advertisement


MQA part the 3rd - t't't'timing...

DimitryZ said:
I have a dozen HDCDs and they sound great through the Musical Fidelity decoder.
Click to expand...

Yes, much the same here via the open source decoder or the player I have that can deal with HDCD. Alas, some of those discs sound clearly degraded when played on a *non* HDCD decoding player (or software). Because of obvious peak compression.

The key point being that they are NOT "compatable" in terms of continuing to sound as good as a plain CD version could on most players. Thus they play games with the term "compatable" to simply mean "it will play and you'll hear some music on plain players".

I'm quite sure the inventor did NOT intend this. None of the HDCDs from him that I have show the problem. But some others are clearly just using it as a gimmick, without a clue. SOP for many in the biz, alas. Just as with clipping, gross compression, etc.

Most crazy of all was perhaps the Beach Boys DVD-A that had HDCD encoded audio when it could easily have had high res audio *because that's what DVD-A was for*.

These pages illustrate some of what I found about HDCD some years ago.
http://www.audiomisc.co.uk/HFN/HDCD/Enigma.html
http://www.audiomisc.co.uk/HFN/HDCD/Examined.html
 
clownfish said:
Thanks, understood.

I guess it’s unlikely given they are producing about a dozen formats for free, but I agree that’s hardly enough to base a study on.

They may be more likely to share if it’s the case than others, especially as they don’t appear to have a dog in the fight.
Click to expand...

Part of the 'provenance' problem is that what they sell you as a 'non MQA' version may have been generated *from* an MQA version. Not what went *into* MQA encoding. So if MQA *did* degrade the sound, you'd hear that from *both* and have no way to tell this was the cause simply by a listening comparison.

Decades of experience with analysing the output of the music biz made me , long ago, to conclude that you need to be vary wary at times about what they do. The pages on my website document example after example of the ways they have messed up what gets sold. Many releases are great, but many are not and show obvious flaws.

Sunlight is the best disinfectant.
 
TBH I tend to feel that plain boring old Information Theory implies the same lesson for MQA as it does from HDCD.

That if Audio CD is a 'tight box' for good audio that any attempt to hijack some of the bitrate for non-lpcm purposes changes the size of a tight box in the hope stealing from Peter to pay Paul won't get noticed. The result is either a tiny-to-zero improvement for those using the magic decoder ring, or causing a degrading of lpcm replay.

Given the magic of the "Music must be 1/f" presumption gives some space. But not all music is the same. And - as per HDCD - not all producers of music have a clue, either.

So, experience leads to "The more toys you give them, the more ways they find to mess up the results." That may not be the intent of the inventors, but it will happen. They just wiggle sliders. No idea what happens inside the box.
 
MQA may be harmless because it may sound almost indistinguishable from FLAC.

Still, the Norwegian Hifi-journalists from Stereo+ detected very small differences that favoured LPCM Flac - less listenening fatigue and less blur (deblurring, ahem).However, they also said that these differences may only be apparent on highly resolving stereo systems. They also said that MQA may suit bright-sounding speakers. So perhaps MQA is 'armless, as Jim puts it. But why pay a premium for mostly harmless MQA?

EDIT
To be fair, some people like the MQA-versions better tnan LPCM Flac - that is fair enough - and I have no qualms about letting them have their MQA pudding if I can have my LPCM pudding - by that I mean a pudding not derived from MQA.
EDIT

Imagine a situation where MQA files are the predominant file format. I buy the most recent album by Bruce Springsteen and go home to play it on my old DAC. But it sounds worse than I am used to. I paid for an album that is supposedly better than LPCM and hi res FLAC but on my DAC it sounds more like MP3. In order to get the true MQA quality that I paid for, I have to go buy a new DAC at extra cost. In reality I pay extra twice for what I can get today at a quality that has already been certified by the record company, warts and all. I see no need for a third party to try and save my music for me by first tampering with it - so it loses its lossless status - and then by flavouring it an extra time to make it sound the way MQA and their algorithm think all music should sound.
 
Rosewind said:
To be fair, some people like the MQA-versions better tnan LPCM Flac - that is fair enough - and I have no qualms about letting them have their MQA pudding if I can have my LPCM pudding - by that I mean a pudding not derived from MQA.
Click to expand...

Yes, but... What - for example - if they actually prefer the MQA simply because of the alteration in the combined results of altered ADC and DAC filter dispersions? The point then being that they could make such a change for themselves and applying it to plain LPCM without needing MQA to encode and decode.

According to the patents, HDCD claimed to give effectively a 'better' response by autoswitching reconstruction filter. To give better 'time resolution' for transients, but then use a more conventional filer for the rest of the music to avoid the (short) transient filter producing excess distortion.

However so far as we can tell this wasn't ever done. The decoders in the HDCD DACs never included it, and I've never found an HDCD with the code for it.

Yet some back in the day praised the improved transient response. :)

So a problem here is that MQA is a 'bundle' whose details are hidden. As a result someone may well like/prefer it, but assume the wrong reason. That clouds being able to tell if something else would have delivered for that listener without any need for MQA or the added 'noise hill' etc. As things stand, we can't tell.
 
Comparison of two editions of the same track on Tidal:
(https://www.faktiskt.io/phpBB3/viewtopic.php?f=9&t=64960&start=540 page 19)
file.php


and:
file.php
 
The look different. Inc some 'digital black' on one version. But what versions are they? My Icelandic (?) isn't up to making any sense of the conversations.
 
EDITED:

They SEEM TO BE different masterings.

The Swedish text preceeding what Zombie copied reads,

"MQA må vara bluff och båg, men det är ju inte konstigt att MQA uppfattas låta bättre. Uppenbarligen olika mastringar hos Tidal:"
=
MQA may be fraud and deception, but it really isn't strange that MQA seems to sound better. Apparantly different masterings at Tidal:"

....
I have no way to tell whether this statement is correct.
 
Last edited:
The obvious difference is that one version is clipped and has some digital black. Gain too high if nothing else. However I'm not sure if they show the same section or the entire track for the stats.
 
I amended the post above as I can't answer your questions. So we are back to "provenance" and which (parts of) tracks are shown in post #106 above.
 
Last edited:
I asked the poster, who posted three years ago, and he answered that it is a part of the song and that he is absolutely sure that both samples are exactly the same part of the song.
 
Rosewind said:
EDITED:

They SEEM TO BE different masterings.

The Swedish text preceeding what Zombie copied reads,

"MQA må vara bluff och båg, men det är ju inte konstigt att MQA uppfattas låta bättre. Uppenbarligen olika mastringar hos Tidal:"
=
MQA must be fraud and deception, but it really isn't strange that MQA seems to sound better. Apparantly different masterings at Tidal:"

....
I have no way to tell whether this statement is correct.
Click to expand...
I find this to be a strange conclusion.

A system that makes better sounding masters available to audiophiles (as the writer seems to mean), is nonetheless labeled as "Fraud and Deception."

:)
 
Zombie said:
I asked the poster, who posted three years ago, and he answered that it is a part of the song and that he is absolutely sure that both samples are exactly the same part of the song.
Click to expand...

Curious, then, that the 'loudest part' displays report different times.

And digital black is something I tend to expect at a leadin/leadout, not in the bulk of the track. Mind you that track is also clipped, so has clearly been messed about.

It is possible to interpolate to reduce clipping effects if you scale down. But I've not seen any arguments that MQA would do that. I'd have expected MQA encoding a clipped track would be asking for trouble as the clipping usually generates spurious HF from the 'flats' it simulates.

So the same 'song' yes. I have many versions of some 'songs', but have them because they differ.

So that doesn't ensure the 'non MQA' version *is* what was presented for MQA encoding. On the evidence shown, I doubt it was.

In this case the MQA version might sound better simply because it wasn't clipped!
 
DimitryZ said:
I find this to be a strange conclusion.

A system that makes better sounding masters available to audiophiles (as the writer seems to mean), is nonetheless labeled as "Fraud and Deception."

:)
Click to expand...

If anything we might argue that 'deception' has been practiced by whoever supplied the clipped non-MQA version as the clipping isn't likely to be what hit the mics! Its a generally result of the LOUDNESS obsession of some in the music biz. Wind up to '11'. 8-]
 
Does that mean that as part of the MQA process that they’ve actually sourced a better version of the track.
God help that MQA might have actually improved the final listening experience for a user….
 
Rosewind said:
The Swedish text preceeding what Zombie copied reads,

"MQA må vara bluff och båg, men det är ju inte konstigt att MQA uppfattas låta bättre. Uppenbarligen olika mastringar hos Tidal:"
=
MQA must be fraud and deception, but it really isn't strange that MQA seems to sound better. Apparantly different masterings at Tidal:"
Click to expand...
Slight correction: "MQA may be fraud..."
 
Mr_Sukebe said:
Does that mean that as part of the MQA process that they’ve actually sourced a better version of the track.
God help that MQA might have actually improved the final listening experience for a user….
Click to expand...

Well, if those generating the MQA *had* "sourced a better version" then it prompts the questions:

1) Is that version available to us so we can bypass the added cost of MQA and any risk of MQA being poorer than that input? if so, how do we determine which version is is, so can choose it? (And use it to assess MQA.)

2) if it isn't - how did the people generating the MQA version get it, and why would the music company agree to let them have a profit from when the music company could gain a profit without such a 'cut' by offerring it direct?

It also simply puts another facet on: You can't assess MQA fully *unless you also have access to what went into MQA*. Otherwise you may be assessing some other cause of a 'difference' that has zip to do with MQA. And for all we know, what was fed *into* the MQA encoder might have been preferred by listeners to both of those 'compared'.

Bottom line: No provenance => useless outcome that can mislead.

If 'God' can help us with this, then he can speak up... Maybe he'll issue an MQA version. 8-]
 
You must log in or register to reply here.


advertisement


Back
Top