MQA and Naim

jmtennapel posted:

Hmack posted:

I don’t want to appear to be evangelical about MQA, because I amn’t. I just happen to think that Tidal MQA masters sound good and are worth having if you have access to MQA unfolding and subscribe to Tidal.

Funny, I have the opposite feeling. 

You mean you are evangelical about MQA?

Only kidding. I guess it’s possible that the specific kit used for MQA unfolding makes a difference. How do you listen to MQA files?

As in the case of a typical issue of MQA misconception, I guess that jmtennapel would use the Tidal app to listen to the MQA files on a MAC, and that explains why he does not like it understandly.

Innocent Bystander posted:

Do the fully decoded hi res files sound identical to, or better than, or worse than, the original hi res album not put through MQA?

I think you know the answer to that..unless the A in MQA stands for alchemy then the answer is obviously no...  but does it sound good enough? For some or even many lossy codecs can certainly sound good enough. just like AAC-LC .. and with the latter the BBC use that on their HD streams... although they know it’s improved with lossless bit perfect FLAC streams .. but so far they have only offered limited trials with FLAC.. so for now the Beeb are sticking with lossy none bit perfect HD streams. The masses appear not bothered about extreme fidelity/bit perfectness. as long as it sounds good... and so who can blame them... after all it’s only the snooty Radio3 brigade and a tiny minority of ‘audiophiles’ that appear to notice such things.

I wasn’t so much thinking ‘good enough’, rather whether an individual actually is able to hear a dufference (their ears, their system), and if so whether that is negative, recognising the possibility that some people may like actually prefer the altered sound: The latter suggets itself because some people on the forum despite only a partial unfold without an MQA DAC have suggested that MQA sounds good, or sounds better. Although different, it reminds me of  Rob Watts observations in relation to ground plane modulation within a DAC, Saying, IIRC, it can manifest itself as the music sounding ‘brighter’, which some people prefer to the ‘clean’ sound.

Again for many it may well be the sound is what they are expecting and so - it sounds fine. A high bit rate MP3, AAC can sound very enjoyable and musical - however compare that with an equivalent lossless master on a quality system with revealing speakers and the difference is quite marked - albeit dependent on the master. However on a system that may have issues with room reflections/speaker matching, or limitations in the electronics then I suspect in many cases the highbit rate MP3 will sound preferable to the fully bandwidth losslesss file - if you like there is less information that has to be adequately rendered... and less information to interact with the dominant information we are listening too

Now MQA may well work on similar ideas (although not as I understand it related to psychoacoustic masking) - especially with DAC reconstruction, and DAC/DSP modulation noise. The less data being transferred and processed for a given sound output may produce less ground plane modulation and other interaction artefacts on lesser DACs and transports and so sound preferable, even offset against the added noise from MQA decoding and processing.

It is the case with some HD lossless files there can be apparently a significant amount of ultrasonic noise encoded within the media. This is not useful and by definition being noise contains no information, but has to be processed - and MQA works to effectively remove this and so is absent from the processed encoding - and as such I can definitely see how this may well sound better on some systems.

Interestingly when I spoke to Naim a little while back about MQA, one of the main concerns was at the time the required MQA libraries  were deemed inefficient by Naim and they felt the added extra processing required using these libraries would work against much of what Naim values in digital replay - that is simplifying the digital processing and thereby lowering the ground plane modulation and other electromagnetic modulations .. therefore there was not seen to be much if any benefit but plenty of downside. Now I don't know what the current state of play is with the MQA libraries and whether they have improved in a way that would be necessary for Naim to seriously consider them.

Simon-in-Suffolk posted:

even offset against the added noise from MQA decoding and processing.

 

So MQA actually adds noise to the original recording? 

Peter Dinh posted:

Simon-in-Suffolk posted:

even offset against the added noise from MQA decoding and processing.

 

So MQA actually adds noise to the original recording? 

I would like to put my question another way - Is it due to the MQA design that noise will be added inevitably? And there is no way to get around it?

The only way to get around it is to increase the bandwidth of the sampled media... but then you will loose the compression... 

Peter Dinh posted:

Simon-in-Suffolk posted:

even offset against the added noise from MQA decoding and processing.

 

So MQA actually adds noise to the original recording? 

Absolutely... it’s an artefact caused by folding the aliasing frequencies into the sample media bandwidth. Careful reconstruction filtering is applied to minimise the impact of these aliased frequencies within the pass band.

Quite a good summary here in fairly basic language 

https://audiophilestyle.com/ca...s-and-cautions-r701/

Thanks Simon, so maybe we do not need to be bothered with MQA with all the hassles it creates.

Well it’s a choice isn’t it... if you enjoy it then fine... just need to ignore the marketing spin, and probably more importantly  that future DAC hardware fully differentiates between MQA and PCM filter reconstruction, as we see in the examples an MQA optimised DAC will not play back PCM as well as an optimised PCM DAC. Also MQA files reduce the quality in terms of numbers of bits and added noise compared to PCM files... so clear encoding labelling should be used to avoid unnecessary compromises for those wanting lossless media. So CD master quality it definitely isn’t.

The marketing stated benefits are largely theoretical and unproven, and don’t seem to bear upto scrutiny... I personally prefer solid foundations for my audio playback, rather than theoretical psychoacoustic benefits, that a premium is paid for.

Peter Dinh posted:

Thanks Simon, so maybe we do not need to be bothered with MQA with all the hassles it creates.

Except that it is no trouble at all for the listener, costs the same on Tidal, and sounds better, to some people at least, than standard CD quality. For me that trumps theoretical imperfections.

Peter Dinh posted:

Thanks Simon, so maybe we do not need to be bothered with MQA with all the hassles it creates.

As always a great explanation Simon, thanks for the insight. Always thought MQA would be the next holy grail....

Simon-in-Suffolk posted:

Well it’s a choice isn’t it... if you enjoy it then fine... just need to ignore the marketing spin, and probably more importantly  that future DAC hardware fully differentiates between MQA and PCM filter reconstruction, as we see in the examples an MQA optimised DAC will not play back PCM as well as an optimised PCM DAC. Also MQA files reduce the quality in terms of numbers of bits and added noise compared to PCM files... so clear encoding labelling should be used to avoid unnecessary compromises for those wanting lossless media. So CD master quality it definitely isn’t.

The marketing stated benefits are largely theoretical and unproven, and don’t seem to bear upto scrutiny... I personally prefer solid foundations for my audio playback, rather than theoretical psychoacoustic benefits, that a premium is paid for.

Simon - can I ask if you have listened to MQA or are you basing your opinion of MQA sound quality on the supposition that any digital file that is not bit-perfect cannot possibly sound good, or on papers that proport to debunk the very concept of MQA? I notice that you have provided links to one or two such papers.

It is inarguable that the resolution of MQA master files is not quite as 'high' as that of the original hi-res bandwidth files. However, whilst the files that are offered by Tidal as MQA masters may not be quite the bandwidth of 'standard' hi-res files, they do provide a higher bandwidth than standard CD quality equivalent files. Tidal does not (as yet anyway) use MQA on its standard CD quality files.

I do take slight issue with your post that argues: 

"A high bit rate MP3, AAC can sound very enjoyable and musical - however compare that with an equivalent lossless master on a quality system with revealing speakers and the difference is quite marked - albeit dependent on the master. However on a system that may have issues with room reflections/speaker matching, or limitations in the electronics then I suspect in many cases the highbit rate MP3 will sound preferable to the fully bandwidth losslesss file - if you like there is less information that has to be adequately rendered... and less information to interact with the dominant information we are listening too"

I can envisage situations where your statement will hold absolutely true, and I might very well have completely gone along with your logic had I not heard the benefits (to my ears) of MQA in respect of Tidal's MQA masters. I have listened to high bit rate MP3 files, and whilst they can sound 'pleasant' on some hardware, distinct limitations can normally be heard when replayed on high quality equipment. I have not heard the same limiting issues when listening to MQA high res masters. My system details can be found in my profile, and I should add that (because of a temporary problem with the Krell power amp in my second system) I have been testing my Brooklyn DAC and MQA playback in my main system which comprises Lindemann amplification and Magneplanar speakers. I believe that this system can be claimed to be a pretty decent quality system with revealing speakers. 

I am also unsure about your claim that inclusion of an MQA filter will result in a DAC that will not play PCM as well as an optimised PCM DAC. Now the Mytek Brooklyn+ is a fine DAC although it is admittedly by no means state of the art. However, even the Brooklyn+ includes both an MQA filter and a number of separate PCM filters. I would have thought that the inclusion of multiple filters (in terms of development and cost of implementation) such as those on Chord DACs for instance would have a similar effect on value for money and ultimate performance. Now although the Brooklyn+ is by no means state of the art, the Meridian Ultra and the DCS Rossini and Vivaldi DACs certainly come close to that definition. Are you suggesting that PCM quality of these DACs is compromised by the inclusion of an MQA filter?

Now for a slight backtrack on some of my previous posts. I had previously suggested that the Brooklyn+ DAC sounded slightly better than my old Chord Hugo DAC. I have just returned from holiday to find that my 'broken' Chord Hugo has been returned to me and is back in working order with a new battery. I can only assume that issues with the battery must have had an effect on the sound of the Hugo over recent months, because it now definitely sounds better than it has for some time. I haven't had time yet to fully compare it with the Brooklyn+, but I am now not so sure that the Brooklyn+ is the better DAC with PCM files. I recall that you used the term 'musical' when posting a preference for the Hugo1 over the newer Hugo2. The term 'musical' can mean different things to different people, but I agree that the term can be applied to the Hugo1 - there is a 'subtlety' and 'refinement' to the Hugo that just might not be quite there with the Mytek, although the Mytek has some advantages in other areas.

As I have said before, my main focus is on CD quality and high res PCM, since that occupies by far the bulk of my listening material. MQA would just be a bonus for me. My Linn Klimax DS/1 is simply better than the Hugo1 or Mytek Brooklyn+ whatever the format thrown at it. The Hugo1 and Brooklyn+ are both very good value and pretty close to each other in terms of sound quality. They run each other pretty close, and although I am of the opinion that Tidal MQA masters are simply 'better' than the non MQA equivalents on Tidal, I believe that I may now just slightly prefer the Hugo1 with PCM files and this would probably be decisive in respect of my decision about which DAC I will retain for my second system - I'm just not sure yet.

What I am sure of though, despite the documents I have read purporting to de-bunk MQA and the extremely acerbic almost evangelical anti-MQA posts I have seen in a number of forums, I firmly believe that MQA (and specifically MQA in respect of Tidal high res masters) is very worthwhile having. There are also a number of highly respected authors and high-fi equipment designers who are prepared to voice their positive views about MQA.

I may well end up not having the ability to unfold MQA in any of my systems (if I choose to retain the Hugo1 in my second system). However, if this happens then it will be down to limitations of the DAC rather than issues with MQA, and if I were now in the market for a high end streamer or DAC for my first system, then I would definitely be auditioning the likes of the DCS Rossini and Vivaldi, or indeed even the Meridian Ultra specifically because of their inclusion of MQA alongside the likes of the Linn Klimax, Tidal Camira and Naim ND555 - and if a hardware or software option comes along (possibly not Roon) to allow me to fully unfold MQA before it reaches my Klimax DS/1 or Hugo then I'll definitely be pursuing that option with conviction.     

Hmack, when I had a Tidal account I used to listen to MQA out of curiosity,  but it often sounded somewhat artificial and processed ... and this was consistent across various platforms, although some tracks did sound rather attractive on my more basic equipment, .. particularly with my iPad Tidal app with my ear buds . If you search back I wrote about it on this forum at the time when MQA was new to Tidal.

Good luck with your DACs, just be careful on their reconstruction architecture or you could end up in an expensive cul de sac.

But don’t get me wrong, just like AAC, Ogg Vorbis and MP3 I think MQA is a clever compromise lossy compression technology with some media rights management advantages , but as a design engineer what I don’t like is the apparent marketing hood winking and questionable peer reviews of some of MQA Ltd claims with the industry. Just be honest and say MQA is a new licenced lossy codec that supports a degree of backwards compatibility with limited compromise, and I think a lot of people will them leave MQA to its own devices...what many of us object to is its extravagant hyperbole to make it sound  something it isn’t.

Hi, [@mention:1566878603876589] Was your audition of Tidal MQA Masters (or locally stored MQA files) prior to the ability to make the 1st unfold in the Roon Core, i.e. before release 1.5?

Naim’s implementation of Tidal integration doesn’t allow access to any of the MQA versions, only taking the standard versions, & CD level will only be if you have the HiFi level subscription. This is an API library access control, I confirmed with Naim about a year ago.

So for MQA playback in a Naim Streamer/DAC environment, you either need to be used a Roon playback path or connecting with a PC using the Tidal app. Let us discount the 2nd path, as this will be sub-optimal given the noisy PC. So with a Roon Core, the MQA files can be selected in Tidal (or played back from local filestore, if you have the MQA files).

Now many are experience SQ differences in the platform they run their Roon Core on, before we even get to the Endpoint technology, with dedicated hardware surpassing commodity hardware.

However, I have now experienced 3 systems where playback of the MQA versions had excellent SQ - all had a Roon Core (one on a Windows based NUC, one on QNAP NAS & one on ROCK), one had a ND555 with 2x555DR (282 with 500DR), one was bridged UPNP into an NDS/555DR and the 3rd was Meridian DSP8000 (Ruby Editions) with the full MQA path (all unfolds). There is definitely something there.

Recent support for MQA has been from dCS & Mark Levinson, so it will be interesting to see what other announcements come out next week, with CES 2019 taking place.

Hi, I listened to MQA on various software renderer providing first level of over sampling reconstruction, ie so called ‘unfolding’ .. this was with streamed and locally cached content.

Yes it could sound quite attractive the way it processed some audio... it’s hard to tell whether that from the original master, or how MQA was processing the audio, as it was hit and miss. I wish MQA would have the confidence to demonstrate a range of side by side comparisons of lossless vs MQA of the same original master so one could objectively hear the relative merits of each.

i guess it might depend on ultimately how you listen to audio... there is added high frequency imaging distortion with MQA, necessarily on how the processing algorithm works, and I guess if one picks up on that it’s going to stand in the way and add a degree of artificiality to the audio. However if one doesn’t pick up on that or one’s speakers and/or system don’t resolve to that degree I guess this becomes moot. I certainly accept not many systems have a true resolution of 24 bit and so robbing a few bits for the MQA sub band encoding isn’t going to be too problematic, however at 16 bit, I think this can become more of an issue... 

But yes interesting to see what happens at the Consumer Electronic Show, given the slow down in China, it will be interesting to see the big bets manufacturers are backing. MQA has had a few false starts.. I wonder if this year MQA Ltd get their breakthrough.

Yesterday TIDAL release new Android APP that can do MQA 1st Unfold. I just try it on my HTC U11 (does not have any MAQ capability), so far so good.

Using software to do MQA 1st Unfold is clever for streaming on computer (Roon) or smartphone (TIDAL); I usually prefer MQA version of TIDAL music with or without MQA capable DACs.

Dear All, Thank you for all the posts - there are very inspiring but I have a very simple question: Can we easily admit that quality of listened music in FLAC files streamed by TIDAL is better than MQA files. Simply yes/no assuming that we have capable system (the same for both type of files)

supmario posted:

Dear All, Thank you for all the posts - there are very inspiring but I have a very simple question: Can we easily admit that quality of listened music in FLAC files streamed by TIDAL is better than MQA files. Simply yes/no assuming that we have capable system (the same for both type of files)

If I understand your question correctly, then the simple answer is 'No' if the same MQA capable DAC is used for comparison. 

One thing I am sure of is that most Tidal MQA Masters sound better than the same non MQA files on Tidal when played via the same DAC (in my case the Mytek Brooklyn+).

Whether or not Tidal MQA files sound as good as non MQA files on Tidal when using a non MQA streamer/DAC (such as Naim) is not of importance because you will simply be able to play the non MQA FLAC version. 

Well, listening sessions, equipment comparisons and audio memory (at least mine) are fickle things!

I suggested above that my newly refurbished Chord Hugo1 was sounding splendid on my main system when fed by my Sonore microRendu and Gustard U12 USB to Spdif converter, and so it is - much better than it ever did on my second system. So much so that I was convinced that when I swapped in the Mytek Brooklyn+ DAC today (without the Gustard) in place of the Hugo, the Mytek could not possibly be as good and that I would emerge from the comparison with the Hugo as my clear favourite. However, I was completely wrong. The Mytek today also sounds better than I remember it sounding before the Christmas period, and specifically better than before in respect of its 'musicality' (subtlety, refinement and lack of listening fatigue) - the very attributes that I find compelling when listening to the Hugo. The Mytek now sounds to me very nearly the equal of the Chord in these respects, but also has a bigger (not louder) sound and soundstage with clearer separation of individual instruments.

Coupled with the fact that the Mytek offers the bonus of DSD and Tidal MQA (the Chord does DSD but not when used in conjunction with the Gustard) and sampling frequencies up to 352.8 kHz (although I'm not sure that I hear any benefits of anything above 192kHz or possibly even 96 kHz), the Mytek now looks like winning the day in the end. I'm just going to have to repeat these tests when I get my second system up an running again.          

supmario posted:

Dear All, Thank you for all the posts - there are very inspiring but I have a very simple question: Can we easily admit that quality of listened music in FLAC files streamed by TIDAL is better than MQA files. Simply yes/no assuming that we have capable system (the same for both type of files)

Personally, I can hear no difference between Non- MQA FLAC or MQA files from Tidal when using my Mytek Brooklyn +. But then again, I can't really tell the difference between standard CD quality files or hi-res files either.

Graham in Sussex posted:

supmario posted:

Dear All, Thank you for all the posts - there are very inspiring but I have a very simple question: Can we easily admit that quality of listened music in FLAC files streamed by TIDAL is better than MQA files. Simply yes/no assuming that we have capable system (the same for both type of files)

Personally, I can hear no difference between Non- MQA FLAC or MQA files from Tidal when using my Mytek Brooklyn +. But then again, I can't really tell the difference between standard CD quality files or hi-res files either.

I think differences that are in reality quite subtle can often be exaggerated unintentionally when they are described, and that applies to my own comparisons as well, as well as across the gamut of hi-fi reviews in general. 

The differences that I hear when comparing MQA and non-MQA files from Tidal are indeed often quite subtle but nevertheless they are there, and to my ears (admittedly a relatively old design) definitely worthwhile. I too often struggle to hear major differences between different resolution versions of PCM files, although I can normally appreciate the difference between, say, 16 bit, 44.1kHz and 24 bit, 192kHz using the same master. I struggle (for instance with the Norwegian 2l files) to hear the difference between say 24bit, 96kHz and 24bit, 192 or 352.8kHz.

I would describe the differences I hear in respect of both MQA hi-res and PCM hi-res files as being in the area of 'refinement' and 'subtlety' of the sound' (the very opposite of what some people call a more 'hi-fi' sound), rather than any magical transformation where 'one hears all sorts of things that weren't there before'. Nevertheless, the words I use to describe the differences may convey different meanings to different individuals.

I find that MQA (and some hi-res PCM) files are worthwhile having on my Mytek Brooklyn+ DAC, and some hi-res files are worthwhile having on my Hugo1 and Linn Klimax DS/1, However, I accept that the differences for many may not be quite as dramatic as some reviews (including my own) might seem to convey.      

I presume  preference for MQA by numbers of people is a consequence of its tweaking of the sound - it is clearly not because the playback is more accurate. So something in the filtering, for want of a better description, is producing an effect that some people find desirable. As I have suggested before perhaps it is time to start looking at deliberate digital manipulation of the signal to ‘improve’ it, as can be done effectively with some room issues, but in a more subtle way. If the MQA sound truly is a desirable one maybe that should be incorporated into the original mastering of all digital recordings - No need then for MQA capable devices at home, and even vinyl made from those recordings could benefit...

Innocent Bystander posted:

I presume  preference for MQA by numbers of people is a consequence of its tweaking of the sound - it is clearly not because the playback is more accurate. So something in the filtering, for want of a better description, is producing an effect that some people find desirable. As I have suggested before perhaps it is time to start looking at deliberate digital manipulation of the signal to ‘improve’ it, as can be done effectively with some room issues, but in a more subtle way. If the MQA sound truly is a desirable one maybe that should be incorporated into the original mastering of all digital recordings - No need then for MQA capable devices at home, and even vinyl made from those recordings could benefit...

Except that the folding/unfolding concept would still be required (from a bandwidth perspective) in order for service providers such as Tidal to be able to offer their hi-res MQA Masters relatively inexpensively.

Whilst I have enjoyed the (hi-res) Tidal MQA Masters, I haven't yet heard anything to suggest that an MQA master sounds better than a PCM file of equivalent resolution. Others do claim to have heard this, but I haven't personally carried out enough tests nor had sufficient local MQA material to hand in order to allow me to carry out this testing.