@ setevev1
Hi Steve,
thanks !
Meridian shows what sampling frequency have the masters which were used to produce the MQA.
Magnificat is made from a 352.8 (DXD) file, Nielsen from 44.1 (CD). The LEDs show this information.
And the MODE LED:
blue LED means - the MQA Studio file - the studio has signed it off
green LED means - the MQA file - the artist personally has signed it off.
The audio is identical.
(Mr. Lindberg from 2L was so kind to answer my questions)
In the CD case you see the only one LED, the MQA mode LED (because it is the same as 44/48 kHz white LED)
Have look at the User Guide (lower, right corner)
I like the 2L recordings very much, too
I installed latest firmware on 818v3 and DSP8000SE. Release notes indicate speakers should indicate Studio, but I don’t see anything. Have Magnificat and others, both albums and samples.
Thanks for the reply. it sounds like the typical generic response a manufacturer would provide. My dealer is investigating the issue as we speak so we will see his version. It took a while to get it over Christmas while MA was waiting for some stock items to come from the motherland. I would just hope that the Christmas production included the latest.
I have to accept your observation as this stuff is all subjective, that being said you would need some point of reference ie. hearing the non MQA file on that system or an intimate familiarity with that system. I have had the opportunity of hearing MQA on a few occasions and on material that I am very familiar with the difference was noteworthy. This was not just my observation but that of several co-workers and several clients who had an opportunity to hear a demonstration by Meridian. I will say that some tracks that I heard did not do much for me but they were tracks that I am unfamiliar with and not to my tastes so I have no point of reference.
Please define what you mean by lossy in context of MQA. there is no loss of musical information in MQA files but as said previously you will not get the information above 20K which is not music.
“Lossy” has become a characterization that doesn’t fit the new world well.
Of course it is lossy in the sense that the full rectangular dataspace of 24 bits * 192k samples cannot be folded into 24 * 44 and then restored.
Compressibility requires either that data is redundant (like PKZIP) or deemed unimportant (like MP3). (All zeros is one form of redundancy.)
MQA origami is based on an assumption of what out of the rectangular dataspace is unimportant. The belief is that this assumption is sounder than the assumptions of MP3.
So it would “data-lossy” but not “sound-lossy”.
All the arguments to show that it is data-lossy is beside the point.
The only thing that matters is sound-lossy, and that has to be evaluated. Although the reasoning appears plausible.
Some object on another basis, that compression is not useful, “I have plenty of bandwidth to download 24*352 here in my home in Switzerland.” That is not a technical judgment but a business one. And wildly out of touch with reality, in my view.
Anders,
That’s a good way of looking at it, data loss vs sound loss - I agree that MQA seems to be data lossy, but not sound lossy (at least not much) There is some sound loss I believe in the first stage of the folding process, the 48Khz to 72khz range. I believe this range is compressed but the MQA guys suggest the compresson at this step improves SQ but tightening up time smearing.
Lossy as in if you MQA a high res file, say 24/96, you cannot recover that file back. This is a simple mathematical definition of lossy. As for musical info, this is something of debate.
An assumption that requires you to make assumptions as to what can be heard and what cannot.
Couple of points:
1- I am all for a lossy compression that allows me to stream better sounding content with less bandwidth - no argument there
2- Lossy means losss, it’s math, period. A 24/192 file will recover the info it will recover according to Nyquist. There are many ways to avoid the phase shifts from the brickwall filter requred. Again this is math, not something of opinion.
3- The folding, the loss produced in the spectrum, IS something of opinion in the sense that you might be able to hear it.
Again, I’ve listened to MQA and the sound was better than the original file - ar least the ones Meridian played, I had no say. But so does playing through HQPlayer or Audirvana with upsampling.
Why do you say this? If you download the various MQA and PCM files available from 2L and play them though a DAC without MQA decoding capability does that not do just this? Or am I misunderstanding something?
I have purchased a couple of albums from 2L (192/24 PCM) and they are absolutely incredible sounding.
Really, maybe some of the best sounding files I own. I would be pleasantly surprised to hear an MQA version of the same file and go “wow, that’s better…”.
I’ve done this in Tidal and the MQA version are clearly better to my ears. All the 2L albums are available in Tidal in MQA for the few lucky ones that got the Master setting early. So give a bigger selection of music to test. Also the latest Enya as well. There’s a few tracks where it’s harder to tell but for most, the MQA always sounds better, more natural. And this is without an MQA dac. Pretty easy to switch back and forth to HIFI vs MQA.
Take a pristine 24/192 recording, encoded it with MQA (this means deblurring plus folding), decode back to PCM 24/192 (this is what you can’t do today), do a diff (might need to adjust levels if they were modified by MQA) and see the spectrum and listen to the diff.
It would also be nice to compare the two files when just the folding part of MQA has been used, but that is very unlikely to ever be possible since MQA is not likely to provide us with any of those details or a binary that just folds/unfolds.
Today the only way to decode MQA is to play the file. MQA does not allow for software decoding to a PCM stream. In fact, Auralic had this already working for CES 2016 and MQA pulled their license at the very last minute.
!