likes to see people stealing from the artists and I subscribe to streaming myself for this reason (and still buy albums not available on streaming) but I don't know how this will prevent copying because computers can output the MQA 24/96 streams digitally.

 

It's been interesting to watch the developments and follow the conversations from fellow passionate music lovers.

 

 

Link to comment
8 minutes ago, Em2016 said:

The labels may require streaming services to charge a premium for MQA streaming

If the ultimate motivation of the labels is combating piracy (as some have speculated) then maybe it is in their interest not to charge a premium for MQA. 

Pareto Audio AMD 7700 Server --> Berkeley Alpha USB --> Jeff Rowland Aeris --> Jeff Rowland 625 S2 --> Focal Utopia 3 Diablos with 2 x Focal Electra SW 1000 BE subs

 

i7-6700K/Windows 10  --> EVGA Nu Audio Card --> Focal CMS50's 

Link to comment
23 minutes ago, rickca said:

If the ultimate motivation of the labels is combating piracy (as some have speculated)

 

Understood but to my very simple mind, this would only work well if they limited the unfolding to only the endpoint, the MQA DAC.

 

But computers can output 24/96 digitally so how can they combat piracy allowing that? Even the Bluesound Node 2 can output 24/96 via SPDIF outputs.

 

If it's to combat piracy, they shouldn't allow any PC decoding and completely limit MQA decoding to only DACs. But I do understand that the PC's power is needed with some DACs to do the first unfold as some DACs aren't powerful enough to do the full decode. If you're going to combat piracy then do it properly. Make it part of the spec for all MQA DACs to have enough processing power to do the full decode and limit any digital outputs to 24/48. I guess that still won't stop piracy though. Unless they encrypt the 24/48 signal from the computer so that only an MQA DAC (that can do the full decode) can read the signal. But then somebody will eventually crack that too lol. Whack a mole like it's been forever

 

Streaming subscriber numbers are continuously growing (so we're told) so that's potentially fighting piracy on it's own, even without lossless streaming, let alone a Hi-Res streaming service (in the cases of Spotify and Apple Music anyway). That's why I don't see why the labels would be so rushed with MQA, even the MQA team obviously want this adopted as quickly as possible.

 

Link to comment
35 minutes ago, Em2016 said:

If it's to combat piracy, they shouldn't allow any software decoding and limit decoding to only DACs.

OK perhaps I should say the initial goal is to control distribution, IOW to authorize distribution of hi-res only in MQA format.  Some have speculated that ultimately MQA may introduce some form of DRM.

 

We really don't know the long-term strategy.  All we can see is the current tactical plan.  It will evolve.

 

Some people think that this is all paranoid nonsense.  I ask those people, what is the business case behind the adoption of MQA?

Pareto Audio AMD 7700 Server --> Berkeley Alpha USB --> Jeff Rowland Aeris --> Jeff Rowland 625 S2 --> Focal Utopia 3 Diablos with 2 x Focal Electra SW 1000 BE subs

 

i7-6700K/Windows 10  --> EVGA Nu Audio Card --> Focal CMS50's 

Link to comment
3 minutes ago, rickca said:

Some have speculated that ultimately MQA may introduce some form of DRM.

 

This is a very fair point. What we see right now with MQA streaming may not be what we see 12-18 months from now.

 

Food for thought.

Link to comment
3 hours ago, Em2016 said:

 the whole folding of the region B (and region C, for completeness although there's not much happening in there) into region A to get a smaller file size looked clever to me.

 

Ironically, that was not even invented by MQA, but by the Japanese, quite a few years ago.

 

Link to comment

Maybe they were inspired by the folding of layers of steel in samurai swords.  Origami was the gentler explanation.

And always keep in mind: Cognitive biases, like seeing optical illusions are a sign of a normally functioning brain. We all have them, it’s nothing to be ashamed about, but it is something that affects our objective evaluation of reality. 

Link to comment

So, still couldn't escape from being nearly exclusive format for Meridian DACs. Still very limited catalog and availability....

 

I suspect "It's DRM!" here pretty much killed MQA anyway.

 

I still remember Bob's Q&A in this very forum, and he pretty much side-stepped/dodged all of the important questions, other than angrily said "IT'S NOT DRM AND ANYONE WHO SAYS SUCH HAS MALICIOUS INTENT TO MY COMPANY" in red-colored words while other explanations and stuffs told us different picture.

Link to comment
7 hours ago, Em2016 said:

Understood but to my very simple mind, this would only work well if they limited the unfolding to only the endpoint, the MQA DAC.

 

But computers can output 24/96 digitally so how can they combat piracy allowing that? Even the Bluesound Node 2 can output 24/96 via SPDIF outputs.

 

If it's to combat piracy, they shouldn't allow any PC decoding and completely limit MQA decoding to only DACs. But I do understand that the PC's power is needed with some DACs to do the first unfold as some DACs aren't powerful enough to do the full decode. If you're going to combat piracy then do it properly. Make it part of the spec for all MQA DACs to have enough processing power to do the full decode and limit any digital outputs to 24/48. I guess that still won't stop piracy though. Unless they encrypt the 24/48 signal from the computer so that only an MQA DAC (that can do the full decode) can read the signal. But then somebody will eventually crack that too lol. Whack a mole like it's been forever

The output from a "core" decoder sent to a renderer embeds a simple control bitstream in the LSB. This includes a flag informing the renderer that the audio data is scrambled. Curiously, the Dragonfly doesn't seem to implement this feature and ignores that flag. The rendering code in the Bluesound handles it when called directly, but I don't know if the device can work as a separate renderer in normal operation. If it can, and someone wants to try, I have a test file with scrambling enabled.

Link to comment
35 minutes ago, mansr said:

The output from a "core" decoder sent to a renderer embeds a simple control bitstream in the LSB. This includes a flag informing the renderer that the audio data is scrambled. Curiously, the Dragonfly doesn't seem to implement this feature and ignores that flag. The rendering code in the Bluesound handles it when called directly, but I don't know if the device can work as a separate renderer in normal operation. If it can, and someone wants to try, I have a test file with scrambling enabled.

 

Thanks. If I still had my Bluesound Node 2 I'd definitely offer to help.

 

So do you mean that (for example) currently with Tidal MQA streaming, the 24/96 MQA digital audio output from a PC  has audio data scramble disabled, but the audio data scrambling could potentially be enabled at any time?

 

Link to comment
15 minutes ago, Em2016 said:

If I still had my Bluesound Node 2 I'd definitely offer to help.

Did you ever try playing MQA through it using software decode?

 

6 minutes ago, Em2016 said:

So do you mean that (for example) currently with Tidal MQA streaming, the 24/96 MQA digital audio output from a PC  has audio data scramble disabled, but the audio data scrambling could potentially be enabled at any time?

That's correct.

 

One possibility, and I'm speculating here, is that they'd start enabling the scrambling whenever a compatible renderer is detected. Once this feature is in place, they could also start tagging certain tracks with an instruction to refuse decoding unless a descramble-capable renderer is used. Broadcast TV works much this way with respect to HDCP, so the concept is not new.

Link to comment
5 minutes ago, mansr said:

Did you ever try playing MQA through it using software decode?

Nope sadly I sold my Bluesound last year well before MQA support came to it. You could try asking someone here: 

 

 

8 minutes ago, mansr said:

Once this feature is in place, they could also start tagging certain tracks with an instruction to refuse decoding unless a descramble-capable renderer is used.

Yep probably will happen exactly as you speculate, especially if fighting piracy is one of label's main drivers for the adoption of MQA.

Link to comment
14 hours ago, mansr said:

They downsampled to 88k sample rate. The resulting 0-44k frequency content was then split into two bands. The lower band, 0-22k was dithered to 15 bits while the upper band, 22-44k, was lossily compressed to 8 bits per sample and stuffed into the low byte of the 24/44 MQA stream. On playback, the compressed data from the low byte is decoded to partially restore the 22-44k band and merged with the uncompressed 0-22k band to produce an 88k sample rate signal. This is then upsampled to 176k or 352k, depending on the DAC, and dithered to 16 bits.

Yes.

I follow your explanation which agrees with my own understanding of MQA in terms of the lossy compression of hi rez information above 44k or 48k sampling rates into the low byte of the stream.  However, the rest of your assertion simply does not agree with the explanation provided by Bob Stuart and Morten Lindberg:

 

http://www.2l.no/hires/documentation/2L-MQA_Comparisons.pdf

 

They state that the MQA version folds recordings done at 352k down 3 times to 44k. That is considerably different from your version.

 

So, quite simply, and with all due respect, I just do not see any reason to believe you.  You state you have concrete evidence, but where is it?

 

 

 

 

 

 

Link to comment
11 hours ago, rickca said:

Some have speculated that ultimately MQA may introduce some form of DRM.

 

Once again: we need to remember that 'DRM' is not synonymous with 'copy protection.' The term 'digital rights management' exists exactly because it is more inclusive. 'Authentication' is by definition a form of 'DRM,' and it's specified right in the MQA acronym. So there's no need to speculate: MQA definitely includes "some form of DRM."

 

3 hours ago, mansr said:

One possibility, and I'm speculating here, is that they'd start enabling the scrambling whenever a compatible renderer is detected. Once this feature is in place, they could also start tagging certain tracks with an instruction to refuse decoding unless a descramble-capable renderer is used. Broadcast TV works much this way with respect to HDCP, so the concept is not new.

 

This is very interesting. It underlines why any DRM is too much DRM - because the slope is so slippery.

 

DRM typically represents the use of technology to create entirely new extra-legal 'rights' for the publisher. For example, MQA, at minimum, gives publishers the new 'right' to 'authenticate' the audio stream and the DAC, without further permission or approval from the consumer. There's nothing in copyright law to support this.

 

And, by the way, there's also nothing about 'audio origami' that necessitates it.

Link to comment
4 minutes ago, Fitzcaraldo215 said:

They state that the MQA version folds recordings done at 352k down 3 times to 44k. That is considerably different from your version.

They are not being entirely truthful.

4 minutes ago, Fitzcaraldo215 said:

So, quite simply, and with all due respect, I just do not see any reason to believe you.  You state you have concrete evidence, but where is it?

The best evidence is the actual software doing the decoding and rendering. I obviously can't post it here for copyright reasons, and even if I did, you could choose to believe I'd faked it.

 

Other evidence is in the analogue captures of DACs playing MQA files compared to the original high-resolution masters. The MQA playback contains high-frequency components not present in the original. On closer inspection, these components are clearly images of lower frequencies created by the leaky upsampling filters used in the "rendering" process. Lots of people have made such recordings and analysed them, so there's no need to take my word for it.

Link to comment
1 minute ago, Rt66indierock said:

The Ten Years War

The first eight studio albums by Black Sabbath in vinyl with a crucifix containing all eight albums in MQA will be available September 29th. All have been on my MQA list.  I think its great MQA supporters can show their support by wearing the crucifix.

 

I just saw this.  I had a good chuckle in that if this is what MQA is about then it is truly the walking dead format...

Hey MQA, if it is not all $voodoo$, show us the math!

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...