MQA is Vaporware

[Shadders]

5 minutes ago, Digital Assassin said:

However, this is not the story told at the beginning.

 

Here is John Atkinson quoted, with two totally untrue today fact:

 

As MQA needs to be applied at the mastering stage in a recording's production, it doesn't improve the sound quality of your existing CD collection. It is really only relevant to downloads.

 

https://www.stereophile.com/content/ive-heard-future-streaming-meridians-mqa#iVGw34mUHUGtC4bm.99

 

The story kept on changing, and keeps on changing to this day.

Hi,

MQA modifies the data - processes the data to de-blur (their inference) - so this means that for a specific recording all you need is the processing stages for the recording, to apply yourself. It has been stated that the final filter in your DAC has negligible effect on temporal blurring considering all others in the chain.

 

http://www.theabsolutesound.com/articles/beyond-high-resolution/?page=3

 

So, if the record company told us the relevant filters in the original recording chain, we could in fact implement de-blurring through a simple computer program operating on the CD data - so you can stream it once processed.

Please correct the above if incorrect.

Regards,

Shadders.

[Shadders]

4 minutes ago, soxr said:

ways to deblur pcm while not licensing anything from Bob.


 

Hi,

I am not sure you can licence/patent an inverse function, or other numerical processing, which is the reverse of a system you do not own.

Will be interesting if MQA lawyers try to challenge any reverse engineering which is just number crunching to reverse an effect from the 70's, 80's, 90's or 00's recording.

Regards,

Shadders.

[Shadders]

15 minutes ago, Fitzcaraldo215 said:

Sal - If MQA succeeds in driving all other formats out of the market, imposing its dictatorial will and allegedly inferior sound quality on all of us, leaving us with no other choices, I will have to say I was wrong.  Until then, I see no basis for your imagined fears coming to fruition.

 

And, even then, I still have my whole existing library of thousands of discs, many of which I have not even heard yet, unless the MQA police are going to come by and burn them all, sort of like Farenheit 451 did with books.

 

It may be "clearer to you every day" as you get your mind ever more riled up about it. But, I see no rational basis for your speculative conclusion.  I think the likely outcomes for MQA are either at best market niche status in (some not all) streaming and little else or a complete flop in the marketplace.

 

But, anyway, you had best stock up now on all the pre-MQA discs you can lay hands on and put them in your hidden fallout shelter.

Hi,

I think what is being inferred is that if all record companies promote MQA then :

1. Every CD we purchase in the future will be MQA encoded, and still play on CD Players.

2. MQA Ltd are stating that the 3 bits dithering is inaudible so we will have to accept MQA encoded CD's which, provide to those not interested in MQA an inferior sound.

3. The record companies may embrace the MQA marketing and only provide MQA encoded CD's, as a single pressing is cheaper than 2 different pressings.

4. The record companies buy into this as it will allow DRM to be inherent in CD's/downloads, so they can implement control at a later stage once we are accustomed to MQA only CD's etc.

Regards,

Shadders.

[Shadders]

3 minutes ago, Fitzcaraldo215 said:

Yes, thank you.  I have understood this hypothetical scenario for quite some time.  I do note that there are a number of "if ... then"s and "may"s in it, as is appropriate, since there is no hard evidence that is the game plan.  And, we could go on forever with alternative scenarios that might lead to other conclusions.  

 

I think there are many holes in the logic of this case you have presented, but many of the themes throughout this thread seem to have a common speculative, conspiracy to monopolize by fiat theme.  Namely, that a little company, probably not yet a profitable one, in conjunction with big media publishers, streamers, downloaders, etc. can simply impose their will on us - all of us - with the insidious knowledge that this product is actually inferior soundIng.  But, they like it because it does offer the possibility of a sneaky DRM, which would only be sprung on us unsuspecting idiots at a later date.  And, of course, I am not even beginning to discuss government laws and regulations to prevent precisely the sort of conspiratorial monopoly being suggested here.  

 

Meanwhile, we as consumers would not protest.  We would just blindly, like sheep, continue to buy this stuff as if our life depended on it, making the fat cats at the top of the chain, including MQA, ever richer and richer.  Meanwhile, no other suppliers would dare start an alternative, competitive venture offering a different or a traditional non-MQA format or technology, even though that would allegedly sound much better to consumers and be highly preferred.  After all, consumers  (except us real smart ones here in this thread, of course) are all so easily brainwashed.  So, MQA would prevail no matter what, and quite easily. It is a no-brainer, as we used to say.

 

Look, I have had a long business and technology career.  I have degrees in Economics and Business Administration from a top tier US school.  I have studied market dynamics and structure, including consumer markets involving technological products.  I think the theories here just do not understand how capitalism and markets work, what is feasible and what is not.  Theories here are generally wild, whipped up hysteria based on irrational fears involving a new technology that is a threat to the status quo.  Fears, including quite irrational ones, are commonplace with new technology.  I have a world of experience dealing with precisely that.

Hi,

My interpretation is that people who care about sound quality are in the exceedingly small minority, so any change to CD's would go unnoticed to the masses. It would be easy for the record companies to implement MQA on CD's, as most people would not be bothered.

People are happy with MP3, and prefer the sound to standard CD, as they being the younger generation grew up on the sound. (i have not got a reference for this, just reports in magazines - Hifi News).

If it is marketed as being better, and is a change, then people who love MP3's, might just believe it. The hifi press seems to have done.

It is not a conspiracy theory, just a concern that it will occur, and a small protest from Hifi people will not stop the implementation.

Regards,

Shadders.

[Shadders]

7 hours ago, Digital Assassin said:

MQA on CD is beyond pointless...you have to purchase new hardware to hear the so called "unfold".

 

This industry is filled with some very smart people who produce very stupid ideas.

Hi,

If MQA offers DRM, then for the future it allows more control, for the present they can offer "supposed" hi resolution etc., all in a package that can play on CD players (no MQA decoding), or on a DAC with an MQA capability. All for one single pressing.

If the record companies offer high resolution downloads (non MQA) and MQA, and CD's that are redbook as well as MQA CD's, then ok - we have a choice.

If they are convinced that MQA offers more benefits such as DRM and extra revenue - and most people cannot tell the difference between MQA on CD and normal redbook CD, then we may lose that choice.

With DVD and Blu-Ray - the copy protection has been cracked, and we can copy to hard disk for ease of viewing.

With MQA on CD, and downloads - you lose quality of recording. (CD redbook reduced from 16bit to 13bit at best).

Regards,

Shadders.

[Shadders]

12 minutes ago, PeterSt said:

While someone is in the mood anyway, here's another Peter with some thoughts :

 

I actually never ran into a hi-res album which worked for me. OK, those which do can be counted on one hand. I have close to a 2000, and I never ever play any of them. It is a waste of (quality playback) time.

Not so with MQA. And I am sorry to say it, but while MQA the most clearly sounds different, more and more it comes to my mind that it is bearable for me (and my system) because it is no hires at all. As a matter of fact, it is the only common denominator in all MQA against all Hires I have of the same albums. Well, when we like to be negative, it is.

 

As I told elsewhere, I built in a spectrograph function which can be run ad-hoc - or in batch over selected albums, and I did not find one so far which really looks like hires. Piles of noise, yes.

 

Still, I am on the verge of cracking MQA to the respect of making it sound better than normal CD (notice that I can tweak quite a lot because of having the software under my control, as well as the hardware). I guess that at this moment, when I would show it to 100 people, it would be 50-50 for what one likes better. Give it a few more months and MQA might start to win.

 

The other sad conclusion from the above and only if I am right on it, is that the change in sound can only spring from DSP. It is only that the DSP might not be in there explicitly, but be the result of the mangling to pack (encode) the format and its result at unpacking it, amongst which correlated ultrasonic frequencies and piles of noise.

I should also emphasize that I am NOT using any of the proposed filters, just because I don't use the intended hardware for it. One could also say : I go my own way with it, leave the rendering be, but have my own rendering. I even shut off any MQA D/A converter for that (rendering) matter, so the converter does not know it's MQA.

 

Is that a new approach or isn't it ... 

Peter

Hi,

With regards to the other thread on the technical aspects of MQA - the aliased upper frequency (22.05kHz to 44.1kHz), and the lossy coding, then MQA will sound different - but better ???

I had a surround decoder which had Hall effect mode, science fiction film mode, action film mode etc., which was just DSP code.

If temporal smearing was the only issue, then we do not need MQA. Just reverse the blur and provide the file as required.

What MQA is as you have indicated is just modification of the process to get the best sound, and the claim it is what they heard when they recorded the album, is a false statement.

I had thought to build a DAC with a DSP that produced even order harmonics to make the recording sound "nice".

I have Pure Audio blu ray discs (24bit/96kHz) and they sound ok - slightly different - but i have heard CD's sound just as good.

So, MQA is just like a DSP mode on your surround sound processor - and anyone can do this if they have the right software, to obtain an equivalent effect.

Regards,

Shadders.

[Shadders]

40 minutes ago, PeterSt said:

 

Haha, yes. But please mind this phrase :

 

 

Shatters, maybe I should put forward to you (because you don't know me much) that I am working for over 10 years each and every day, testing for better sound quality. Most of the working day comprises of "designing" tweaks and which comprises the full audio chain (we really build each component of it), applying the tweaks for real (build them) and lastly listen for 3-4 hours each day. MQA, at first, also was such a tweak because I believed in the set up (believed is not really past tense here). Along the way I got to learn more of it (like we all did) and instead of debunking MQA I am more of the positive kind and try to see the possible virtue of it. No matter the virtues could not be intended in the first place !

Of course all starts with not hearing any super downsides readily. Do note though that MQA sure was unlistenable for me, at first. But because I have a bit of listening experience, one can say that when I hear something good in a "tweak", no matter the bad masks it mostly, I will try to undo the bad so the good remains. I do this with speakers, with amps, with playback software, with operating systems and with DACs. And cables.

And MQA.

 

So what's wrong for me with MQA is that it shows elements which are clearly for the better and this is nothing about filtering because the filtering is not engaged. Thus do notice that exactly nobody is talking about this (at least not on this forum) and what I do, is in fact illegal (because not intended by MQA). I apply my own filters and additionally the DAC applies nothing at all (it is NOS R2R).

 

To be the most clear : I am 100% against DSP of whatever kind, because it kills the sound (but I'd allow it for very low frequencies - just saying). So we agree on that one. But the indirect or unintentional DSP applied to MQA, works out for the better - or at least it seems that it can go towards that direction.

Btw, for those listening to MQA via Tidal's Desktop Player, forget any comparison with that because it s*cks all over (for me).

Hi,

From your comments - you stated MQA was not listenable at first. So have you become accustomed to it ? Now you can determine the best parts - the tweaks that work ?

Whether it is the DSP code, FPGA code, filters, algorithm etc., MQA is supposed to deblur the signal, so we hear it how the people in the studio heard it. What MQA seems to be doing is deblur and "more", and the "more" is not proven to be of benefit (theoretically), although it sounds "nice".

So the "more" is nothing but effects, and the code can be implemented on any platform ?

Regards,

Shadders.

[Shadders]

5 minutes ago, PeterSt said:

 

I think that is a mis conception, but easy to adopt for those who hardly can know (because not knowing the internals of MQA). The deblurring is supposed to come from MQA's filters, although I don't think (mind the think) that this is stated anywhere explicitly. So if we put this the other way around : if this is not coming from MQA's filters, it is to come from a Master which lacks the ringing already. I think people stipulated that, but what I recall from it, it remained stipulation only. You could also say : if THAT could be true please !

 

So yes, MQA is supposed to deblur the signal, but this is supposed to be done in the rendering part, hence with the filters we see passing by these days. But careful, because these filters assume (this is implied) that no ringing is there in the first place, and that this thus has been removed in an earlier stage. And I don't think this is the case.

 

Looking at the MQA filters as a stand alone phenomenon, is also not allowed - I'd say. They - or whatever it is for real in real MQA renderer life - are supposed to work in conjunction with a specific DAC and - and this is MQA's story - also anticipating the ADC used. This, and now it is my story, is too much of ballony to be able to be the truth. Maybe someone thought that it could work like that, but it already can't because a multitude of ADC's can have been in order to record the (studio) album.

Now what IS told (by MQA) is that the "deblurring" (but IIRC the term is not used as such) is happening in anticipation of the ADC used, and now it is BS to apply that at the rendering stage. It should be in the Master (file we receive) already. So see, now things become vague because it is not consistent any more. Nice story, nice feature, but not consistent UNLESS it is in the master file for real. And the masters *are* different, as far as I can see. Always.

 

On a side note, a previous time I talked about this, Miska jumped in with the apodizing filter and that it exists for a long time already. This is related, but I feel to see how it is the same.

One other thing would be that the MQA guys should think that they can perform better than any self-respecting ADC. And if I would be working on such a project, I would surely think the same. Anyway notice that if that would not be the thinking, all is moot and a made up story.

 

Right, with this post I feel 10x more vague than MQA itself. I am sure you will agree.

Including inconsistencies.

But anyway all these things are on my mind, and nothing is decided for what's really happening.

Hi,

Yes - my interpretation is that the master file you receive has had the de-blurring already implemented. The filters in each device is an inverse filter (MQA requirement in code/hardware) of the DAC IC FIR filter (assumed FIR) such that the smear from that FIR filter in the DAC, is minimised.

The only way that MQA Ltd can deblur the master file is if they know the recording chain equipment ?

Or is it best guess from analysing the master file ?

Who knows. In any case - we have a process that is perhaps good engineering (removal of temporal smear and sounds "nice") and a process that makes it sound "nicer", which is perhaps superfluous engineering.

The secrecy of the entire process does not help, so we are being told it is better, and we have to use subjective experience to assess this. But, from some of the basics we do know, some aspects of the implementations (aliasing), are bad engineering - but it sounds "nice". Perhaps the bad engineering aspects are nice because as per LP/vinyl, it is noise which is preferred ?, and all MQA is this approach ?

Regards,

Shadders.

[Shadders]

23 minutes ago, PeterSt said:

 

Hi Shadders,

 

This is not so easy to explain, as it is a combination of things plus that it is (thus) combined with the playback software and its possible settings. You might not know it, but XXHighEnd can be tweaked to such an extent that even I need more than a life time to find any "best" setting, might it exist. And these settings relate to the filters used, which filters inherently don't ring, but *are* different (for the few to choose from).

 

While the above is one aspect only, I also deal with ever changing / improving DAC features. In this case this is minor because IIRC since MQA only the internal USB Isolator has been added (of course this is a great change, but it is relatively minor still).

 

What really did the job was a new USB cable, which was almost made for the purpose. Well, sort of, but I really saw it coming because I knew what I was working on. This is all related to the USB cable I was using previously, and which was too "direct" for MQA. In itself hard to explain, but think in terms of "too well following". This is no joke and I see you thinking "how can one make an USB cable doing that". Well, I did, and I did it explicitly.

So if you now envision MQA to be too much forward and also envision the reasons for it, you can work on a cable lessening that. But to be really honest, I did it out of curiousity to see what could be achieved with a USB cable if I'd make it for audio instead of "the best USB". The latter was Clairixa and the former the Lush. Either name represents its sound very well.

So if you know could hear MQA as how I perceived it myself (all hard and way too direct) then you can see how the one cable may change that character when the cable is exchanged for the other.

 

Btw, in case you seek for proof of others and MQA ... nobody has this (XXHighEnd version) yet.

 

Thanks,

Peter

Hi,

I thought that MQA Ltd were very stringent on the DAC implementation - so, the tweaks that are applied in any instance may be in breach of the MQA Ltd requirements.

The USB cable - i have not used USB audio - but the galvanic isolation has been shown to reduce jitter.

Maybe, if the USB cable has just an effect as MQA, then MQA improvements are minimal ?

Regards,

Shadders.

[Shadders]

15 minutes ago, PeterSt said:

 

Haha, Yes, if there would be a DAC implementation. But there isn't any. So his is software only, and quite explicitly so. Of course it helps if the DAC isn't again adding any filtering, and our NOS1a/G3 does not do that.

Clearly you recognized that I am only in this thread so sell DACs. 

 

In the end you will be right. This now is because the software should pass on the audio stream unaltered when an MQA DAC is the endpoint. But the software just does not support that. It always passes on the audio stream in altered fashion or otherwise my filtering does not work. And this was just a prerequisite for MQA. They knew.

 

It has been a subject before :

What's wrong with the MQA spec and software players which can decode, is that software players these days (mine ahead and this time HQPlayer as a second) want to filter in-software. If I can't do that, our DAC is of no use as it has been developed explicitly to receive filtered (reconstructed) audio data. It is NOS all right, but not allowed to operate like that.

We can say that after all facts I am forced to use software only, because software + hardware or hardware alone, is just not a workable option (and I really tried). And so I am now approaching it via a nice backdoor. And it even could work for the better.

It was about better sound, right ?

or ? ... 

Hi,

I see - you are implementing your own MQA decoding in software and presenting it using your own NOS DAC.

Te hardware with or without software can be workable - FPGA implementation of a filter - is more efficient. But then, some processor instructions may be just as efficient in some circumstances, or specific for a task such as AES encryption/decryption, video decoding functions.

I suppose as MQA is a subjective experience, then a partial, or near complete implementation is just as good.

Regards,

Shadders.

[Shadders]

28 minutes ago, mansr said:

The evidence so far suggests that there are in fact no DAC-specific filters.

Hi,

Thanks - i understood that each DAC and hence the inherent IC that has been used, requires specific implementation of software to reverse the filter of that DAC IC ?

Some of the hifi texts online indicate that the final temporal smear introduced by the DAC filter is minimal, but to obtain the pretty light on the front of the DAC stating MQA, then full implementations required.

Regards,

Shadders.

[Shadders]

2 minutes ago, mansr said:

That's what they want you to believe. Nevertheless, the impulse responses measured from the Dragonfly and Mytek Brooklyn are nearly identical to the filter coefficients extracted from Bluesound firmware. The deviation is well within expected bounds from noise and filtering associated with recording the analogue output.

Hi,

Thanks. So, this seems to be marketing claims, or was this claim made by the interview with MQA Ltd ? In any case, this is not very promising, false claims.

Regards,

Shadders.

[Shadders]

1 hour ago, mansr said:

It was also the lack of licensing restrictions making VHS the pick of the porn industry.

Hi,

I used to work in the telecoms area.

One of the senior managers wrote to all that "it is disgraceful that 50% of the then internet traffic was due to porn".

A colleague wrote "Yes, it is bad, they should bring back Sunday school".

I then responded, "Do i get a recognition award".

No one laughed. B*st*rds.

Regards,

Shadders.

[Shadders]

Hi,

Isn't the issue with regards to open, being that Sony/Philips published every detail on their standard for CD, so everyone can implement their product accordingly.

With MQA it is a secret closed standard, so no one except those who have signed the NDA can implement, and only MQA Ltd can process a sound file to encode.

Then there is the peer review aspect, where MQA is not peer reviewed, but still claims benefits against other open standards, but we can never know to confirm.

There is a saying in the security area " security through obscurity". Proponents of a system say it is secure as no one knows how it is constructed. Problem is that people will always try and reverse engineer and crack the system, and they generally succeed, or a very basic back door exists.

Regards,

Shadders.

[Shadders]

2 hours ago, soxr said:

Sampling at higher rates than 44.1K has some advantages: the AA filter can be outside the treshold of human hearing range, it can be less steep, ... and any DSP artefacts can be shifted to ultrasonic range where you can't hear them.

So basically highres is not about more music content (bit depth mainly determines noise floor), but about avoiding audible artefacts in the baseband as you have more places to hide mistakes into parts of the data that can't be heard.



 

Hi,

Just to check on this, are we saying that current recording/mastering is at such a high sample rate that the filters used do NOT have temporal smear ?

Can it be confirmed that the temporal smear is in fact "group delay" of the filter - and the filters used for recording/mastering have an all pass filter response for those frequencies that the ear is interested in ?

As such, MQA can really only "tart up" previous recordings and at the same time implement aliasing which is detrimental ???.

Regards,

Shadders.

[Shadders]

52 minutes ago, Fokus said:

 

No. Typical AA and AI filters are linear phase, thus with a constant group delay.

 

'time smear' is the 'problem' that is suggested when you look at the system response to an  unit impulse.

Hi,

I was of the understanding that an all pass filter had a constant time delay across the frequency band of interest, hence no smearing/delay/phase change in the frequency band of interest.

I would have thought that linear phase, which introduces a change in phase (albeit linearly) with regards to frequency, means, as an example, that at 100Hz there is no phase change (no delay), and at 10kHz there is a 45deg phase change, hence a delay when compared to 100Hz.

Is it that linear phase filters provide minimal temporal smear, but it still exists ?

Regards,

Shadders.

[Shadders]

19 minutes ago, Jud said:

 

In linear phase filters, time through the filter does not vary by frequency.  In minimum phase filters, group delay is minimized but varies by frequency.

Hi,

Thanks - but if the phase changes with frequency, and we have two signals in synchronisation, 100Hz and 10kHz, the 100Hz has no delay (0deg phase) but the 10kHz signal has 45deg phase (assume lag) and this introduces a 12.5uS delay ?

Regards,

Shadders.

[Shadders]

23 minutes ago, Fokus said:

 

Linear phase = constant (time) delay over the entire frequency band.

 

If you have 45 degrees at 10kHz, then you'll find 0.45 degrees at 100Hz. No phase distortion. No temporal distortion.

 

It are the minimum phase filters so beloved by Meridian/MQA that cause phase distortion. Not that this matters much.

 

 

Hi,

Thanks- have responded to Jud on this.

Regards,

Shadders.

[Shadders]

5 minutes ago, Jud said:

 

I don’t know what the specific delays are, but yes, a minimum phase filter delays one frequency more than another, so it is a “dispersive” filter.  This is thought by some people to possibly give an illusion of depth.

Hi,

Are you sure this is correct. If the filter is an all pass, then all frequencies have the same delay (minimal group group delay) in the frequency range of interest.

So a minimal phase filter may delay one frequency more than another (but minimally), but a linear phase filter incurs greater phase changes (assumed) albeit linearly with frequency.

Is a minimal phase filter called as such, since its phase change is less than other filters such as linear phase ?.

A phase change will equal time delay for that signal. They represent the same thing ?

Regards,

Shadders.

[Shadders]

4 minutes ago, Jud said:

 

Yep.

Hi,

Thanks - just looked up the wiki - states linear phase has constant group delay - so i need to check this out.

Regards,

Shadders.

[Shadders]

5 minutes ago, mansr said:

An all-pass filter is any filter with flat amplitude response across all frequencies. The phase response can be anything.

Hi,

Thanks - yes - just looked up the book i have - constant time delay, but varying phase delay. The group delay graphs for the various orders are constant. I need to examine more closely.

Regards,

Shadders.

[Shadders]

Hi,

Does anyone know what anti aliasing filter was used for CD encoding in the 1980's ?.

Was it Bessel, Butterworth, or other, and what was the order and cut off frequency etc ?.

Thanks and regards,

Shadders.

[Shadders]

9 hours ago, Don Hills said:

 

"Other", usually 8th to 12th order elliptical. Typically 35 to 50 dB down at Nyquist. Several companies produced improved performance filters to retrofit to early ADCs, such as the Apogee filters for the Sony PCM-1600 and F1 range.

Hi Don,

Thanks - much appreciated - will try and simulate the filter.

Regards,

Shadders.

[Shadders]

On 17/07/2017 at 1:34 PM, Shadders said:

Hi,

Thanks - yes - just looked up the book i have - constant time delay, but varying phase delay. The group delay graphs for the various orders are constant. I need to examine more closely.

Regards,

Shadders.

Hi,

I have worked out my mistake, group delay is the differential of phase change across the frequency band. So constant group delay is linear phase change.

The time differences between a 100Hz and 20kHz signal for a 4th order bessel filter with cut off frequency of 22.05kHz is approximately 3.5uS - so not very much.

I have yet to simulate an 8th order elliptical filter - so i assume that this is what MQA is supposed to address - the temporal smearing of the signal due to the phase changes across the frequency band ?

Given the higher sample rates used today - i assume that temporal smearing is not an issue so we don't need MQA ?

Regards,

Shadders.

[Shadders]

8 hours ago, Jud said:

 

An experiment:  Set up a free test installation of Audirvana Plus.  Set upsampling preferences as you like, but leave the “pre-ringing” setting at 1.0 (linear phase).  Listen to whatever you like.  Then adjust pre-ringing to 0.0 (minimum phase) and see if you hear a difference.  (You may wish to upsample to the maximum input your DAC will accept, so as much of the upsampling as possible is performed by iZotope 64-bit SRC bundled with Audirvana Plus.)

Hi,

Thanks. Seems to be a Mac only program. Also, do not have a DAC, but do have a surround sound processor - so testing is limited.

Regards,

Shadders.