Why MQA is bad and Roon (shouldn't bother) (shouldn't be bothering) shouldn't have bothered with it :)

(Petri K.) #77

I don’t know how this works technically, but I have done some listening tests with my little MQA-DAC (Pro-Ject Pre Box S2 Digital). I found out that the MQA sounded worse than Red Book with this DAC in my opinion. I don’t understand what’s the point of MQA if it just sounds worse. I just don’t seem like the sound signature of the MQA, which is kind of like adding a very slight veil and smoothing over everything. I’m sure someone will like it though.

(Sean) #78

Not a problem Petri and subjective opinions should always be respected - I don’t care either side of the argument to be honest, so I’m neutral (in terms SQ that is)

But my questions are only related to the technical nitty gritty.

Also, your DAC is doing rendering/up-sampling past the 1st unfold, which is exactly what I’m asking we don’t consider in my questions (there’s heaps of that analysis on CA Forum). I’m only asking about the 1st unfold.

(Henry) #79

That is OK, I don’t take forum discussion personally! But apart from not being openly hostile to MQA, how have the likes of me aided it. I subscribe to Tidal, I am a Roon lifer and I bought an Explorer2. That is it and probably no more or less than many here. I occasionally stream MQA and I have samples in my library. Again, not much different to you probably. If this happens it isn’t because of me or anyone pro or indifferent. It is because the industry wants it and they are still in control.


Tidal and the likes are incentivised to adopt it because it saves bandwidth cost, hence uptake in streaming is likely. Without customers there would be no streaming. If you stream you by definition support it and you are facilitating the control you refer to.

I don’t steam, at all.

(Sean) #81

Hmmm I may have to do my own analysis. There are so many strong opinions out there, it’s hard to trust anyone’s opinion. Better off seeing these differences with my own eyes lol

This album is on Tidal, only in MQA 353kHz. The master file is available for purchase at 353kHZ (DXD) and 24/96kHz and CD quality.

I can then compare vs the output of Roon’s 1st unfold.

By compare, I don’t mean with my ears - I mean using something like Audio Hijack and Spek on MacOS, to compare the music content in the file.

(John B) #82

I asked the same question here and got this reply.


(Sean) #83

I don’t know where he got the information to say “Yes the unfolded MQA can have more information than CD quality”

But I’ve just seen it with my own eyes. Roon’s 1st unfold looks pretty close to the 24/96kHz version I purchased. Unless I’m missing something or doing something very wrong - a high possibility, this is not my area of expertise.

All come from the same DXD (353kHz) master.

All you need is 20 minutes and a credit card to prove to yourself.

For Mac users:

  1. Buy all the copies you need (the artists will love you) - https://shop.klicktrack.com/2l/439809

  2. To capture the MQA 1st unfold from Roon’s Tidal stream - https://rogueamoeba.com/audiohijack/

  3. To compare the outputs of CD quality vs 24/96kHz vs 1st MQA unfold - http://spek.cc/

(Peter Lie) #84

It should. I don’t think there is any surprise here. That’s why MQA Core has more information than Red Book CD in terms of kHz.

(Sean) #85

‘Should’ because MQA Ltd (or any other company that may have a conflict of interest) says so is different to seeing it with one’s own eyes.

But yes, the 1st MQA unfold looks to me better than CD quality, all other things being equal (same master) and putting aside file size and all the other well known discussions (all valid but not part of my questions here).

An interesting exercise for myself. I encourage everyone else to convince themselves. Use the 3 steps I used above (for MacOS).

(Peter Lie) #86

I said it should because an in-depth study carried out by a prominent author in the anti-MQA camp actually supported what you’ve found. It’s the other aspects they criticize it for.

(Sean) #87

Noted. If you had shared that link to me earlier when I asked for such info, you could have saved me from purchasing more copies of this album !! :grin:

All said in good fun of course.

I have trust issues anyway. It was a nice exercise to see with my own eyes.

And nobody should trust me. Check for yourself if you’re skeptical.

(Peter Lie) #88

I assumed you’ve read it before since you’re familiar with CA discussions.

(Sean) #89

I definitely missed that one. Doesn’t matter, it’s always nice to personally verify anyway.

(Sean) #90

For completeness, just found Archi’s comments (that confirms what Peter said above) that are relevant to my questions (we all know about all the extra stuff that is also important to consider but not part of my questions):

"Objectively with the songs I examined, the software decoder works well to reconstruct what looks like the equivalent 24/96 download."


"Bottom line: TIDAL/MQA streaming does sound like the equivalent 24/96 downloads based on what I have heard and the test results"

(Jeff) #95

Bob himself in an interview said it’s close to 18/96 unfolded and the CA guys like @jussi_laako have measured closer to 17/96. Those are the 2 resources I used to make my comment.

(Henry) #96

You stand by your principles and that is fine. But I was streaming long before MQA came along as well as buying music in various formats and occasional excursions out to watch live performances. I make my contributions and am comfortable with some of that going to places like MQA if Roon and Tidal get their cut too. I know how MQA are getting their product out there. As I have already said, we were never the ones they had to convince.

(John) #97

Just chiming in with a reminder to keep things civil. I’ve deleted a few posts/responses that delved into the inappropriate territory. Please try and stick to the technical discussion at hand here, and do not make assumptions or accusations on someones tastes or morals based on someone stance about how they like their 1’s and 0’s delivered to them.


(Anders Vinberg) #98

Sean, to your original question about MQA first unfold only vs. Redbook:
The technical description is quite simple.
MQA unfolded includes frequencies from 22k to 48k (roughly) that are not included in Redbook.
To fit that in, it sacrifices some of the 24 bits, and it is selective about how much of the high frequency material is included.

(Sean) #99

Noted. I respectfully ignored what MQA Ltd have said (slight conflict of interest) and just wanted to see with my own eyes.

While I share a lot of the valid long term concerns, I’m quite impressed with the 1st unfold.

And that’s with music with a much higher dynamic range than I normally listen to.

And to go back to my own questions that I answered myself, the 1st unfold offers more music content than CD quality streaming (when the same master is compared).

Unless you’re not a fan of the actual master it came from and unless your music itself has greater than 16bit dynamic range, having Roon do the 1st unfold of a Tidal stream is essentially like streaming the 24/96 FLAC file that you’d buy from HDTracks or 2L or wherever (assuming same master of course).

Important note: all of this is only relevant to the 1st unfold that Roon now does. None of my questions (and my own answers) relate to any MQA stuff beyond the 1st unfold. There’s enough of that info out there.

I think I’m done with this little experiment. Back to listening to some music.


You have take into consideration that ‘de-blurring’ filters are applied during encoding stage. This one helps to reducing ‘ringing’ effects. Test has shown different type of digital filters applied during recording can significantly affect the tonal quality of the music contents. Most studios uses sharp FIR filters while MQA uses minimum phase slow roll off similar to Ayre implementation.

MQA is not about hearing more music contents but the effect of ‘de-blurring’ make up the most significant part of the SQ. If ‘de-blurring’ is to apply on a standard Hi-Res PCM, it will definitely sound like one or even better coz it is lossless.

To fully complete the ‘de-blurring’ process one need a renderer to do it; this ultimately translates the best SQ MQA has to offer.