I’m not sure I completely follow this argument - simply because I don’t have the technical expertise to do so. However, it does appear to align with my own subjective appraisal of MQA in my systems and environment.
As a matter of interest - has anyone here downloaded the high resolution music samples available from 2L and attempted to compare these samples? I have, and on each of my systems (sonore microRendu/Mytek Brooklyn+ and Linn Klimax DS/1), I am unable to consistently pick out a difference between those identical source recordings with 96kHz and 192 kHz sample rates. That is - a difference of any sort rather than a subjective preference.
This may partly due to the fact that the 2L music itself is not really to my taste, but I don’t think so. Has anyone else had the same experience?
I have heard the full MQA on a pair of Meridian 8000 ‘Ruby’ (Meridian’s 40th Anniversary versions, limited to 40 pairs) and yes wonderful - but not sure whether it was the MQA, the speakers, the room treatment, mains conditioners etc that were making it all special.
For me - unfold to 24/96 is good enough for now - most of the Tidal Masters are 24/44.1 or 24/96. Very few are 24/192. Then if there are, I either already have the PCM version in 24/192 or can find.
Of the 6k albums in my library, over half are 24-bit and continues to grow.
I did not think about this network bridge, however I did consider just doing Roon and getting the first unfold that way. It was a nice solution, and one gets the library/search function of Roon which so many find excellent. I wanted to experience full three unfolds, so went with a mqa dac.
Its too bad they were not able to turn off the mqa decode so that you could hear the difference and take out the other factors. I should be able to do that with my setup when I get it.
I agree that the misleading or incorrect information that was out there did not help. They should have made the effort to clearly explain things.
However, I disagree on the open source part. We live with closed source for lots of things, like Naim Patents, Dolby, Phillips CD, Google, numerous other manufacturer patents, I am sure Linn has lots of them, and there are many hidden fees in the audio chain. In the capitalist system, folks need to get paid. Things that are paid for with government taxpayer money should be open source, however if we are to encourage investment in research, then there has to be a way to get paid. There are not enough smart ideas in government to spurn all the cool research ideas. Government would have to fund it also. Not sure that is the best path. “Necessity is the motherhood of invention” as they say.
I do object to some folks getting filthy rich however. The founders of things like Google, who have enhanced our lives immensely, I have no problem. However those who manipulate numbers on a spreadsheet, or buy and destroy companies and jobs, not so much. As a victim of consolidation which cost me half my life’s pension, I see no gain in this. The number of consolidations that went belly up seems extraordinarily high. I also object to the $50 million golden parachutes that those executives who clearly did no good get. They deserve something, perhaps more like single digits millions, but not 50-100. Sorry off topic a bit, but somewhat germane.
Good point.
For a truly level comparison, you’d have to find a set of tracks that were all recorded originally in no higher than 16/44.1.
Then the unfolded MQA files would be in the same resolution as the non MQA treated files.
And you’d need to get quite a lot of them because some tracks respond better to MQA treatment than others.
But there are a lot of them - over a million now on Tidal I think.
In case it’s of interest, I copy below an email I received from MQA this weekend.
It only mentions 2 unfolds (not three) - although I have seen various descriptions elsewhere that slice the process into finer parts.
But this email is from the Horse’s Mouth (i.e. MQA, the organization) - including their own typos:
"MQA playback takes place in two stages:
MQA Core Decoding– “Authenticates” the stream providing provenance that the stream/file has been “signed off” buy the content owner, guaranteeing that you are getting what happened in the studio. Also, at this stage is the “First unflod” providing resolution up to 24bit / 96kHz. MQA Core Decoding can be done in software, ie Tidal Desktop Player, Amarra, Audirvana, or ROON music players as well as MQA Certified Hardware including dCS Network Bridge and Meridian’s 218. All of these will output an MQA Cores stream over their respective digital outputs to ANY DAC. MQA Core Decoding is a part of MQA’s backwards compatibility.
MQA Rendering – Provides specific DAC management that is dictated by the ADC used in the actual Master’s recording. Any additional “unfolding” to higher resolution above 24 bit/96kHz, depending on the master is also done at this stage. These steps MUST be done in hardware and are ONLY available on MQA Certified Hardware.
What the above means to your system is that you can play MQA streams at resolutions up to 24 bit/96Khz from Tidal but you will not hear full Master Quality due to the Naim NOT being MQA certified and thus not being able to perform MQA Rendering. This will still be a noticeable improvements over playing non- MQA content."
NOTE: For those who do not want to listen to MQA, or do not like the sound of MQA, please do not shoot the messenger (i.e. me) - I’m not saying everything in this email is true, only that that is what MQA, the organization, describes it as.
I’ve never even heard MQA so have no idea whether it’s actually good or not!
For those who think MQA is immoral, I guess we could discuss whether it’s immoral even to consider listening to it - but only if you’re not rude and stroppy about it!
This is confusing to me. The two steps described in the email are confusing with the link that you posted. The link suggests that step on in the email is acutally two steps. From the link I got that the MQA Rendering is for portable devices. The email does not refer to Full MQA Decode, which is the one with the highest sq according to the link. Its pretty confusing. I am not speaking from authority here, just trying to understand this all. MQA does need some help with technical writing, or I need some help with technical reading.
FWIW, I’ve discovered that I can use the Tidal desktop App on my 2015 iMac connected to my Nova using the optical out to optical in and it does that first “unfold”. It sounds pretty good, but I need to do some more direct comparisons between CD rips, hi-res downloads and MQA. I’m picking I could multi-room to my NDX2, which is in another room, and try that too.
It authenticates the sound as modified at the MQA encoding stage at a studio where the sound profile can, and I would say can’t help but be, modified by the MQA encoding processing.
This is NOT the same as the original mix master as would have been heard in the recording studio when the track was created. Some of the marketing info hints that this is the case and is highly misleading in my opinion.
This encoding and decoding destruction and reconstruction does have a reliance on the DAC reconstruction filter function. As almost commercial DACs have this function built in, there is a reliance on the physical DAC. This of course is usually done by software in the DAC itself.
See my post above… it is quite possible to do the first oversample in Naim firmware, and even create a specific reconstruction filter response software programme that is switchable in with the SHARC reconstruction filter processor to do full MQA oversample decoding (it certainly does NOT need to be done ‘hardware’… which is kind of a consumer marketing over simplification)… However I suspect the extra faffing around and possible risk to overall sound performance may deter Naim, at least for now. Currently Naim use a very simple software programme for their ‘hardware’ reconstruction filter running in the SHARC… its only a few lines long… this keeps it simple and reduces the risk of added processing noise artefacts.
When I spoke to a Naim designer about this a couple of years back, he said the MQA libraries were not sufficiently ‘mature’ for Naim standards at that time, and it was a case of a watching brief at that time.
When I quizzed the engineer on this comment, the answer was something along the lines that the required MQA libraries were too inefficient for Naim, and given the Naim standards for processing noise, they felt incorporating such software into the DAC streamer almost certainly would have a detrimental effect.
Now I have no idea whether MQA have optimised and improved their libraries to meet Naim standards since then…