MQA Users & Discussion Thread

[davewantsmoore]

4 hours ago, frednork said:

Just for my own info, isnt this exactly what the roon guy says they do? If not, how do they differ from HQplayer aside from more filters as most people seem to think they are different.

Your (4 year old) quote from Brian, would indicate they are the same.

 

... but the current roon docs, which Grant quoted say different.

 

Quote

Roon currently does not support processing DSD signals directly--if the processing is requested, Roon begins by converting DSD into an extremely high-resolution form of PCM called DXD”

 

You can read the whole thing here:

https://help.roonlabs.com/portal/en/kb/articles/conversion-to-pcm

 

In short... it does what I described for "how it's done" ... but instead of working with the audio at DSD rate and 64bit depth..... it converts it first to DXD rates (~384khz) and 64bit depth.

 

... which it can either output at this rate, or (as you see at the end of the doc) you can put it back through a SD modulator to any DSD rate you like.

 

 

All that being said.... because it's happening inside the computer, (as opposed to realtime) then it's hard to say whether one approach is inherently inferior to the other, or by how much.   The devil is in the detail.

 

This is all very different from days gone by where the idea that "oh, well you've resampled it, and so lost all the 'magic' of DSD" (for example inside a DSD audio workstation).

[frednork]

32 minutes ago, davewantsmoore said:

Your (4 year old) quote from Brian, would indicate they are the same.

 

... but the current roon docs, which Grant quoted say different.

 

 

You can read the whole thing here:

https://help.roonlabs.com/portal/en/kb/articles/conversion-to-pcm

 

In short... it does what I described for "how it's done" ... but instead of working with the audio at DSD rate and 64bit depth..... it converts it first to DXD rates (~384khz) and 64bit depth.

 

... which it can either output at this rate, or (as you see at the end of the doc) you can put it back through a SD modulator to any DSD rate you like.

 

 

All that being said.... because it's happening inside the computer, (as opposed to realtime) then it's hard to say whether one approach is inherently inferior to the other, or by how much.   The devil is in the detail.

 

This is all very different from days gone by where the idea that "oh, well you've resampled it, and so lost all the 'magic' of DSD" (for example inside a DSD audio workstation).

 

Ok thanks, no wonder I am confused, so they used to but no longer? Also the messaging on the app is not clear if the above is the case as the switch implies it is doing something different.

 

[davewantsmoore]

8 minutes ago, frednork said:

so they used to but no longer?

 

Maybe?!

 

(As I said, about Roons and facepalms)

 

8 minutes ago, frednork said:

 

The switch "native dsd processing" basically says it does it.

 

I'd say it's the docs (resample to ~384khz) which are wrong.

[frednork]

Yep, I think it does do it

 

this is with switch on

 

this is with switch off

 

[Volunteer]

On 07/06/2021 at 1:05 PM, El Tel said:

One of the things that alarms me, is that mqa has reportedly been seeking to get its DAC partners to force all conversion through the mqa decoding path, even the open formats. Why? It's clearly a crap idea to force compliance in this way when manufacturers like Lumin currently push non-mqa streams through a non-mqa DAC path in their kit. If mqa make this happen, they will have the ability to influence the sound of non-mqa streams and potentially provide a false picture of format superiority.

On 07/06/2021 at 8:42 PM, LHC said:

 

Are there any real evidence that DAC manufacturers are "force" to take this route? Evidence of such coercion would strengthen the case you are making. 

 

Not forcing, but certainly bordering on coercion... A Stereophile.com review (https://www.stereophile.com/content/aurender-a10-network-music-playerserver-measurements) of some Aurender kit a while ago uncovered that mqa "recommended" that all streams pass through the mqa upsampling process regardless of format. Aurender apparently implemented this. I'm certain that this is a standard approach for the majority of mqa DAC manufacturers with the notable exception of Lumin.

 

So now you need to not only ditch Tidal (and all the subterfuge of the HiFi tier tracks masquerading as not being mqa when they are 44.1/16 mqa CD), but you also need to acquire a DAC that either doesn't do mqa at all, or alternatively invest in kit from the likes of Lumin who steadfastly manage mqa and non-mqa through different processes. This is the only way you can be certain that other formats aren't impacted deliberately or accidentally by ham-fisted mqa upsampling.

 

It seems my conjecture on mqa hobbling other formats is not as far fetched as some believed or hoped.

[LHC]

10 hours ago, El Tel said:

Not forcing, but certainly bordering on coercion... A Stereophile.com review (https://www.stereophile.com/content/aurender-a10-network-music-playerserver-measurements) of some Aurender kit a while ago uncovered that mqa "recommended" that all streams pass through the mqa upsampling process regardless of format. Aurender apparently implemented this.

 

The A10 test bench review is an interesting use case study. The comments section is also interesting to read. Notably Stereophile highlighted the mistake made: "However, the A10's misapplication of the MQA reconstruction filter to non-MQA files stored on its internal drive means we must withhold a full recommendation for the A10 until this problem has been corrected.—John Atkinson" This criticism was by John Atkinson, a vocal champion of MQA. One could take Atkinson's word that this was an honest mistake on the part of the manufacturer (although they did not admit it as so). 

 

I think one could view the Stereophile test bench here as providing a template for testing MQA enabled DACs to ensure this mistake is never repeated again. So I don't imagine it will be ongoing concern (unless Stereophile or other testers abandon this approach in the future).

 

10 hours ago, El Tel said:

I'm certain that this is a standard approach for the majority of mqa DAC manufacturers with the notable exception of Lumin.

 

It would certainly be worth checking other DAC manufacturers to see if they have also made the same 'misapplication'. Without checking we don't know if Aurender is the exception or the norm. The MQA site lists 44 DAC partners so there are plenty of manufacturers to check. 

[Guest deanB]

Luxman D-03X. Is the signal being MQA'd if you're playing a compact disc?

[Volunteer]

22 minutes ago, LHC said:

 

The A10 test bench review is an interesting use case study. The comments section is also interesting to read. Notably Stereophile highlighted the mistake made: "However, the A10's misapplication of the MQA reconstruction filter to non-MQA files stored on its internal drive means we must withhold a full recommendation for the A10 until this problem has been corrected.—John Atkinson" This criticism was by John Atkinson, a vocal champion of MQA. One could take Atkinson's word that this was an honest mistake on the part of the manufacturer (although they did not admit it as so). 

 

I think one could view the Stereophile test bench here as providing a template for testing MQA enabled DACs to ensure this mistake is never repeated again. So I don't imagine it will be ongoing concern (unless Stereophile or other testers abandon this approach in the future).

 

 

It would certainly be worth checking other DAC manufacturers to see if they have also made the same 'misapplication'. Without checking we don't know if Aurender is the exception or the norm. The MQA site lists 44 DAC partners so there are plenty of manufacturers to check. 

Notwithstanding the comments stating that applying mqa upsampling on all formats was a direct recommendation from mqa ltd. This is the worry. Especially if later iterations of whatever filters/algorithms etc become the standard for mqa upsampling further hinder non-mqa material.

 

I wholeheartedly agree with you on checking the other manufacturers' implementations and whether they have actioned mqa ltd's recommendation.

 

My trust and respect for Lumin is huge; their implementation is one that all others should aspire to and be judged against.

[sir sanders zingmore]

16 minutes ago, El Tel said:

Notwithstanding the comments stating that applying mqa upsampling on all formats was a direct recommendation from mqa ltd. This is the worry. Especially if later iterations of whatever filters/algorithms etc become the standard for mqa upsampling further hinder non-mqa material.

 

somehow it doesn’t surprise me that MQA would recommend implementing a filter which makes alternatives sound worse. Good business strategy, but it makes a mockery of dac manufacturers if they can’t tell that it makes their product perform worse. 

[LHC]

51 minutes ago, deanB said:

Luxman D-03X. Is the signal being MQA'd if you're playing a compact disc?

 

I hope this is not an ongoing trend .

 

Interesting to compare the circuit for the D-08u which apparently is not a MQA unit. 

 

[alistairm]

Wow... I haven't read the entire thread, but this was sufficient for me to reactivate my Qobuz subscription and give it a run through Roon. 

 

I don't pretend any great knowledge or sophistication with digital and settled on Roon a while ago as the simplest way to mangage and serve up my digital library. I just run Roon off my main PC, over ethernet into an APL SNP-SR network player (based on an Auralic Aries) and then into an APL DSD-MR Mk2 DAC (which is apparently not MQA compatible). 

 

It did not take much back-to-back listening to work out that in my setup the Qobuz streaming sounds better than Tidal. Tidal still sounds good, but the Qobuz is just a little more detailed and open. 

 

I use Soundiz to sync playlists between Spotify, Apple Music, Todal and Qobuz and it does a good job. My main playlist (just stuff I "star" because I like it when I hear it on Radio Paradise) has 394 tracks in Spotify. Tidal has 381 of those tracks, and Qobuz 332. So for now I'll be keeping Tidal as well as it seems to have a bit more content in the genres I listen too, but where possible I'll listen through Qobuz. 

[sir sanders zingmore]

12 hours ago, deanB said:

Luxman D-03X. Is the signal being MQA'd if you're playing a compact disc?

Looks to me as if everything is routed through mQA no matter what

[sir sanders zingmore]

1 hour ago, alistairm said:

Wow... I haven't read the entire thread, but this was sufficient for me to reactivate my Qobuz subscription and give it a run through Roon. 

 

I don't pretend any great knowledge or sophistication with digital and settled on Roon a while ago as the simplest way to mangage and serve up my digital library. I just run Roon off my main PC, over ethernet into an APL SNP-SR network player (based on an Auralic Aries) and then into an APL DSD-MR Mk2 DAC (which is apparently not MQA compatible). 

 

It did not take much back-to-back listening to work out that in my setup the Qobuz streaming sounds better than Tidal. Tidal still sounds good, but the Qobuz is just a little more detailed and open. 

 

I use Soundiz to sync playlists between Spotify, Apple Music, Todal and Qobuz and it does a good job. My main playlist (just stuff I "star" because I like it when I hear it on Radio Paradise) has 394 tracks in Spotify. Tidal has 381 of those tracks, and Qobuz 332. So for now I'll be keeping Tidal as well as it seems to have a bit more content in the genres I listen too, but where possible I'll listen through Qobuz. 

My view was that if I’d started with Qobuz and trialled Tidal there  would almost certainly be tracks missing from Tidal. So some of it is to do with the environment you have “grown up” in. 

Of course there was stuff in my tidal playlists that wasn’t available in qobuz. My solution was to simply buy that stuff and be rid of MQA once and for all

Edited August 3, 2021 by sir sanders zingmore

[brumby]

I just paid my $229 for a year of Qobuz.  If I find a few tracks I want to hear missing, I'll simply buy them or go to Spotify lossless if and when it ever emerges.  The only glaring Qobuz omission I've found so far is Eva Cassidy and I have all her stuff ripped from my CDs anyway.  I don't want anything "proprietary" between me and my music - or between me and anything else - so that lets out MQA - and Tidal.

[Volunteer]

1 minute ago, brumby said:

If I find a few tracks I want to hear missing, I'll simply buy them or go to Spotify lossless if and when it ever emerges. 

I think the other Qobuz thread has some info about raising the issue of missing artists/albums/tracks with Qobuz directly. Sometimes it is geo-licencing, sometimes stuff is missing due to deals not being done with artists/record companies or disputes thereon, but sometimes it's an oversight and they sort it.

[alistairm]

Not to go too far off topic, but I've cancelled my Tidal subscription. Hard not to after watching the video from the OP. 

[rantan]

15 hours ago, sir sanders zingmore said:

somehow it doesn’t surprise me that MQA would recommend implementing a filter which makes alternatives sound worse. Good business strategy, but it makes a mockery of dac manufacturers if they can’t tell that it makes their product perform worse.

100%

 

IF I ever get into streaming, I will specifically avoid any streamer that has MQA built in. The same will apply for streaming providers.

 

As @brumby correctly says, avoiding anything proprietary in the "music chain" is a non starter for me.

Edited August 3, 2021 by rantan

[tripitaka]

Tidal financial results 2019

https://www.digitalmusicnews.com/2021/01/13/tidal-2019-earnings-report/

 

For FY2019, total revenue was 166M, at a net loss of 52M (compared with a 37M loss for 2018), bringing accumulated company net liabilities to 132M.

 

Regardless, the company was still valued at $302M in May this year when purchased by Square.

https://www.forbes.com/sites/korihale/2021/05/04/what-squares-350-million-tidal-acquisition-means-for-its-music-industry-ambitions/

 

Perhaps the most noteworthy thing above is speculation about all the ways in which the new owner, Square, might exploit the Tidal brand to leverage other product lines, such as Square's phone payment App.

 

Regardless of where the future leads, I don't think the integrity of audio formats and welfare of artists are likely to feature prominently in all of this

 

[alistairm]

Well... Square just agreed to pay $39B in scrip for a business that sells people stuff they don't need and can't afford and can pay for later. So Tidal went cheaply! 

[rantan]

49 minutes ago, alistairm said:

Well... Square just agreed to pay $39B in scrip for a business that sells people stuff they don't need and can't afford and can pay for later. So Tidal went cheaply! 

 

Sorry for the OT, but with respect, Afterpay is quite handy for many people who sometimes need to pay 4 instalments for more expensive items.

 

I have used it in the past to my advantage in the above scenario to buy audio gear which was heavily marked down. I knew I could afford to pay it off over 4 fortnights but didn't have the whole amount just laying around at the time.

Don't confuse the 10% of vacuous lipstick buyers with the 90% who know very well what they can afford.

Edited August 3, 2021 by rantan

[sir sanders zingmore]

5 hours ago, rantan said:

 

Sorry for the OT, but with respect, Afterpay is quite handy for many people who sometimes need to pay 4 instalments for more expensive items.

 

I have used it in the past to my advantage in the above scenario to buy audio gear which was heavily marked down. I knew I could afford to pay it off over 4 fortnights but didn't have the whole amount just laying around at the time.

Don't confuse the 10% of vacuous lipstick buyers with the 90% who know very well what they can afford.

that’s an extremely harsh assessment of a whole lot of sometimes very vulnerable people who

a) perhaps are in such dire straits that they are forced to live on whatever credit they can get and/or

b) are vulnerable to the extremely predatory behaviour exhibited by these lenders. 
 

As much as we love to hate banks, they are at least subject to a decent amount of oversight. Afterpay and their ilk are not bound by the same regulations. 
 

[BugPowderDust]

50 minutes ago, sir sanders zingmore said:

As much as we love to hate banks, they are at least subject to a decent amount of oversight. Afterpay and their ilk are not bound by the same regulations. 

This.

 

You don’t need to be Nostradamus to see that afterpay uses an ability to avoid stringent credit guidelines in a way that you wouldn’t be hard pressed to say is predatory.

 

[tripitaka]

We are getting off topic, we need to keep our outrage directed uniformly towards towards MQA

[alistairm]

26 minutes ago, tripitaka said:

We are getting off topic, we need to keep our outrage directed uniformly towards towards MQA

Amen. 
 

And my incredulity at Afterpay’s valuation was in no way meant to reflect adversely on any product that enables @rantan to indulge in his audio addiction!

[awayward]

I can sort of understand the angst towards MQA, I may be a simple man but MQA music has not restricted my access to music or cost me more to hear it, and IMHO some MQA music sounds better than the equivalent non MQA formats. I don’t need or really want it, but it exists for the moment and I’m confident it will eventually die as it has not dominated as the creator and supporters expected, just my 2 cents worth.