MQA Users & Discussion Thread

[tripitaka]

3 minutes ago, wklie said:

 

My understanding is this and I have not seen any evidence that this would be wrong: With HiFi Plus, you pay more and can get the Master quality level (up to MQA-coded 24/44.1 and 24/48 delivery in FLAC).  With HiFi (non Plus), you pay less and are not allowed to get the Master quality level.  I assume HiFi quality is always the same resulting in 16/44.1 FLAC regardless of which subscription you have.

 

Well that's all I need to know, in that case HiFi NonPlus is exactly what I want, regardless of cost - so cheaper is a bonus

[PKay]

4 minutes ago, wklie said:

 

My understanding is this and I have not seen any evidence that this would be wrong: With HiFi Plus, you pay more and can get the Master quality level (up to MQA-coded 24/44.1 and 24/48 delivery in FLAC).  With HiFi (non Plus), you pay less and are not allowed to get the Master quality level.  I assume HiFi quality is always the same resulting in 16/44.1 FLAC regardless of which subscription you have.

Unfortunately there isn't enough clarity and the customer is required to assume.  It should be clearly stated.

[wklie]

23 hours ago, a.dent said:

It just seems odd that Roon is processing the two MQA files differently.

 

If Hi Fi is a max of CD resolution why would Roon upsample unless it is being told to do so by Tidal? This means the file is probably the 13 bit MQA file and is definitely of lower quality than an equivalent CD rip or Qobuz file.

 

Why does Roon not upsample the MQA 96 and 192 files and send them straight through as 44.1?

 

1. Roon always uses a consistent set of MQA decoding logic. 

 

2. You got the 16/44.1 MQA file.  Roon MQA Core decoding is activated as expected.

 

3. MQA 96kHz or MQA 192kHz music are delivered in 48kHz FLAC.  The MQA signalling in the 48kHz file is destroyed by asking for a 44.1kHz file at HiFi quality instead of the Master quality.

[wklie]

20 hours ago, Eggcup the Dafter said:

it would be easier for them to keep the standard res files than to resample from MQA 48/24, as far as I can see.

 

Every streaming service provider needs to deliver the same music at multiple bit rates.  Tidal needs to deliver 96kbps AAC, 320kbps AAC, 16/44.1 FLAC, 24-bit MQA for different subscription types or settings for the same music.  Remember that in addition to HiFi Plus, HiFi, there is also the lossy Premium subscription.

 

Even Spotify has multiple bitrates to choose from for the same music with the same lossy subscription: 96kbps, 160kbps, 256kbps, 320kbps.

Edited April 29, 2021 by wklie

[Grant Slack]

2 hours ago, Snoopy8 said:

This was so predictable!  Cannot beat the inconvenient MQA measurements, let's close the ASR thread.  No idea why there is advance notice to close in about 8 hours' time.?

 

What next?  Perhaps claim they are MQA neutral, and thus all MQA discussions will be banned at ASR ???

 

Hi Snoopy,

 

Where is the thread closure notice?

 

cheers

Grant

[Guest Eggcup the Dafter]

2 hours ago, wklie said:

 

Tidal is not doing the upsampling.  It it delivering to you a 16/44.1 MQA file.  MQA Core decoder (in Roon or any MQA software or MQA digital output hardware) outputs 88.2kHz.  This is normal.

It’s worth remembering the “for user convenience” upscaler in this context. Yes, some MQA DACs may not even be doing the “unfold”.

 

What’s more, if I remember the Stereophile article correctly we need 96kHz before gaining that timing advantage MQA claims to deliver.

 

I wonder what Bob Stuart really thinks about this - is it really what he intended all those years ago when he started on this path?

Edited April 29, 2021 by Eggcup the Dafter

[Guest]

22 minutes ago, Grant Slack said:

 

Hi Snoopy,

 

Where is the thread closure notice?

 

cheers

Grant

https://www.audiosciencereview.com/forum/index.php?threads/mqa-deep-dive-i-published-music-on-tidal-to-test-mqa.22549/page-91#post-763590

[wklie]

On 28/04/2021 at 2:06 PM, El Tel said:

I tried to see months ago if I could go through Roon, to Tidal, and pull one of the Led Zep tracks that was labelled MQA and push it through to one of my Sonos devices (non-MQA) and see if it would just play the 44.1/16 FLAC which Sonos is very capable of doing. The signal path indicated that the source was Tidal 44.1/16 FLAC and I mentioned all this on the Roon forum. One of the support guys picked-up on my findings and screen captures and said he thought that Roon would be handling the first MQA unfold and passing the rest over to the Sonos to play as a native FLAC. He suggested the signal path was not showing the real situation and said he would raise a ticket to the devs about it. Haven't seen any further outcome...

 

I found your posts there.  No, I don't think this is what he said.  Roon MQA Core decoding cannot occur in your scenario because Sonos do not support over 48kHz.

 

Since no MQA Core decoding occurred, Roon should be sending the MQA CD just as your native platform gets the same MQA CD from Tidal, if you picked the same version in both players.  There are four versions of the same album.

 

He said he suspected it is a display bug with the 24 bit thing, but I disagree on this.  I claim that it is not a bug, I claim Roon is functioning and reporting correctly regarding the bit depth processing in that scenario.

Edited April 29, 2021 by wklie

[davewantsmoore]

46 minutes ago, Snoopy8 said:

https://www.audiosciencereview.com/forum/index.php?threads/mqa-deep-dive-i-published-music-on-tidal-to-test-mqa.22549/page-91#post-763590

 

LOL... what a **** show.

 

The MQA encoder definitely made a mess of his files.

[LHC]

1 hour ago, Eggcup the Dafter said:

It’s worth remembering the “for user convenience” upscaler in this context. Yes, some MQA DACs may not even be doing the “unfold”.

 

What’s more, if I remember the Stereophile article correctly we need 96kHz before gaining that timing advantage MQA claims to deliver.

 

I wonder what Bob Stuart really thinks about this - is it really what he intended all those years ago when he started on this path?

IMO the answer is 'no'. In my naive pov Bob Stuart started MQA with noble intentions, but the practical reality of turning it into a profitable business made those intentions into roadkills. 

[Nada]

13 hours ago, LHC said:

IMO the answer is 'no'. In my naive pov Bob Stuart started MQA with noble intentions, but the practical reality of turning it into a profitable business made those intentions into roadkills. 

 

A codec developed with  "noble intentions"   like flac is

1. free

2. no DRM

3. actually  adds a useful function

4. doesn't waste audiophiles time

5. lossless

6. honest

 

I dont know the intentions behind creating MQA but to me its a  festering blight.

 

Its wasted my time. Time I wont get back. Time I could have been enjoying listening to music.

Edited April 29, 2021 by Nada
Edited to target the file not the man.

[davewantsmoore]

1 hour ago, Nada said:

2. no DRM

3. actually  adds a useful function

 

DRM is a useful function.

[LHC]

Bob Stuart's peers do hold him at a higher regard.

 

https://www.stereonet.co.uk/features/inside-track-bob-stuart-mqa

 

For the first time ever in its twenty-year history, an audio engineer has won the Royal Academy of Engineering’s Prince Philip Medal. Of all the possible recipients in the hi-fi industry, it surely had to be Bob Stuart. He got the gong for, “his exceptional contribution to audio engineering, which has changed the way we listen to music and experience films”, no less. Previous recipients include the inventor of the turbojet engine, Air Commodore Sir Frank Whittle; geothermal power innovator, Lucien Bronicki and the electrical engineer who revolutionised fibre optics, Dr Charles Kao. Could it be that the engineering community is suddenly taking sound seriously?

 

“One never truly knows which aspects prompted this award,” Bob tells me, “but it’s given once every two years to an engineer who they feel has made a contribution to engineering, education and knowledge, I guess. It’s interesting, the people who’ve won it in the past have done quite different things. I think the two things that are unusual in my career are the work we did around standards – Meridian Lossless Packing (MLP) in DVD-Audio and Blu-ray, and more recently Master Quality Authenticated (MQA).”

Edited April 29, 2021 by LHC

[Guest Eggcup the Dafter]

4 hours ago, davewantsmoore said:

 

LOL... what a **** show.

 

The MQA encoder definitely made a mess of his files.

They responded, arguing his files weren't prepared properly. The basis for that argument tells us as much about MQA as his tests did. The full response and his own reply are in the first post on the ASR thread which is linked somewhere on the first two pages here.
I'm guessing that at least some files in the batch processing might be similarly mangled. Hopefully the majors are doing QA on their files.

 

 

1 hour ago, davewantsmoore said:

 

DRM is a useful function.

If it protects all of the rights holders with no undue problems for customers. I'm not sure how the multiple licensing and cost model with money going to MQA Limited for a lossy and possibly SQ damaging streaming service, which is not open, not properly independently verified to work, and that pays **** all to the artists, can be seen as useful.

 

I still want the right to back up my purchased media though - and, importantly, in the context of this particular debate, the right to purchase.

[Guest Eggcup the Dafter]

28 minutes ago, LHC said:

Bob Stuart's peers do hold him at a higher regard.

 

https://www.stereonet.co.uk/features/inside-track-bob-stuart-mqa

 

For the first time ever in its twenty-year history, an audio engineer has won the Royal Academy of Engineering’s Prince Philip Medal. Of all the possible recipients in the hi-fi industry, it surely had to be Bob Stuart. He got the gong for, “his exceptional contribution to audio engineering, which has changed the way we listen to music and experience films”, no less. Previous recipients include the inventor of the turbojet engine, Air Commodore Sir Frank Whittle; geothermal power innovator, Lucien Bronicki and the electrical engineer who revolutionised fibre optics, Dr Charles Kao. Could it be that the engineering community is suddenly taking sound seriously?

 

“One never truly knows which aspects prompted this award,” Bob tells me, “but it’s given once every two years to an engineer who they feel has made a contribution to engineering, education and knowledge, I guess. It’s interesting, the people who’ve won it in the past have done quite different things. I think the two things that are unusual in my career are the work we did around standards – Meridian Lossless Packing (MLP) in DVD-Audio and Blu-ray, and more recently Master Quality Authenticated (MQA).”

I'd like to remind you in the context of that remark, that the late Michael Gerzon (also of Ambisonics fame) did much of the acclaimed work on noise shaping dither, and did the original development of MLP that underlays that part of Bob Stuart's reputation. Gerzon was a true genius and he needs to be remembered in any discussion of Meridian's work in this area. I just wonder, given Gerzon's interests and background in recording, whether the original germ of MQA was actually Gerzon's as well, though Stuart obviously did the actual work underpinning and developing it?
 

That's not to undermine Bob Stuart's career, which even if it is leading to problems for some of us with MQA now is still massively important in the development and implementation of digital audio as we know it now. He deserved the award.

 

[Ittaku]

Right about now I can smugly say I'm glad I've always bought my own high-res audio files and not succumbed to the temptation to stream.

[Guest]

Some interesting comments on AS from the GoldenOne on the now closed ASR thread

https://audiophilestyle.com/forums/topic/30381-mqa-is-vaporware/?do=findComment&comment=1128812

 

Quote

Prior to this, one of the moderators requested a phone call with me. They said (and no I am not joking) that there were too many people criticising Amir for the MQA related comments he'd made and that they couldn't keep up with the moderation, so they wanted to lock the thread and wanted me to start a new one, in which they would not allow anyone to bring up the statements Amir had made regarding MQA. Because ASR/Amir needed to "remain neutral".

 

Amir flooded the ASR thread 2 days ago, which in turn led to its closing down....

Edited April 29, 2021 by Snoopy8
Typo

[NonPlayableCharacter]

14 hours ago, wklie said:

 

I found your posts there.  No, I don't think this is what he said.  Roon MQA Core decoding cannot occur in your scenario because Sonos do not support over 48kHz.

 

Since no MQA Core decoding occurred, Roon should be sending the MQA CD just as your native platform gets the same MQA CD from Tidal, if you picked the same version in both players.  There are four versions of the same album.

 

He said he suspected it is a display bug with the 24 bit thing, but I disagree on this.  I claim that it is not a bug, I claim Roon is functioning and reporting correctly regarding the bit depth processing in that scenario.

 

Once again, Peter, your contributions are awesome. Thank you.

 

I apologise for misrepresenting the exchange that I had on the Roon forum - I did not go back and read it again before my post on here. It was a genuine mistake whereby I had mistakenly conflated several other exchanges on Roon forum, from memory alone, and distilled them into what I wrote on this thread. I had no intention to make the situation appear worse than it already is.

 

This whole mess is still leaving me bewildered and confused.

Edited April 29, 2021 by El Tel
Providing clarity.

[Guest Eggcup the Dafter]

15 hours ago, wklie said:

He said he suspected it is a display bug with the 24 bit thing, but I disagree on this.  I claim that it is not a bug, I claim Roon is functioning and reporting correctly regarding the bit depth processing in that scenario.

If Roon is set to pass through MQA, that might be another explanation.

You'd then see the file play at 44.1/24 rather than 16 as that is the original file format. 

 

DId you/can you tell Roon to only play hi-fi and lower from Tidal, as you can in the Tidal player?

[davewantsmoore]

9 hours ago, Eggcup the Dafter said:

They responded, arguing his files weren't prepared properly. 

 

Yes.  They cited errors from the encoder, which he hadn't seen.

 

Overlooking the issue that he wasn't made aware of almost all the errors ...... it is reasonable to conclude that we didn't witness the expected behaviour of the encoder.

 

That being said, I agree with most of his conclusions about MQA generally.

 

I will bet my mother that the encoder will deliver much more reasonable performance, if "used properly".... although, I'd expect that performance to look very "minimum phase" (we got a glimmer of that through his tests, and from what other stuff has been shown about MQA) .... MQAs argument is that (minimum phase look) doesn't matter.... and I don't think that argument is complete bunk.     Whether that is "lossless" or not, is somewhat semantics.

 

OTOH, the fact that it seems potentially true that the encoder isn't always doing a good job with the whole "batch convert everything" is not a good thing .... (we don't know this, but it's not a completely unfair speculation) is worrying ..... especially if we end up in a world where everything is MQA from the get go (so not even the artists themselves have a non-MQA copy).

 

Obviously, the biggest obvious thing going on here, is "why didn't I get the errors .... what did I do wrong.... how do I fix it.... etc.".

 

In some ways it's understandable that the MQA encoder doesn't handle such files ..... but in other ways its "scary".

 

 

As I've said a gazillion times.   I think this line of enquiry, (while it does have merit, of course) has a danger ......   it leads the groupthink to the point where it says   "if MQA has acceptable performance then everything is ok, and we'll go along with it" ....... and/or   "MQA is flawed, but if another performant system was introduced (or MQA fixed the performance) then that would be fine".

[NonPlayableCharacter]

10 minutes ago, Eggcup the Dafter said:

DId you/can you tell Roon to only play hi-fi and lower from Tidal, as you can in the Tidal player?

 

@a.dent did all the tests using Roon in the last couple of days. He setup a throw-away Tidal Hi-Fi sub, so it would not have gone above HiFi quality in any case.

[Guest]

An almost accidental but interesting development last night in my own system. I use a Mark Levinson No519 CD/Streaming Player predominantly through Roon.

 

The No519 is a full MQA decoder, which has to be enabled in the settings of the unit. Upon getting this player nearly two years ago, one of the first things I did was enable MQA capability. Digital in my system is good, but not as good as I have experienced with other streamers at various times. Perfectly acceptable though.

 

I don't ever "seek" out MQA tracks, but as we know they regularly come up via TIDAL. 

 

Last night I decided to turn MQA functionality off entirely in the No519 settings. It wasn't an planned process of turning it off and then playing known tracks via Roon (NAS stored files) and Qobuz files, but something I had been meaning to do reading these threads recently. But I did that, and lo and behold, I noticed vastly improved sound quality across the board. It didn't immediately occur to me why as I hadn't set out specifically to do this experiment. But as I was listening away, track after track sounded better than what it has previously. Make of this what you will, but MQA functionality will be staying firmly turned off in my system and this particular source component. I will have a little conversation with some tech/engineers I know at Mark Levinson, and get their take on how MQA is implemented, if I can.

I had been an early advocate of MQA. But my only and very limited experience with it upon first release was with one of the only tracks that had been made available at the time - Adele, Hello, played via Meridian DSP8000SE with native MQA capability. It was a remarkable sound, but no one knew the source or master of that track. It won me over though - but sadly I have never ever experienced something like that with MQA since. It's just seemed to be just another format - I've certainly not heard any perceived benefit or improvement personally.

 

 

[aussievintage]

14 minutes ago, Marc said:

Last night I decided to turn MQA functionality off entirely in the No519 settings. It wasn't an planned process of turning it off and then playing known tracks via Roon (NAS stored files) and Qobuz files, but something I had been meaning to do reading these threads recently. But I did that, and lo and behold, I noticed vastly improved sound quality across the board.

 

Not MQA, but my soundbar has Meridian sound processing, and the sound I like best, is when I turn it all OFF !!     I didn't buy the sound bar because of the Meridian stuff  as past experience with any "improvement" processing has always been disappointing.   MQA, to me,  is just another example.

 

17 minutes ago, Marc said:

Adele, Hello, played via Meridian DSP8000SE with native MQA capability. It was a remarkable sound, but no one knew the source or master of that track

 

Yes, choice of source material for demonstration purposes is critical - for sales.    Listening to custom chosen demos is a very bad way to determine if you like certain equipment.

 

 

[wklie]

2 hours ago, El Tel said:

This whole mess is still leaving me bewildered and confused.

 

No need to apologise.  This thing is complex and secretive, and that's why I try to provide answers wherever I can from my experience of integrating this technology.

Edited April 30, 2021 by wklie

[Guest Eggcup the Dafter]

1 hour ago, Marc said:

Last night I decided to turn MQA functionality off entirely in the No519 settings. It wasn't an planned process of turning it off and then playing known tracks via Roon (NAS stored files) and Qobuz files, but something I had been meaning to do reading these threads recently. But I did that, and lo and behold, I noticed vastly improved sound quality across the board. It didn't immediately occur to me why as I hadn't set out specifically to do this experiment. But as I was listening away, track after track sounded better than what it has previously. Make of this what you will, but MQA functionality will be staying firmly turned off in my system and this particular source component. I will have a little conversation with some tech/engineers I know at Mark Levinson, and get their take on how MQA is implemented, if I can.

I had been an early advocate of MQA. But my only and very limited experience with it upon first release was with one of the only tracks that had been made available at the time - Adele, Hello, played via Meridian DSP8000SE with native MQA capability. It was a remarkable sound, but no one knew the source or master of that track. It won me over though - but sadly I have never ever experienced something like that with MQA since. It's just seemed to be just another format - I've certainly not heard any perceived benefit or improvement personally.

 

 

One thing that I think is starting to come out of this round of MQA "research" is that not all MQA devices are the same. There are a few things we can pick up on immediately:

1) the fact that in some MQA DACs, all files whether MQA or not are pushed through the MQA decoder or its default filter. That has the capability of harming the performance with non-MQA tracks (and hey presto! MQA sounds better)

 

2) the upscaler. According to the performance graphs supplied by MQA to Stereophile, the timing advantages of unfolding only start at 96kHz or so, and therefore a simple upscaler to 88.2 may not really be an issue, but it seems to me to be a breach of the promise that MQA files get unfolded and that timing advantages are calculated from "between samples" through MQA processing. They say it's "low blur" and developed with listening tests - what happened to all the science and analysis, if they're devising part of the process through listening choices?

 

3) How are they really adapting MQA to the particular DAC? According to mansr over at AudiophileStyle, the filters look the same when measured, and only a small number of them are used by the encoder. Is that the result of adaptation being fitted by MQA (so the output looks the same) or are the filters the same and something else being done? Do MQA just sit on the firmware for each DAC for a month and pretend do adapt them?

 

If they really are adapting the DACs to MQA, I would assume that all MQA DACs would sound the same with high resolution masters. With my Dragonflys (Red and Cobalt) this actually seems to be the case: they sound different with PCM (being my normal obtuse self I actually prefer the Red) and similar with at least some MQA files, Tidal giving no idea of target resolution in its player. I'd be interested to know if that is the case for DACs with chipsets from different manufacturers.