Fair enough. A lossy encoding. I do have to admit I have non-clarity on what non-Shannon means…
Exactly. MQA argue that all they throw away is noise under the noise-floor of the encoding (that’s the quantization bit). I think there are also one or two corner cases where they throw out additional detail because the encoder runs out of encoding space, but this is claimed to be rare.
I just listened to the MQA version of Purple Rain. Sounded way better than any other version I have heard so far. Lossy or not, it sounded good!
Ok thx. This is not unreasonable if you consider that the encoder, when faced with a 44.1/16 master, could conceivably do some upsampling itself to 88.2/24 and then “fold” that back to 44.1/24.
For example, something like this must have happened with Dire Straits’ “Brothers In Arms” which I understand was recorded to 44.1/16. The DSD version from MoFi sounds fantastic.
I agree it sounds great. I don’t think it sounds better than the high res version that came out not too long ago, but I would have to compare them closely. On the other hand, Amy Winehouse’s “Back To Black” is markedly better in MQA, and that’s a pretty compressed and distorted recording at that. That is one recording I would definitely buy in MQA format if it were available for purchase in that format.
2 Likes
That recording is so bright and thin to begin with, the softening that MQAing does could only help it.
The only “Purple Rain” MQA that I see is a 2017 Deluxe Edition release based on a 2015 24/96 remaster. If so, that is loudness war junk.
DR5 0.00 dB -7.51 dB 4:38 01-Let’s Go Crazy (2015 Paisley Park Remaster)
DR7 0.00 dB -8.64 dB 3:53 02-Take Me With U (2015 Paisley Park Remaster)
DR6 0.00 dB -8.53 dB 5:13 03-The Beautiful Ones (2015 Paisley Park Remaster)
DR7 0.00 dB -8.78 dB 3:59 04-Computer Blue (2015 Paisley Park Remaster)
DR5 0.00 dB -7.66 dB 4:15 05-Darling Nikki (2015 Paisley Park Remaster)
DR8 0.00 dB -9.55 dB 5:53 06-When Doves Cry (2015 Paisley Park Remaster)
DR9 0.00 dB -10.35 dB 2:49 07-I Would Die 4 U (2015 Paisley Park Remaster)
DR8 0.00 dB -8.76 dB 4:23 08-Baby I’m a Star (2015 Paisley Park Remaster)
DR7 0.00 dB -9.24 dB 8:45 09-Purple Rain (2015 Paisley Park Remaster)
http://dr.loudness-war.info/album/view/131033
What you are hearing is extreme dynamic range compression – not any intangible MQA attributes.
AJ
2 Likes
So given the choice the studio still MQAs the crap brick walled version instead of the uncompressed 24/192 “crown jewel”. Not sure how MQA is exciting to some if that’s the kind of stuff they plan on releasing.
2 Likes
Completely agree. Amy Winehouse’s “Back To Black” probably qualifies as even deeper junk DR-wise. But the music is so good that any improvement is welcomed in my book. And the MQA version does sound better than the 24/96 “remaster”.
I don’t think there’s an uncompressed 24/192 version available.
Sure there is.
http://dr.loudness-war.info/album/view/138291
Though it might have got pulled from HDTracks after his passing or for some other reason.
1 Like
Circa 2011-2014, WEA pushed out for high res digital download sales some “flat transfers” of many majors in its back catalog (Fleetwood Mac, Prince, The Cars, et al.) Since then, most/all have been pulled in favor of revised and modernized remasters.
AJ
2 Likes
Ah yes, now I remember, I have the VH flat transfers from back then as well.
We can only hope that one day the studios release two versions of their material: one compressed brick-walled version for maintstream streaming and radio etc, and one high DR flat version for high fidelity playback. I think MQA was meant to be this latter option, but it seems the studios don’t have the same vision.
I heard that Adele’s latest release was compressed for CD, flat for vinyl! You would’ve thought the opposite was needed… 
Ok, that’s the one I own. After purchasing it I realized it had disappeared from HDTracks - now I know why.
That digital downloads do go “out of print” for non legal reasons is ridiculous. But the record label conglomerates seem fearful that too much choice will be confusing to consumers. And that is another reason why a single inventory MQA future is a distinct – and dangerous – possibility.
AJ
1 Like
Of course re the online arguments and free publicity. My point is I can’t even tell what the tech is.
The non Shannon sampling is applied in 24-48k bandwidth as 1x sampling(48k sample) at 8 bit. Bandwidth above that are just up-sampling.
Just tried out Tidal app, the non MQA DAC shows 88.2k not 44.1k recorded version of Bruno Mars track. Looks like software first unfold will always report 88.2/96k MQA core.
Now, can this consider up-sampling since this happens in the first ‘unfold’? The question is why it is done this way or because the MQA core must always operate at 88.2/96k?
When one convert from 96k to 48k without any form of digital filter applied in the first place, then you can take 48k and ‘unfold’ back to 96k. If digital filter is applied, that information will be lost therefore ‘unfold’ back to 96k is just up-sampling.
In the case of MQA filter is applied at 88.2/96k sampling (cut off at 44.1/48k bandwidth) so anything above that, no information can be ‘unfold’, so they revert to up-sampling instead. Nothing fancy here.