MQA have publicly acknowledged that the max resolution of an unfolded MQA file is 17/96. It can also be less, depending on the source master. This happens in the first unfold. You can’t magically create bits that aren’t there. Read how MQA works. The rest is upsampling. No argument from MQA about this.
The thing is, you get lost in the digital and forget to listen to the analog.
You don’t know the full MQA process and make assumptions based on your understanding to cobble together an explanation that fits your paradigm.
I would much sooner judge with my ears and listen to technical explanations from people with pier reviewed research and development.
What is the difference between the MQA 1. unfold and the full-unfold? As already mentioned it is in the filters and other corrections that are controlled via the metadata to correct the characteristics of the recording ADC and provide a complete, authenticated path from recording to playback. This is NOT about a 100% bit-perfect restoration of a recording. It’s about perfect playback in the time domain and avoiding artifacts caused by the ADC on the one hand and by the reconstruction filters at a cutoff frequency of 20 kHz (CD) on the other. The articafts are around 3.5 kHz and are clearly audible. A resolution of 96kHz shifts the artifacts at least beyond of 12kHz. A 192 kHz resolution shift it into the range beyond 20 kHz. Of course, this also applies to HiRes recordings, e.g. from Qobuz! Only in the direct comparison full-unfold MQA still sound better, at least for me, like the HiRes recordings in the same resolution. The 1st unfold just doesn’t deliver these corrections. But get a Qobuz trial subscription and compare it directly and with the MQA first unfold and the CD quality. If MQA sounds better than a CD, you’ve already won. If Qobuz sounds better, then think about a change or maybe you’ll get a small MQA DAC, e.g. the Dragonfly Red or Cobalt and just try a full unfold yourself.
Hope that helps a little bit
I get it Chris: YOU’VE swallowed all sorts of MQA BS, and when provided with FACTS, you “attack the messenger” - because the facts don’t fit your false preconceived notion about MQA.
What I’ve written about MQA isn’t my opinion, it’s based on MQA’s own papers and interviews with Bob Stuart confirming what I wrote, among others.
Instead of more BS attacking me, let’s see you disprove the content I wrote.
And quoting cute marketing phrases about “blurring” doesn’t count.
All I wrote was that the so called second unfold is a form of filtering/upsampling and MAY not sound superior to non MQA upsampling that can be done without an MQA DAC. The first part of that statement is a fact, the second part is just a suggestion that the user posing the question about MQA try similar upsampling and see what he thinks.
Why does that disturb you so much? Were you unaware the “second unfold” is upsampling to the target rate of the original master? Hard for you to deal with the truth?
This is an excellent explanation in layman’s term.
Great explanation of what is happening during the fold and unfolding of the MQA file.
This is untrue. See video above.
What’s untrue is the video. The guy often doesn’t know what he is talking about- not just about MQA, but about other topics. He also mentions in at least one of his videos (I didn’t rewatch this one to check) about a “third unfold”, which doesn’t exist.
Putting something in a video doesn’t make it true.
What I wrote was correct. Try to do real research instead of relying on Hans’ videos and maybe you will figure it out.
There can be a third unfold if higher resolutions are folded.
Nope. The MQA algorithm doesn’t sample any content over 48khz. Any source content above 17 bits depth and 48khz frequency is discarded-even if the source is, say, a 24/705khz sample rate file. So the max resolution of an MQA file is the same as that of a 96khz master - and at max depth rate of 17 bits . Anything in the original file above the 17/96k level is DISCARDED.
A file that’s been converted to MQA maxes out at 17/96 or 17/88k and that’s folded into a 24/48 or 24/44.1 MQA container file for distribution.
If say, it’s a 24/48k MQA file, the first unfold brings it to 17/96, and anything above that is upsampling and filtering with an MQA filter. The so-called third unfold is just one of the MQA filters. Unlike MQA claims, there is no “unfolding” of more data after the first unfold. The filters are not DAC specific, but a generic set of MQA filters (except for a handful of models where the manufacturer has produced unique filters) and there is nothing especially unique about them.
This is exactly what’s written in the MQA tech papers and also confirmed by Bob Stuart in video interviews. Anything else is just marketing speak that has nothing to do with reality.
What higher resolutions are you referring to? What’s your source for your info? Blah blah put out by people like Hans who don’t know what they are talking about doesn’t count. It’s fine if you like MQA. But please stop parroting their marketing speak that has nothing to do with audio reality.
If you’d care to watch before commenting, I only selected a portion of the video that shows what is happening during the folding and unfolding process. It shows the 2 folds and how they are placed with the compressed file. If there is only upsampling in the 2nd unfold then what is folded in during the 2nd folding process during the packing of the file?
Don’t worry about it. The reality is there is lots of speculation about MQA by people a lot less qualified than Hans, but they will swear that the stuff they are regurgitating from the same tired and unsubstantiated sources is somehow more true.
Just trying to find the “truth”. I’ve never seen the actual folding process more distinctly presented as Hans does in this clip. If, as some say, he’s not very qualified, then I wish someone could show me, graphically (like the video), what is actually happening during the folding and unfolding process and what is the information in the 1st and 2nd folds.
Personally, I see the Fervent MQA detractors are set in the a paradigm of thinking that is not compatible with what MQA is about. That’s OK but that doesn’t make me wrong either. I wonder how much un biased listening is done?
For myself, I just get stuck into the music on my very capable MQA system and am blown away by the quality which in all practical terms makes their arguments mute to me.
Perception is all.
Like I said, do the reading and figure it out. I explained to you correctly the outline of what is happening. “Folding” is just an MQA marketing term. They call the second “unfold” an unfold because if they called it upsampling then all of a sudden MQA doesn’t look so special.
Note that my detractors here can’t actually refute the facts I presented, so they resort to ad hominem comments and other generalites that are off topic and meanigless.
The difference between what I wrote and what you wrote is that I stated facts that are in technical papers and not denied by MQA when specifically asked if they are true. Please find me a single reliable source specifically contradicting what I wrote.
You, on the other hand, are just making unsubstantiated speculation without the requisite knowledge.
Do you want to link to those technical papers and the review process that verifies them? Because that is how this actually works. Note that I made no claims about MQA, only about the people who feel qualified to dismiss it with no more knowledge than myself. If you have that knowledge, please share.
It is very easy to verify. 2L has high-res recordings in a number of formats, including MQA. As others have already done, you can download them and open them up in any number of audio analysis or processing programs and see that there is zero information above 48 kHz in any of the MQA files, whereas it still exists in the original high-res versions. You might also observe that the original high-res versions have large amounts of ADC noise above 48kHz, so that can’t be “folded” into any kind of smaller file anyway.
As well you should. The truth is that there is nothing to the MQA marketing speak called “2nd fold” other than upsampling at the DAC with slow rolloff, high IMD “min phase” filters. Even MQA supporters such as John Atkinson (technical editor for Stereophile) and Jim Austin (Editor of Stereophile) admit this is the case. Bob Stuart himself does as well, in a round about way, but you have to read between the lines of his/MQA non-standard use of signal processing terms (e.g., “resolution”, etc.).
I listen to a lot of MQA through my Oppo 203 which is not MQA capable. I also listen to a lot using my Dragonfly Cobalt that is MQA capable. While I enjoy both, the Dragonfly sounds better to me. I think it sounds as good as Qobuz 96 and 192, just a little different. It seems to me to be a little punchier with a little more bass.
It’s all good. Listen to what you like, not what someone else likes.