About MQA from MERIDIAN AUDIO

[louisxiawei]

Few months ago, I purchased Meridian explorer². The main new feature is to support the new incoming music file: MQA. 
 
However, I don't think the Explorer2 is a decent DAC+amplifier compared to some I have (like ARCAM irDAC). But still curious that has any one pay attention on those new format?
 
Quote from the website:
 
MQA is an end-to-end process which consideres the AD converters from the original recording sessions and the actual DAC used for playback and reduce its digital ’fingerprint’. It sounds best through an MQA decoder which confirms this studio recording as you listen, but even with no decoder you will enjoy the deblur of the recording side. Using pioneering scientific research into how people hear, MQA technology captures the full magic of an original audio performance in a file size that's small enough to stream or download. 
 
Does any one have better MQA decoder unlike my explorer and can share your experience of the new MQA file?
 
 
You can download some MQA file from here: http://www.2l.no/hires/index.html
 
 
Let me know your guys' thought.
 

 

[spruce music]

Have you upgraded the firmware in your Explorer so it is fully MQA enabled yet?  I think it was released a week or so ago.

 

[icebear]

1. there are already threads about MQA ...
2. Who in their right mind would start buying another new file format and a proprietary new DAC to decode this new quote "music origami"

 
 
From the 2L link above:
[About the MQA files MQA uses a process called ‘music origami’ to ‘fold’ a high-sample-rate signal down to a smaller,
lower-data-rate file which can be played back without a decoder. 2L-048 is folded once to 48 kHz 24bit; 2L-120 has no folds; the others are folded three times from 352.8 to 44.1 kHz 24 bit.]

 

[spruce music]

  1. there are already threads about MQA ...
2. Who in their right mind would start buying another new file format and a proprietary new DAC to decode this new quote "music origami"

 
 
From the 2L link above:
[About the MQA files MQA uses a process called ‘music origami’ to ‘fold’ a high-sample-rate signal down to a smaller,
lower-data-rate file which can be played back without a decoder. 2L-048 is folded once to 48 kHz 24bit; 2L-120 has no folds; the others are folded three times from 352.8 to 44.1 kHz 24 bit.]

Well the MQA thing is full of hype though it does genuinely do some neat things.  Promises to be higher fidelity, and appears to provide lower bandwidth streaming of larger files.  So the hope in buying into it is better sound.  I think the jury is out on that being true or not.  I wouldn't paint anyone willing to try it with such a broad brush of negativity however.  The Explorer 2 is a nice little DAC even without MQA.

 

[louisxiawei]

I upgraded my firmware few days ago. But it's a really a matter of comparison between what objectives. 
 
Assuming I can tell the difference between the MQA and normal CD quality file based on MQA decoder DAC. (MQA better than normal lossless file). But if let you compare MQA file on MQA decoder like explorer2 with combination of other serious DAC and Amplifier. Don't really think the music file will make any huge difference. That's my personal opinion. 
 
There are little MQA decoder device out there to let you judge the MQA file objectively in my opinion.

 

[castleofargh]

the purpose is to get a little better then CD resolution without adding too much weight to the streaming of the music. so if you don't plan to use it to stream music, and if you're not into "highres sounds night and day better", then MQA won't change a thing for you. now what I like about this format, it's that unlike DSD, DVD audio, superdisc... it actually serves a purpose for once, streaming. so it's not just a "go buy new stuff for something that is the same as before, just more annoying".

 

[watchnerd]

  the purpose is to get a little better then CD resolution 

 
What does better than CD resolution mean?

 

[spruce music]

   
What does better than CD resolution mean?

Supposedly undecoded MQA will play as redbook.  Yet will have less 'temporal' blur than plain redbook.  Decoded MQA will unfold into higher sample rate files with much improved temporal resolution compared to CD or even other high sample rate files.  And will do this while streaming if need be in file sizes smaller than redbook. 

 

[watchnerd]

 
Supposedly undecoded MQA will play as redbook.  Yet will have less 'temporal' blur than plain redbook.  Decoded MQA will unfold into higher sample rate files with much improved temporal resolution compared to CD or even other high sample rate files.  And will do this while streaming if need be in file sizes smaller than redbook. 

 
What does temporal blur mean?
 
Better impulse response?  If so, that's dependent upon the filter....
 
That can't mean less jitter, as that's a run-time issue.

 

[spruce music]

   
What does temporal blur mean?
 
Better impulse response?  If so, that's dependent upon the filter....

Yes, pretty much better impulse response and yes the filters.  Don't know how much you have kept up with the MQA claims, and patents and AES papers.  Hard to sort out what it is, but filtering is one of the core things along with the folding process for the undecoded file.  They in some cases say they can go back and undo the original ADC filter and implement a better one to get a result that would have occurred with better ADC filtering in the first place.  So they claim with their filtering the blur is less with a 24/96 MQA file than 32/768 PCM.
 

 

[watchnerd]

Okay, but there's no free lunch.
 
A gentler filter slope will lead to better impulse response, but also to more noise.
 
That's just how the math works.

 

[castleofargh]

warning pure fishing from a guy who extrapolated ideas from a marketing video! do not trust!!!

 
I got that strange idea that they were pretty much making one resolution(like CD maybe) from highres, and then out of the highres file they would take a few more bits to actually encode information (higher sample rate)?  is that totally wrong? the guy talking in the videos made me get that idea that they knew how 24bit was a waste of good space, so they would go for enough to sound clean(redbook?)and use the rest to stock other information that would be decoded somehow when turned back into PCM.
so lower sample rate than highres, maybe a little less bits too, a default file that can be read as is (not caring about the rest as it would be super quiet information), and when possible, that extra data would be decoded to turn the file into a little more than it is (more samples?).
 
 
 
ok now shoot me down ^_^

 

[watchnerd]

  warning pure fishing from a guy who extrapolated ideas from a marketing video! do not trust!!!

 
I got that strange idea that they were pretty much making one resolution(like CD maybe) from highres, and then out of the highres file they would take a few more bits to actually encode information (higher sample rate)?  is that totally wrong? the guy talking in the videos made me get that idea that they knew how 24bit was a waste of good space, so they would go for enough to sound clean(redbook?)and use the rest to stock other information that would be decoded somehow when turned back into PCM.
so lower sample rate than highres, maybe a little less bits too, a default file that can be read as is (not caring about the rest as it would be super quiet information), and when possible, that extra data would be decoded to turn the file into a little more than it is (more samples?).
 
 
 
ok now shoot me down ^_^

 
Uh...wut?
 
I can't parse it well enough to shoot it down.  It sounds like you're saying they're doing on-the-fly SRC and bit-depth truncation, in which case, BFD...Roon does this on the fly when I stream high-res stuff from my Roon server to AppleTV.
 
(Then again, Roon used to be part of Meridian, so maybe not total coincidence...)

 

[spruce music]

One thing you could do is read the patent.  Don't have it handy, but have seen threads with links to it.  I think I may have posted it here at one time.
 
I think they are leaving the 13 most significant bits intact.  They fold ultrasonic info into the lower three bits.  Most ultrasonic info is so low in level that is all you need to encode it. Once you split the band at 20 khz the above 20 khz samples would usually be zeros in the upper bits.  So you can throw those away and lose nothing.  They do this in a way you supposedly get 15 bits accurately without decoding.  With decoding the lower bits can be decoded into the ultrasonic portion and put together with the top 13 bits to get a file with higher sampling rates and a few more bits of bit depth. 
 
What they are doing with filtering is a separate issue.  Even when they unfold into some very high sample rates they actually are only going for 30 khz response though supposedly with filtering that has very little time blur (their marketing words). 
 
The use of subtractive dither and compressed sensing are how they manage this in the relatively small file size.

 

[ralphp@optonline]

Now hold on just a minute!
 
I happen to really like the sound of temporal blur(ring),it's that dreaded melodic blurring that drives me crazy. Not mention spacial, dynamic and rhythmic (related to but not the same as temporal) blurring. Plus the newly discovered gravitational blurring (caused by gravitational waves). Makes one wonder how we manage to be able to listen to any music.