Jump to content
IGNORED

MQA is Vaporware


Recommended Posts

10 hours ago, PeterSt said:

Btw, I'm not saying that these albums now sound "better" in MQA fashion. But they are more tolerable (for me). This only counts for these too; I, II, III, IV lose with MQA all over. But the LZ's are typical "first MQA trials" as how I see it. Start out with those and you will ditch MQA right away (I virtually did that too, at the time). But then try Tusk. Or the now disappeared Neil Young's Peace Trail.

 

Most of my listening to those LZ titles are from the original Barry Diament Redbook masters (Page/Marino remasters were horrible IMHO).  I hear more detail in the hirez versions.  Presumably not because of the sample rate, but because the ADCs are better now.  The MQA versions (through a Meridian Explorer 2) produce listening fatigue for me in less than 30 minutes.

 

I purchased the double LP Tusk on the release date.  I remember it well.  Overall, I've probably listened to that title more from vinyl than from digital.  The original CD master shortened "Sara" so the whole double LP would fit on one CD.  Overall the original Tusk CD was dull and lifeless.  When Tusk first appeared on HDTracks (96/24), I was all over it.  After all the downloading was done, and I was able to import the tracks into DAW software, the peak limiting was quite evident.  And while the improved detail in the new remaster was undeniable, the peak limiting sometimes made that additional detail a bit grating.  The MQA version (on Tidal) sounds just as peak limited, but now with pseudo detail on top of the remaster's level of detail, which only enhances the listening fatigue for me.

 

Link to comment
39 minutes ago, Rt66indierock said:

I have to ask why you need to mention “digitus.” John Atkinson showed in his latest release of the Portland State University Chamber Choir (released in August )that the iTunes and Mp3 versions can sound excellent with a great performance, an excellent recording job by John and top notch mastering.

 

It is always fun when one reads your posts upside down. Heck, I was pointing out the contrary. Use MP3 to SOLVE.

 

39 minutes ago, Rt66indierock said:

Moving on to your second paragraph it is hard to make a case for a format if Led Zeppelin’s first four albums sound worse to you. I can only imagine what Led Zeppelin’s fan think and there are a lot of them.

 

You are good at it, or I am good at writing upside down. Your choice.

The first 4 are super, the two I mentioned are not. But I was talking about the MQA versions, right ? there it is the other way around.

 

Btw BD never heard a difference between the two sets. Or he wouldn't admit it. Or he felt picked upon (as often).

 

42 minutes ago, Rt66indierock said:

 

Be careful of what you wish for with the MQA is Vaporware thread. The next one will be “MQA is not Commercially Viable”

 

Oh. I guess that is for you to decide, right ?

Well it is your right anyway.

 

43 minutes ago, Rt66indierock said:

In any case it was made clear to me at RMAF 2017 I have been given complete freedom  to act in any way I see fit to stop MQA. Carte Blanche

 

Whine. You must be really proud.

 

What's up with your mood today ? You can't have it that maybe your own 10.000 threshold is passed ?

Silly.

 

45 minutes ago, Rt66indierock said:

Anybody can post they dug up some more MQA albums. I’m asking you to show us.

 

And make you even more sad ? no way. You would get angry.

Lush^3-e      Lush^2      Blaxius^2.5      Ethernet^3     HDMI^2     XLR^2

XXHighEnd (developer)

Phasure NOS1 24/768 Async USB DAC (manufacturer)

Phasure Mach III Audio PC with Linear PSU (manufacturer)

Orelino & Orelo MKII Speakers (designer/supplier)

Link to comment
28 minutes ago, Samuel T Cogley said:

The MQA versions (through a Meridian Explorer 2) produce listening fatigue for me in less than 30 minutes.

 

Agreed.

Lush^3-e      Lush^2      Blaxius^2.5      Ethernet^3     HDMI^2     XLR^2

XXHighEnd (developer)

Phasure NOS1 24/768 Async USB DAC (manufacturer)

Phasure Mach III Audio PC with Linear PSU (manufacturer)

Orelino & Orelo MKII Speakers (designer/supplier)

Link to comment
29 minutes ago, Samuel T Cogley said:

 

Most of my listening to those LZ titles are from the original Barry Diament Redbook masters (Page/Marino remasters were horrible IMHO).  I hear more detail in the hirez versions.  Presumably not because of the sample rate, but because the ADCs are better now.  The MQA versions (through a Meridian Explorer 2) produce listening fatigue for me in less than 30 minutes.

 

I purchased the double LP Tusk on the release date.  I remember it well.  Overall, I've probably listened to that title more from vinyl than from digital.  The original CD master shortened "Sara" so the whole double LP would fit on one CD.  Overall the original Tusk CD was dull and lifeless.  When Tusk first appeared on HDTracks (96/24), I was all over it.  After all the downloading was done, and I was able to import the tracks into DAW software, the peak limiting was quite evident.  And while the improved detail in the new remaster was undeniable, the peak limiting sometimes made that additional detail a bit grating.  The MQA version (on Tidal) sounds just as peak limited, but now with pseudo detail on top of the remaster's level of detail, which only enhances the listening fatigue for me.

 

 

 

THX, IRRC I am using the original CDs - I'd have to dig to find the actual CD, not the rip

Link to comment
30 minutes ago, Samuel T Cogley said:

The MQA version (on Tidal) sounds just as peak limited, but now with pseudo detail on top of the remaster's level of detail, which only enhances the listening fatigue for me.

 

Maybe not the latter. But that highs generally are "pseudo" I heartedly agree with. Otoh it is heard to see where they emerge.

 

I just finished a version of the software which can leave out the decoding. I thought to me smart and listen to "another version" which maybe is more native than the first unfold, might that do wrong. Man, that s*cks !

This in itself leads to my thinking that this unfold does more than just that. There's just no way that the e.g. 48KHz and 96KHz sound so different. I again did this with LZ (I), Talk Talk (the Colour of Spring) and Machine Head. All super bad.

Btw, the LZI should not be part of this game because this is some kind of remaster (I otherwise don't own) and it sounds heavily distorted on each track. You could say that LZ always sounds distorted, but only for IV this is really so (IMO).

 

 

Lush^3-e      Lush^2      Blaxius^2.5      Ethernet^3     HDMI^2     XLR^2

XXHighEnd (developer)

Phasure NOS1 24/768 Async USB DAC (manufacturer)

Phasure Mach III Audio PC with Linear PSU (manufacturer)

Orelino & Orelo MKII Speakers (designer/supplier)

Link to comment
43 minutes ago, Samuel T Cogley said:

The MQA version (on Tidal) sounds just as peak limited, but now with pseudo detail on top of the remaster's level of detail,

 

I am listening throughout to MQA lately (not because I like it so much but to learn from) and what often comes to my mind is that the filtering I apply is "too square" for any filtering means which is not applied at all by MQA's first unfold. This is quite contrary to what we think, but at the moment it is my perception. Read : this requires more minimum phase filtering than Redbook (which is virtually nothing). It feels like : requires more apodizing and less steep roll off. I already do sort of that, but can't do more than already done now, or I'll lose "frequency".

 

I just finished a fully played Shades of Deep Purple in MQA. I think this again a good example of an album which "works" with it. But could be personal. And possibly I am also getting too much used to actually not the best sound. :$

Lush^3-e      Lush^2      Blaxius^2.5      Ethernet^3     HDMI^2     XLR^2

XXHighEnd (developer)

Phasure NOS1 24/768 Async USB DAC (manufacturer)

Phasure Mach III Audio PC with Linear PSU (manufacturer)

Orelino & Orelo MKII Speakers (designer/supplier)

Link to comment
15 minutes ago, PeterSt said:

 

Maybe not the latter. But that highs generally are "pseudo" I heartedly agree with. Otoh it is heard to see where they emerge.

 

I just finished a version of the software which can leave out the decoding. I thought to me smart and listen to "another version" which maybe is more native than the first unfold, might that do wrong. Man, that s*cks !

This in itself leads to my thinking that this unfold does more than just that. There's just no way that the e.g. 48KHz and 96KHz sound so different. I again did this with LZ (I), Talk Talk (the Colour of Spring) and Machine Head. All super bad.

Btw, the LZI should not be part of this game because this is some kind of remaster (I otherwise don't own) and it sounds heavily distorted on each track. You could say that LZ always sounds distorted, but only for IV this is really so (IMO).

 

 

Hi PeterSt,

Does this mean that a CD encoded with MQA will sound significantly different to a non MQA encoded CD ?

(that is, they have intentionally made the CD version sound bad if not decoded by the MQA chip ?).

Thanks and regards,

Shadders.

Link to comment
1 hour ago, Rt66indierock said:

The next one will be “MQA is not Commercially Viable” do you really want that?

 

Well no, because maybe we're all fed up with MQA threads and treats and threats. But you'd have my support.

Easy to see that there's no market for it.

Lush^3-e      Lush^2      Blaxius^2.5      Ethernet^3     HDMI^2     XLR^2

XXHighEnd (developer)

Phasure NOS1 24/768 Async USB DAC (manufacturer)

Phasure Mach III Audio PC with Linear PSU (manufacturer)

Orelino & Orelo MKII Speakers (designer/supplier)

Link to comment
12 minutes ago, Shadders said:

 

Does this mean that a CD encoded with MQA will sound significantly different to a non MQA encoded CD ?

(that is, they have intentionally made the CD version sound bad if not decoded by the MQA chip ?)

 

Hi Shadders - I don't think it works like that. This will already not be so because MQA Ltd states that the MQA version in CD format (well, 24/48 most of the time) already sounds better. But I say : if you are used to a few things then this undecoded (first unfold not executed) sounds so poor that by now I can't tell what got into them.

But what we also know is that this does not make any sense anywhere because why present a "CD" which is 1/3 larger than the CD - and tell it is good to begin with ? it is not at all (I state) and the file is larger.

 

Some real time fun : I am now playing a normal Redbook. I have the sense of something lacking (the directness).

So I guess we can get used to whatever it is.

Lush^3-e      Lush^2      Blaxius^2.5      Ethernet^3     HDMI^2     XLR^2

XXHighEnd (developer)

Phasure NOS1 24/768 Async USB DAC (manufacturer)

Phasure Mach III Audio PC with Linear PSU (manufacturer)

Orelino & Orelo MKII Speakers (designer/supplier)

Link to comment
2 minutes ago, PeterSt said:

 

Hi Shadders - I don't think it works like that. This will already not be so because MQA Ltd states that the MQA version in CD format (well, 24/48 most of the time) already sounds better. But I say : if you are used to a few things then this undecoded (first unfold not executed) sounds so poor that by now I can't tell what got into them.

But what we also know is that this does not make any sense anywhere because why present a "CD" which is 1/3 larger than the CD - and tell it is good to begin with ? it is not at all (I state) and the file is larger.

 

Some real time fun : I am now playing a normal Redbook. I have the sense of something lacking (the directness).

So I guess we can get used to whatever it is.

Hi PeterSt,

Thanks. The concern is that they offer 44.1k CD's, but encoded with MQA. This will be no different to non encoded MQA CD's. I recall that the 3 LSB's are used for the embedded higher frequencies.

Whatever MQA do, is up to them, but if the record labels decide that they want to implement only MQA CD's - then they can degrade the quality purposefully. Forcing people to purchase MQA DAC's, controlling the entire audio chain.

Regards,

Shadders.

Link to comment
12 minutes ago, lucretius said:

I've never been able to find a digital version of this that doesn't suffer from a high level of tape hiss.

 

I have talked about this countless of times. I think even in the Lush thread. :o

Did you (or can you) try the MQA version ? it is the very first which is as fresh as can be. Just what all these failed version needed. Including the HDCD (especially that !) and the Hires.

So the MQA incarnation of Machine Head is by far my preferred version. Also when played undecoded it would be (I think) but that one just does not "work" (no music). If you play them subsequently (I only did Highway Star) then you just can't tell what actually happened to the / during the Unfold. But I dare bet quite some $ that this is not just unfold to "hires" (96KHz).

Lush^3-e      Lush^2      Blaxius^2.5      Ethernet^3     HDMI^2     XLR^2

XXHighEnd (developer)

Phasure NOS1 24/768 Async USB DAC (manufacturer)

Phasure Mach III Audio PC with Linear PSU (manufacturer)

Orelino & Orelo MKII Speakers (designer/supplier)

Link to comment
1 minute ago, Shadders said:

Whatever MQA do, is up to them, but if the record labels decide that they want to implement only MQA CD's - then they can degrade the quality purposefully. Forcing people to purchase MQA DAC's, controlling the entire audio chain.

 

Shadders, I tend to agree with you. But I tell you : to me this is new since tonight. I must admit that at very first I reported quite similar (I think on the Phasure forum) BUT at that time I could not decode MQA yet, so I could not relate to the decoded outcome and just thought that MQA was nothing much. This now (6 months later !) changed to : If I keep on going in improving MQA by all what I am capable of then possibly I can end up with many MQA albums which I like better than the Redbook counterparts (I really have some already), BUT without that decode stage it is rampage and sheer disaster. Remember, all is relative and in 6 months of time there's an improved SQ version of the playback software around, I play the MQA "isolated" from streaming noise and last but sure mot least : I use a newly created USB cable (out of all made for the purpose).

Lush^3-e      Lush^2      Blaxius^2.5      Ethernet^3     HDMI^2     XLR^2

XXHighEnd (developer)

Phasure NOS1 24/768 Async USB DAC (manufacturer)

Phasure Mach III Audio PC with Linear PSU (manufacturer)

Orelino & Orelo MKII Speakers (designer/supplier)

Link to comment
4 minutes ago, PeterSt said:

 

I have talked about this countless of times. I think even in the Lush thread. :o

Did you (or can you) try the MQA version ? it is the very first which is as fresh as can be. Just what all these failed version needed. Including the HDCD (especially that !) and the Hires.

So the MQA incarnation of Machine Head is by far my preferred version. Also when played undecoded it would be (I think) but that one just does not "work" (no music). If you play them subsequently (I only did Highway Star) then you just can't tell what actually happened to the / during the Unfold. But I dare bet quite some $ that this is not just unfold to "hires" (96KHz).

 

Thanks! I will try the MQA version on Tidal.

mQa is dead!

Link to comment
28 minutes ago, PeterSt said:

 

Shadders, I tend to agree with you. But I tell you : to me this is new since tonight. I must admit that at very first I reported quite similar (I think on the Phasure forum) BUT at that time I could not decode MQA yet, so I could not relate to the decoded outcome and just thought that MQA was nothing much. This now (6 months later !) changed to : If I keep on going in improving MQA by all what I am capable of then possibly I can end up with many MQA albums which I like better than the Redbook counterparts (I really have some already), BUT without that decode stage it is rampage and sheer disaster. Remember, all is relative and in 6 months of time there's an improved SQ version of the playback software around, I play the MQA "isolated" from streaming noise and last but sure mot least : I use a newly created USB cable (out of all made for the purpose).

Hi PeterSt,

I am unsure as to what you are stating.

Are you stating that you are using your own MQA decoding process, and like what you hear, which is better than redbook CD?

Or

Are you using an MQA DAC, and this presents music better than standard redbook CD ?

Or

You are listening to MQA file with no decoding, at 48k/16bits, and these files are better than redbook CD ?

Thanks and regards,

Shadders.

Link to comment
4 hours ago, sullis02 said:

An effect of the magnitude claimed by Meridian, if real,  would not need anywhere near 1500 subjects to demonstrate conclusively in a DBT, do you understand that?

 

I was quite happy with PeterSt's reply, given he has experience in design and manufacture of equipment.

 

Link to comment
9 hours ago, Shadders said:
9 hours ago, PeterSt said:

 

Shadders, I tend to agree with you. But I tell you : to me this is new since tonight. I must admit that at very first I reported quite similar (I think on the Phasure forum) BUT at that time I could not decode MQA yet, so I could not relate to the decoded outcome and just thought that MQA was nothing much. This now (6 months later !) changed to : If I keep on going in improving MQA by all what I am capable of then possibly I can end up with many MQA albums which I like better than the Redbook counterparts (I really have some already), BUT without that decode stage it is rampage and sheer disaster. Remember, all is relative and in 6 months of time there's an improved SQ version of the playback software around, I play the MQA "isolated" from streaming noise and last but sure mot least : I use a newly created USB cable (out of all made for the purpose).

Read more  

Hi PeterSt,

I am unsure as to what you are stating.

[...]

[...]

[...]

Shadders.


Hi Shadders,

 

Quote

 

Are you stating that you are using your own MQA decoding process, and like what you hear, which is better than redbook CD?

Or


 

 

Yes.

And for a certain percentage I think this sounds better. But give me another month (I am serious).

This is the general response, NOT in the context of where you quoted from.

 

Quote


Are you using an MQA DAC, and this presents music better than standard redbook CD ?

Or

 

 

No. I have (fabricated) one, but I never listened to it. This is because I think it is not the way to go, plus a few people including myself have the idea that any second unfold can not exist (and the DAC would be used for that). Further more, too many downsides (no possibility of my own filters, which would be prerequisite #1).

 

Quote

 

You are listening to MQA file with no decoding, at 48k/16bits, and these files are better than redbook CD ?

Thanks and regards,

 

 

In the context of what you quoted from, partly. So this "option" I built in the software (not really a big deal) and allows for playing back MQA albums without decoding. I had good hopes (something like maybe 17 bits instead of 16), especially thinking that now at least the masters would be better. So with the current experience I saw an "opportunity" do do even better than with a possible flawed (1st) unfolding. And there I unexpectedly was wrong. It sounds too strange and with no dynamics (the contrary of when st stage unfolded).

 

So my normal modus of playback (also see my first response) is using that decoder (in my own software) and just decode. I try with the 96KHz and with the 192KHz native sampling rate files (as so-called is buried in there).

The more I am working with it, the more I come to the conclusion that it highly depends on environmental settings and that one day I find the optimum and most MQA sound better than normal Redbook. At least that is what I strive for because I am not against better SQ.

 

Regards,

Peter

 

Lush^3-e      Lush^2      Blaxius^2.5      Ethernet^3     HDMI^2     XLR^2

XXHighEnd (developer)

Phasure NOS1 24/768 Async USB DAC (manufacturer)

Phasure Mach III Audio PC with Linear PSU (manufacturer)

Orelino & Orelo MKII Speakers (designer/supplier)

Link to comment
2 hours ago, PeterSt said:

...  It sounds too strange and with no dynamics (the contrary of when st stage unfolded). ...

 

Encoding MQA audio includes splitting it into <24 KHz and >24 KHz. The coefficients of the Quadrature Mirror filter used for this are unknown. In theory, the effects of the filter are reversed when the two signals are re-combined by the matching filter in the Core decoder. If you are not Core decding, might part of what you hear be an effect of the splitting (encoding) filter? 

"People hear what they see." - Doris Day

The forum would be a much better place if everyone were less convinced of how right they were.

Link to comment
1 hour ago, Don Hills said:

 If you are not Core decding, might part of what you hear be an effect of the splitting (encoding) filter? 

 

Yes, it is to be expected that the QMFs impact the baseband signal quality, especially because it is unlikely that these filters can be optimised both for a lossless split+join and for subjective sound quality.

Link to comment
12 minutes ago, Don Hills said:

If you are not Core decding, might part of what you hear be an effect of the splitting (encoding) filter? 

 

Thank you Don. It really puzzles me for now. This is what came to my mind yesterday :

 

Having in mind what Bob Stuart tells about what we can perceive and what not and what will be in the noise etc. etc., this may flaw in certain situations, which could be mine. Actually I dug my own pitfall at attempting this undecoded playback. Look :

 

My first thought was : okay, so maybe I have now 17 bits instead of 16 for Redbook and the remainder will be niceley dither. Mind you, I try to do a lot with dither as such, for explaining why MQA could work for me. Not the MQA encodes this dither but as long as it is uncorrelated data in there, it just is dither and a lot. Then :

 

My DAC (Phasure NOS1a/G3) resolves to 23.5 bits. Well, the analyzer can dig that out visually, to to certain extent this just is so.

Normal Redbook is upsampled to 24/705600 and theoretically 20 bits are required to do that well (2x2x2x2 upsamping, each stage requiring another new LSB). This 20 bits is resolved hands in pocket by the Phasure DAC.

 

Now on to your text, Don;

While I blindly thought of 17 bits of music data, the remainder being "dither", this is not dither at all if it concerns "folded" high(er) frequency data. Instead it will be as correlated as can be, but in a sort of strange fashion.

 

23 minutes ago, Don Hills said:

the effects of the filter are reversed when the two signals are re-combined by the matching filter in the Core decoder.

 

... Next dare to read this as "of opposite phase". And then "of some kind" but this is what I seem to hear and thus now like to read in your text. :eek: The "folding" would even literally visualize that.

 

If this is so, then Bob et al might have made the slight mistake in thinking that no DAC is able to resolve further than a famous 17 bits. This one can and now all works counterproductive (and counterfeit, haha). Well, for undecoded MQA this would be so and devastating. I'd not only hear an unintended signal, but it would even be capable of working out of phase in relation to the intended signal. And there you have what I hear : less highs, no dynamics and exactly the opposite of what MQA intends (at least on the dynamics part (impulse response).

 

Unfolding

 

Still assumed I am right to some degree with the above, it allows us to learn a few things from the unfolding process and what it actually accomplishes.

 

What the process of unfolding ... oh wait ... Androcell01.thumb.png.29eb9251c9a508db7584fb3e74122353.png can do, is this :

 

 

LZ01a.thumb.png.467ff081c630d2acb2fc492a42312fc9.png

 

This is Whole Lotta Love (LZII). Left is MQA undecoded, right is MQA decoded.

You most certainly won't hear me say that I will be able to hear what happens close to 24KHz, but the least what this will do is remove the filter noise in the 24KHz roll off area. Say that the roll off starts at 20KHz, then, well, it rolls off there, but it also implies noise. You can see this here too :

 

LZ01b.thumb.png.8e6d00acc1431f35d8065f9b99d278e1.png

 

where the right hand one is the normal CD. You can even see that in that version the noise is closer to the audible band, which is logic because this is 22.05 limited, while the left hand screenshot has 24.00 limitation (the raw MQA is 48KHz, not 44.1KHz).

On another note you can see that the band of noise is broader in the CD version and that it is also more noisy. This will be logic because it is there where the roll off area is (say) from 20KHz to 22.05KHz. So here the transition band is smaller and the roll off thus steeper (implies more noise).

 

For those who wonder :

 

LZ01b-2.thumb.png.7c4ef0b77d8bce5eb4428fbebf9834f9.png

 

The frequency itself is not extended (seen from this rough level).

 

What this urges for, is the thinking that not much needs to happen to make a CD more "Hires". But now not from the frequency standpoint but from the view on the roll off filter only.

 

What would people think about a suggestion that MQA encoding is capable of undoing the filter roll off for 22.05 limitation ?

There would not be fake upsampling in order. It would not require other masters which are sampled at a higher frequency to still bring a higher frequency (straight up to 22.05). There would NOT be needed any data space for frequency above mentioned 22.05, because there is no music data present anyway (hey, Otis Redding really never saw Barry Diament's ADC, unless re-taken from some old tape, which is not what I ever see (would be super noisy)).

 

At this moment I can't finalize this idea. But maybe others can stare at these picture and let loose the creativity beast. One last thing though :

 

I think that what I perceive from the leaky filters is IMD distortion, like a super tweeter can (or will) imply that. It is a similar rough sound. But this is a different happening from the additional frequency (I suppose I can perceive that one way or the other, though I certainly won't directly hear above 13KHz).

 

 

 

Lush^3-e      Lush^2      Blaxius^2.5      Ethernet^3     HDMI^2     XLR^2

XXHighEnd (developer)

Phasure NOS1 24/768 Async USB DAC (manufacturer)

Phasure Mach III Audio PC with Linear PSU (manufacturer)

Orelino & Orelo MKII Speakers (designer/supplier)

Link to comment
1 hour ago, PeterSt said:

Now on to your text, Don;

While I blindly thought of 17 bits of music data, the remainder being "dither", this is not dither at all if it concerns "folded" high(er) frequency data. Instead it will be as correlated as can be, but in a sort of strange fashion.

 

 

Surely the folded upper band data are at least scrambled into pseudo-noise and probably encrypted...

Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...