Jump to content
IGNORED

MQA is Vaporware


Recommended Posts

4 hours ago, R1200CL said:

Chris and others. Thanks for taking time to enlighten me 😀

 

 Feel free to read the 832 pages in this thread, any other thread, and even Archimago's blog.  MQA promoters have brought up the same BS hundreds of times.  If after this you have questions that have not been examined and refuted a hundred times, feel free to call Bob Stuart.

Boycott Warner

Boycott Tidal

Boycott Roon

Boycott Lenbrook

Link to comment

Hello, I've been reading this and other forums for some time now and I have done some research myself about MQA that I'd like to share. I tried not to be affected by any positive or negative thoughts posted here or elsewhere.
But after some months of experiments I  do have some serious questions about MQA.

I compared some of the Warner 16b/44.1kHz MQAs with the original non-MQA PCM flacs - that they removed at the very same time those MQAs came online. Luckily - I had some of those PCMs stored locally so I could compare (most people didn't have the chance ... )
In my opinion in most cases the 16b 44k PCM sounded better than the fully 'unfolded' 16b 44k MQA 
This could be the reason the original HiFi files were immediately removed at the same time the MQA's were released : So we couldn't compare and notice. Well I could and I noticed.
(Only in some cases I could say the MQA sounded better; have to mention this as well)

So now I have a question :

Who would like the original Hi-Fi 16b/44.1kHz PCM flacs back on Tidal ? 

(Undo the Warner conversions? I DO !) 

(I am not talking about the higher quality 24b MQA's that were already there before Warner flooded Tidal with 16bs... those sound pretty good in my opinion)

Link to comment

I’m still chasing this 16 bit MQA decoding or upsampling or whatever is happening. (In Roon decoding signal path). 
 

In this article I found this sentence:

“Well, this is an example of what happens when 44/48kHz audio is "unfolded" into 88/96kHz by the MQA Core decoder in TIDAL using their "leaky" filter of choice allowing the ultrasonic distortions to seep through. It's an example of "fake hi-res" MQA unfolding.”


@Archimago or others. 
Can I have an explanation/confirmation of what actually happened here ? 
 

I hope the explanation is a simple as an error in bad MQA SW, and of cause there is no upsampling/unfolding and bit conversation. 
 

Edit

Now I’m confused. 
@mansr You wrote this back in 2017 in comments:

”That's not necessarily a bug. The MQA core decoder has a setting controlling whether to upsample 1x (44/48k) content to 2x (88/96k). Enabling this means the caller always receives 2x rate back from the decoder, which might simplify the setup a bit. The renderer will still upsample the output to 4x or 8x with the same leaky filters, so it doesn't really matter much in the end”.

 

I was so sure MQA decoder didn’t do any upsampling. Only unfolding if the original MQA file was 88.2 / 96 or above. 

 

Link to comment
1 hour ago, R1200CL said:

I’m still chasing this 16 bit MQA decoding or upsampling or whatever is happening. (In Roon decoding signal path). 
 

In this article I found this sentence:

“Well, this is an example of what happens when 44/48kHz audio is "unfolded" into 88/96kHz by the MQA Core decoder in TIDAL using their "leaky" filter of choice allowing the ultrasonic distortions to seep through. It's an example of "fake hi-res" MQA unfolding.”


@Archimago or others. 
Can I have an explanation/confirmation of what actually happened here ? 
 

I hope the explanation is a simple as an error in bad MQA SW, and of cause there is no upsampling/unfolding and bit conversation. 
 

Edit

Now I’m confused. 
@mansr You wrote this back in 2017 in comments:

”That's not necessarily a bug. The MQA core decoder has a setting controlling whether to upsample 1x (44/48k) content to 2x (88/96k). Enabling this means the caller always receives 2x rate back from the decoder, which might simplify the setup a bit. The renderer will still upsample the output to 4x or 8x with the same leaky filters, so it doesn't really matter much in the end”.

 

I was so sure MQA decoder didn’t do any upsampling. Only unfolding if the original MQA file was 88.2 / 96 or above. 

 

Archimago is just trying to show what the leaky filters used by MQA do: they allow/create high frequency aliasing artifacts that didn't exist in the original.

@mansr doesn't participate here anymore. You can ask him over at ASR. 

Main listening (small home office):

Main setup: Surge protector +>Isol-8 Mini sub Axis Power Strip/Isolation>QuietPC Low Noise Server>Roon (Audiolense DRC)>Stack Audio Link II>Kii Control>Kii Three (on their own electric circuit) >GIK Room Treatments.

Secondary Path: Server with Audiolense RC>RPi4 or analog>Cayin iDAC6 MKII (tube mode) (XLR)>Kii Three .

Bedroom: SBTouch to Cambridge Soundworks Desktop Setup.
Living Room/Kitchen: Ropieee (RPi3b+ with touchscreen) + Schiit Modi3E to a pair of Morel Hogtalare. 

All absolute statements about audio are false :)

Link to comment
27 minutes ago, John Dyson said:

If you look at supposedly super-high-quality SACDs or whatever, much of the time you can tell by a spectrogram that the actual music material is still cut-off at 21+kHz and most of the stuff above 23kHz is noise, glitches, super HF tones, or whatever.   However, the important fact is that a lot of material is band limited anyway -- high res sources mostly have the benefit of more noise, not necessarily significantly higher quality.

+1 

 

How a Digital Audio file sounds, or a Digital Video file looks, is governed to a large extent by the Power Supply area. All that Identical Checksums gives is the possibility of REGENERATING the file to close to that of the original file.

PROFILE UPDATED 13-11-2020

Link to comment
1 hour ago, R1200CL said:

 


I was so sure MQA decoder didn’t do any upsampling. Only unfolding if the original MQA file was 88.2 / 96 or above. 

 

It’s all lossy after 88.2 / 96, no part of the original file is recovered after this and it is upsampled to suit the max your DAC can handle eg a 352khz file will stop at 176khz if your DAC maxes out to that, I have never seen any proof of  “the third unfold” yet 

Link to comment
1 hour ago, John Dyson said:

I agree with your sentiments, but being totally balanced on the matter:   Even though the MQA is a distortion process with mild data compression -- it really *IS* sometimes difficult to do a fair comparison, because mastering differences EXTREMELY overshadow any distortion caused by MQA.

 

If you look at supposedly super-high-quality SACDs or whatever, much of the time you can tell by a spectrogram that the actual music material is still cut-off at 21+kHz and most of the stuff above 23kHz is noise, glitches, super HF tones, or whatever.   However, the important fact is that a lot of material is band limited anyway -- high res sources mostly have the benefit of more noise, not necessarily significantly higher quality.

 

----------------

 

Here is where I am going with this -- in reality, with MQA:

 

20-23kHz BW limited source ->  MQA distortion -> distribution -> MQA unfolding (with lingering artifacts) -> listener.

 

The 'natural' digital world is this:

 

20-22kHz BW limited source -> distribution -> listener

 

----------------


Even if MQA starts with >25kHz material without lots of additional noise, even high res sources are often not as good as the sample rate might suggest.   Why bother with MQA from a technical standpoint?   I can see ZERO real technical advantage over FLAC (or other lossless schemes) because any BW savings for MQA are silly/nothing to write home about  -- bandwidth is now VERY cheap.    Even in the post apocalyptic future, FLAC will be better -- at least the technology is open and full quality is reproduceable by publically known  methods.   If someone really does need serious data-bandwidth reduction, then why not use near-maximum quality opus or mp3?   Sure, mpeg-like schemes aren't perfect, but is good enough for casual listening where bandwidth might be more costly or more difficult to achieve. (e.g. portable.)

----------------

This is NOT about MQA hate, but there is simply no true engineering/technical benefit to MQA.   The technology appears to be really cute, but there is no real benefit.

 

Where do you or I benefit from MQA?   Not really anywhere (unless catalogs are restricted, and we cannot get non-MQA anymore.)

John

 


Well...
The PCM tracks (flacs) that are stored locally on my computer AND the MQA tracks (flacs) that are now on Tidal
- DO have the EXACT same number of samples (so tracklength in milliseconds is the same).
- Also the volume level etc is the same. 
- AND the ID from tidal is also the same (where the non MQA's were and where the MQA's are now)

So I am 99% certain I have been comparing files from the same master.

I have been comparing lots of files from different artists from the Warner label and

the original 16b 44k PCMs sound MUCH better than the MQAs that replaced them.

If you listen to the MQA versions only then it's harder to notice sound quality has dropped,
but when you CAN compare with the original PCM flacs, then you'd notice the difference is not subtle at all.
The PCM wins in most cases of the fully unfolded MQA. Sometimes the difference is hard to notice, but the MQA is almost never better than the original PCM.

Of course they knew that people would complain, when comparing. So what did they do ?
=> They released the MQA's and in the mean time immediately removed the PCM's they were created from.
=> That is how they made it impossible for most of us to compare and complain and how they got away with it.

So basically what do we have now with those 16b 44k MQAs ?

a) a sound format that without decoder is much worse than CD
b) a sound format that with decoder is still worse than CD

I don't see the "better than CD" in both cases.


I have also NEVER seen the Studio Dot light up for a 16b / 44k MQA *
Does that mean that no artist/studio/producer likes it ? (I fully agree, the original PCMs deserve the dot)


* You could force a studio dot light up on a 16b 44k MQA when you start from

a 24b / 44k studio MQA and strip the lower byte off; this is discussed in this thread already.
That is actually how they make an MQA-CD from 24bit MQAs :
they simply ditch the byte that contains the +24kHz frequency data as if it was never important after all.

That the quality is then even more reduced doesn't seem to matter. 
All that matters is the studio dot remains on the whole time !

 

Link to comment
2 hours ago, R1200CL said:

I’m still chasing this 16 bit MQA decoding or upsampling or whatever is happening. (In Roon decoding signal path). 

 

3 hours ago, R1200CL said:

Edit

Now I’m confused. 
@mansr You wrote this back in 2017 in comments:

”That's not necessarily a bug. The MQA core decoder has a setting controlling whether to upsample 1x (44/48k) content to 2x (88/96k). Enabling this means the caller always receives 2x rate back from the decoder, which might simplify the setup a bit.

 

I believe xxHighEnd's MQA Core decoder doesn't upsample both 16 bit & 24 bit MQA files if the MQA original sample rate is less 88.2kHz - perhaps @PeterSt can confirm/comment?

We are far more united and have far more in common with each other than things that divide us.

-- Jo Cox

Link to comment
12 minutes ago, R1200CL said:

From a business point of view, I guess it would be in Tidal’s interest to keep both. 

The only people concerned about Tidal’s interest are people at Tidal. MQA will use Tidal as just another rung on the ladder to get to the services that actually bring in money with huge user numbers.  

Founder of Audiophile Style | My Audio Systems AudiophileStyleStickerWhite2.0.png AudiophileStyleStickerWhite7.1.4.png

Link to comment
18 minutes ago, IT Freak said:

I have also NEVER seen the Studio Dot light up for a 16b / 44k MQA *
Does that mean that no artist/studio/producer likes it ?

From MQA site. 
 

Q2. What should be encoded as MQA Studio?
A2. The MQA Studio authentication should be reserved for new or remastered recordings prepared after studio preview or, for recordings where the rights holder is able to definitively assert that the source is the original release.
By implication, the further down the supply chain encoding happens, the harder it is to be certain of provenance and so recordings may only be encoded as MQA Studio when provenance is explicitly supported and provided by the content owner. Where there is doubt over the chain of custody, music should be encoded as MQA.

Link to comment
17 minutes ago, IT Freak said:

From a business point of view, I guess it would rather be in B.S. interest not to make both non-MQA and MQA versions available for streaming at the same time. ;-)

Then you must be certain MQA is paid similar model as the artists. I have no idea. But the contract is between record company and MQA in order to make the MQA available. 
There is one odd option, and that is if Tidal pays an additional fee based on total streamed tracks. Sounds highly unlikely. (How would then Roon pay, if they do at all). 

Link to comment
42 minutes ago, Cebolla said:

 

 

I believe xxHighEnd's MQA Core decoder doesn't upsample both 16 bit & 24 bit MQA files if the MQA original sample rate is less 88.2kHz - perhaps @PeterSt can confirm/comment?


Do you consider this as a “prove” to what you saying? (Yes I agree with you)

 

“Origami is always used when the input sample rate is higher than the ‘transmission rate’.

MQA can be sourced from analogue, or PCM, from A/D modulators or from DSD, but the final output is a PCM stream. Currently, all MQA on streaming services has a transmission rate of 1x (either 44.1 kHz or 48 kHz depending on the ‘family’ of the original).

When the input is PCM, the output stream will have the same bit-depth as the input unless either a) Origami is used or b) the input is DSD or floating-point; in these cases, the MQA stream output will always be 24 bit. So an original at 44.1 kHz/24b will create a 24b file and 44.1kHz/16b will create a 16b file. However an original of 96kHz/16b) will generate a 48kHz/24b MQA file because Origami was used. [1]”

Taken from here. http://bobtalks.co.uk/blog/provenance/provenance-and-containers/
 

What is a bit confusing, is this is about encoding and not decoding. He avoids to speak about that.

But you will hardly find the word upsample on that site. Or anything Bob says.

 

How to understand the term ‘transmission rate’. Hopefully what is shown in pic below.
 

 

 

A7C3F101-3A3A-4CDE-B0F6-CC6884501399.jpeg

Link to comment
9 hours ago, R1200CL said:


Do you consider this as a “prove” to what you saying? (Yes I agree with you)

 

“Origami is always used when the input sample rate is higher than the ‘transmission rate’.

MQA can be sourced from analogue, or PCM, from A/D modulators or from DSD, but the final output is a PCM stream. Currently, all MQA on streaming services has a transmission rate of 1x (either 44.1 kHz or 48 kHz depending on the ‘family’ of the original).

When the input is PCM, the output stream will have the same bit-depth as the input unless either a) Origami is used or b) the input is DSD or floating-point; in these cases, the MQA stream output will always be 24 bit. So an original at 44.1 kHz/24b will create a 24b file and 44.1kHz/16b will create a 16b file. However an original of 96kHz/16b) will generate a 48kHz/24b MQA file because Origami was used. [1]”

Taken from here. http://bobtalks.co.uk/blog/provenance/provenance-and-containers/
 

What is a bit confusing, is this is about encoding and not decoding. He avoids to speak about that.

But you will hardly find the word upsample on that site. Or anything Bob says.

 

How to understand the term ‘transmission rate’. Hopefully what is shown in pic below.
 

 

 

A7C3F101-3A3A-4CDE-B0F6-CC6884501399.jpeg

That 'transmission rate' is just how the flac file looks like that you are streaming before any unfolding is done


All MQA's are
- 16/24 bit 44.1/48k MQAs wrapped in 16/24 bit 44.1/48k flacs 
- xx bit 88.2/176.4/352.8kHZ MQAs folded and wrapped into 24 bit 44.1kHz flacs
- xx bit 48/96/192k/384kHz MQAs folded and wrapped into 24 bit 48kHz flacs

So you'll always end up with a 16/24 bit 44.1/48 kHz flac. That is what they call 'transmission rate'.

Unfortunately there have been issues where 16bit 44.1k MQAs were mistakenly wrapped in 24b 44.1k flacs.
That happened at the time they converted those millions of Warner tracks to MQA. They forgot to reset the flag of the flac encoder to 16bit ... and a 24bit flac came out with every lower byte of the 24 bit PCM data equal to 0.
(I checked this for some tracks and that was indeed the case). They did correct this for most albums by now.

And then you also have the MQA-CD ... A format that we should avoid. I did some testing there and chopping of the lower byte of a 24bit MQA makes it sound worse.

Link to comment

This is how MQA is now promoting MQA-CD : 

https://www.mqa.co.uk/newsroom/news/eudora-records
 

I extracted this from that page :
 

MQA-CD

The performances were recorded in September 2019 at the Auditorio San Francisco in Ávila, Spain and captured in DSD256. The original master was then encoded in MQA and converted directly into CD format (44.1kHz/16bit). MQA-CD brings these benefits:

- MQA preserves delicate time resolution in the sound, capturing the spatial acoustics of the venue. MQA avoids ‘smearing’ distortions which are inevitable when making a normal CD;
- MQA-CD is fully compatible, no special equipment is needed to play it and many of the sound benefits are available to everyone;
- If the CD is played back using an MQA-enabled device, the file can ‘unfold’ to as high as 352.8 kHz, delivering the best possible sound approved by the label.

Purchase the MQA-CD of Gaëlle Solal's 'Tuhu' from the Eudora Records website.


However when you buy those digital tracks (flacs) those are 24bit flacs.
It is however still possible they packed 16bit MQAs inside 24bit flacs (they did that before)
so I decompressed the flacs to 24bit WAVs to have a look at the 24 bit PCM stream.
 

If that 24bit PCM stream represents a 16bit MQA then every third byte would be ZERO.
And this is NOT THE CASE. So we are listening to 24bit MQA and not to 16bit MQA-CD here !


The tracks are available on Tidal as well ... same thing here : 24bit MQAs not 16bit MQA-CD 

So one is promoting 16bit MQA-CD by letting us listen to 24bit MQA tracks ??? 🤥

 

P.S. I did the test and chopped of the lower byte of the 24bit MQA to make it a REAL 16bit MQA-CD
The official MQA Tag Renamer app confirmed the 16bit MQA-CD I produced was genuine and
the MQA-DAC also accepted it : showing the MQA symbol, sample rate AND studio dot as well.
... but it sounded worse than the original 24bit MQA.
This seems logical because the byte holding the folded data is not there any more.
so what does the DAC do with MQA-CDs ? They call it 'unfold' between apostrophes (see extracted text above).

Link to comment
1 hour ago, bogi said:

The lowest bit of the MQA CD contains MQA control stream so the PCM part is at most 15 bits.

Yes you confirm what I was talking about : MQA-CD is 16 bit. 
That Eudora album is promoted as MQA-CD but the tracks available for download / streaming are genuine 24 bit MQA ones. NOT 16 bit MQA-CD ones !

http://www.bobtalks.co.uk/blog/science-mqa/16b-mqa-what-is-it/

Extracted from there :

Quote

In 2) above, an MQA 16-bit file was made by first optimally encoding to 24 bits and then removing the lower 8 bits. But if we know the file is only for MQA-CD, then the encoder uses different optimisations to squeeze even more from the CD.


Now those files are not only for MQA-CD since they released 24bit MQA's on Tidal and for purchase.
So... the 16bit MQA-CD is created by removing the lower bits of the 24bit MQA.

They don't let us hear how the real MQA-CD would sound like ... why is that ?

Because the stripped byte holds the folding data and without it it sounds bad.

My point is they talk about MQA-CD (16b) but the example they give is NOT MQA-CD as it is 24b.

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...