I’ve noticed when playing Tidal mqa on my SP1000 the green light far outweighs the blue light. Very few mqa albums seem to be certified by the performer and are just run through the sampling machine. I feel Neal has it right.
Had 3 Pono’s including one I had signed by a few Prog musicians.
A post was merged into an existing topic: Tidal - All Neil Young albums before 1982 are gone. Thanks TIDAL!
Found out couple days ago about Tidal been ditching and replacing the lossless RedBook FLAC’s to MQA. That might explain my increased listening fatigue! Yeah, I don’t use MQA DAC, and won’t get one, lossy is lossy (pseudo lossless). its like MP3 all again. 
Can’t get Qobuz here in Finland but luckily there is Deezer HIFI lossless which i just subscribed and find it sounding more airy and relaxed than these 13-bit MQA’s. Too bad no Roon integration for Deezer. So, if i continue with Deezer, no Roon option for me. 
It’s funny. I would start a new thread titled “MQA not disappointing” but I feel like that would be needlessly provocative. I now have a bunch of Schiit (read: seriously opposed to MQA) DACs including a Gumby MB and a Bifrost 2, as well as an SMSL M500 (which has its issues with audible clicks during format changes, but serves just fine as an MQA and non-MQA DAC.
So I set up my gungnir and my M500 into my Mjolnir 2 along with my favorite Valvo 6201 tubes and my HD-6XXs. I run them both from the same Pi2AES, the Gumby via BNC/SPDIF and the M500 via USB. At first I listen to only non-MQA tracks from my local redbook library, then I listen to non-MQA Tidal content. Gungnir is lovely, M500 is workmanlike, same track over and over again. Then I listen to hi-res, mostly 24/96 from local. Again, Gungnir wins in a country mile. I go back and switch the inputs to make sure it’s not insanity. It’s different sure— I believe the BNC is a way better input for the Gungnir, and the SMSL is nothing amazing to write home about. I switch the balanced and unbalanced SE. Yep, there are differences, but nothing to get all het up about.
Then I switch into MQA land. Beethoven’s 1802 Testaments. Argerich’s Chopin. Coltrane. Lots of Coltrane. The SMSL is doing something different. It’s definitely not draining or disappointing. It’s a little… umm… maybe overproduced, like when you’re at a concert and the tubes in the stacks have all kinds of harmonic distortion and it sounds nice. But it’s not bad - and the detail retrieval is great. But I’m not sold on it… at any price anyways.
I do like MQA reproduction on many but not all tracks. But if the price of admission is that the redbook recordings go away, I will be seriously bummed. Especially for historic but not popular content where I think there isn’t that much call for a good look at remastering it again if at all. But some of the recordings are lovely - in different ways.
Basically, I get being an MQA skeptic, even a hater of the business model. But I think you owe yourself a chance to listen side by side to some of the well recorded new items in a setting where they can shine. And I think most people will probably consider that there is something real there, sometimes.
And how much does a “setting where they can shine” cost?
$1, 239 and 23 cents!
But seriously - I hope your question was rhetorical!
Ah, but it’s not!
MQA is not at all like MP3.
“…there is Deezer HIFI lossless which i just subscribed and find it sounding more airy and relaxed”
Now - that statement I can’t argue with! Who am I to tell you that you don’t hear what you hear?
However, I could point out that you may possibly be experiencing ‘expectation bias’ which leads you psychologically to the view that ‘lossy’ MQA must sound worse. Your pejorative reference to ‘13 bit MQA files’ sort of leads me towards this possibility. It may also be the case that your hi-fi system (you don’t mention what you use) may be particularly unsuited to playing MQA files. Of course there is also the possibility that MQA files genuinely sound worse to you and you might just be able to distinguish between MQA and non MQA files in a blind test.
Maybe not a plethora of possibilities, but certainly more than the single one that you suggest?
MQA has never been 13 bit resolution. This claim is due to a misunderstanding of the figures in one of the MQA patent applications and has been propagated by audiophiles ever since. As for the companies 16b/44.1 kHz products, Bob Stuart has written on his blog that they are all audibly 16b or greater.
…if one sprinkles pixi-dust on them. (And for full disclosure, the exact same dust can be sprinkled on any other non-MQA stream, whether it’s interpolation filtering or upconversion, or both.)
Assuming one has the means to sprinkle this ‘pixie dust’, which includes their filtering corrections to the A/D converters, the overall filtering chain on both the A/D and D/A sides, noise shaping, subtractive dither, and in the case of MQA-CD, some form (proprietary and undisclosed) of higher frequency packing. Following their claims on their website, they have been collecting data on the A/D converters in historical use for years, so would have the data base to know how to apply their A/D “correction”. Given all of that, I don’t see that anyone else is in a position to reproduce the same product. You are still free to like it or not.
At a risk of starting another religious war… But. The entire statement of theirs (delivered by you) is nothing but a cool-sounding 100% marketing mumbo-jumbo. I am not an audio signal processing guy, but I am an RF modem signal-processing guy in the area of modern commercial and military communication systems. And what you’re talking is a standard signal processing bag of tricks every experienced signal-processing designer has in their bag. As for the
One would be stupidly insane collecting the nonlinearity data of each DAC and ADC “in a database” (what, in the EVM form, and then?). The way it is handled - after you build an end to end system, you “sound” (test) it and generate a calibration/compensation digital filter. Modern design dictates that. MQA guys claim doing it too (while some reverse-engineering indicates that they actually reuse the same filters again and again across multiple DACs within a family). Which at the end is the right thing - DAC variations are often negligible compared to many other effects (so yet another MQA marketing baloney).
I am. It’s useful to know the field before applying RF modem design issues to audio. Psychoacoustics never enters military comms. The detail in MQA processing is under wraps obviously. But their processing begins with an analysis of the music file and a choice of appropriate filters for both sides of the chain based on it. As far as DACs, I believe their ‘last stage’ processing is within the DAC itself for licensed DACs, and not dependent on the upsampling filter choices that precede it. (IIUC)
BTW, audio converters and errors have changed substantially over the years since CDs originated.
Well, the argument quality speaks pretty loud to what kind of an expert one is. 
Here are the “simple facts”: (1) MQA was invented with hi-res in mind, unlike MP3. Its claim-of-fame is to intelligently remove (and then restore) inaudible portions of the “wider, high-res” audio baseband. And, for the sake of discussion, it does it well. Again, entirely to reduce transmit bandwidth/storage size.
However, (2) It does require MQA-compatible reconstruction. Without it - for the listeners who do not have a MQA-certified DAC - the sound is noticeably degraded (compared to “real” hi-res). (3) Over an MQA DAC, the MQA technology cannot produce “improvements” to the audio recording-reproduction that a non-MQA process can’t and does not offer as well.
“Physics and information theory belong to everyone, not just to MQA, Ltd” (And do not be mistaken, my goal here is not to flame with @robbi_burdeck, but rather to share my opinion/experience with the forum. But I am thankful to @robbi_burdeck for healthy stimuli.)
Without an MQA DAC, the music file is played back as CD, not high res. If Roon or another core decoder is used, the sound is expected to be intermediate. The company states this and says the licensed DAC gives best quality.
A non-MQA process does not “correct” for the A/D and D/A filtering. If those are important, your statement is inaccurate. Moreover, the use of complementary short filters on both sides of the chain is not available without a complete codec. The use of short filters is tied to MQA’s ideas on the importance of time resolution.
I’m glad you referred to these as your opinion since there are no verified or published listening tests supporting any of this, either your opinions or MQA’s claims. It would be useful if those were done.
“A non-MQA process does not “correct” for the A/D and D/A filtering. If those are important, your statement is inaccurate” - Adaptive digital filtering is a standard feature of most modern digital-to-analog converters, whether implemented based on FPGA+DAC, integrated general-purpose SoC, or dedicated audio DAC chip (like ESS Sabre). They all have the ability to convolutionally correct the signal, most do it (sometimes after a few iterations of firmware.) Filtering serves multiple purposes and can be/is implemented at encoding, decoding and/or on both sides.
“The use of short filters is tied to MQA’s ideas on the importance of time resolution.” - Again, MQA does not have exclusivity in any aspect of sound reproduction, and especially not in constriction of impulse response filters.
You correct filtering by pre- or post-filtering with another filter(s). In this case, one of several designs that removes the pre and post ring of the original filter, thereby changing the combined impulse response. There are other aspects that can be changed, for example removal of modulation noise and (some) correction for failures related to dithering. You might read through their white glove treatment on restoring old CDs to get some idea of the processes involved.
The concepts of MQA have been published in a peer reviewed journal. Listening tests are still important.
Sorry, but if they “begin” with a choice of filters applied post-recording, how can MQA be:
1 - Lossless;
2 - Unaltered;
3 - Just like the author “intended”, as they always say?
Pray tell.
Well,
Can you please explain the process of MQA folding and unfolding? Maybe that helps to understand, why there are statements like MQA is 13 bit, MQA is not lossless, but on the other side MQA is HiRes.
Agh.
The simplest place to start is reading Bob Stuart’s folding analysis on the MQA website under “Bob talks”. From there you can read their patents and peer reviewed papers for an overview.
The 13 bit comment is a misunderstanding of their patent diagrams, and is explained in the text of the patent. Patents are in any case only illustrative and not the actual product.
MQA obviously is not lossless in the sense of reproducing the exact output prior to encoding or it couldn’t reduce the data rate. It is said to be lossless because it retains the audible data plus much of the existing ultrasonic region but not the noise, neither that at very high ultrasonic frequencies nor that lying at bit depths well below the noise floor of the recording. The retained region, according to their research, is the same as what you would hear in a standard PCM high res recording except that additional inaudible noise would also be present.
I don’t know if that helps.