MQA is Vaporware

[barrows]

Does this change any views?

 

http://www.computeraudiophile.com/f8-general-forum/big-mqa-news-really-tidal-and-software-decoding-etc-31151/

 

Yes, it scares me that MQA might actually catch on. I am a firm believer that MQA is a solution for a nonexistent problem: I do not need to compress my music for any reason, storage is very affordable, and I prefer to purchase the music I listen to rather than stream. Looking at the potential for additional artifacts from the MQA process, i would rather not have those in my playback. Better sound quality? Is The Emperor naked?

Perhaps it is time to buy more music before it is too late...

[barrows]

I object to the term "master" in this context entirely. MQA encoding is not a bit perfect copy of the master, and using that terminology is very misleading.

 

This reminds me of the lie: "perfect sound forever"

[barrows]

MQA is Da Bomb! I just listened to Empire State of Mind in MQA through the Tidal desktop app (software decoding) and the initial bass lines that open the song damn near reached out of my desktop speakers and slapped me!

 

Whoever says this is vaporware needs to check this out.

 

 

Whoever thinks desktop speakers are a reliable way to judge sound quality is delusional.

[barrows]

it's the DRM issue that people are worried about

 

While DRM is worrisome, I am ore concerned that MQA adds alias artifacts to the file which are not present in the music. One might like the sound of these additional artifacts, but they are artifacts and will obscure musical details.

[barrows]

Introducing is hook marketing word. If there is no additional cost then Tidal is absorbing the costs associated MQA or MQA is not getting any royalties. You tell me which one is happening. MQA can license hardware and software collect royalties on unit sales. And they can license content providers and collect royalties. Can you see any other ways to bring in revenue?

 

As it is DRM, perhaps content providers are paying something for it.

[barrows]

A proper listening test for MQA would A/B a single track where the track is the same source for both versions (same master, same original sample rate). Make sure it is a track which demonstrates lots of low level details (hall ambience will be especially effective, also soundstage depth) as these are what will be lost in the artifacts created in MQA. Also, listen really closely for differences in timbre, especially on bowed strings and brass instruments. Do a level test first as well with an SPL meter, any sound pressure differences will invalidate any results.

 

I am well aware of the historic significance of JBL loudspeakers, indeed, many of my musician friends still own the JBL 3-way studio monitors and I have heard them quite often. A near filed listening environment is unsuitable for testing things like ambience retrieval and soundstage depth. I have also heard JBLs, high end horn speakers, while they do offer excellent dynamics, they are very weak in portraying instrumental timbre accurately. You are aware that the company which makes JBLs (Harmon Group) also makes a line of higher performance speakers under the Revel name, right?

[barrows]

Yeah I know about revel but everytime I play my JBL's I can't seem to want to change anything, just happy with what I am hearing. If I do upgrade it will be the new JBL 4312E just announced at CES.

 

Cool, that is all that matters!

[barrows]

"alias artifacts" ??

 

If you are interested in what is happening with MQA's compression methods, go back in this thread and look for the link (I think miska posted it) of the article which describes in detail what happens. The MQA process produces alias products (artifacts).

[barrows]

I wouldn't call it aliasing - so not sure what you mean.

 

Did you read this?

 

MQA 192k / 96k There and back again.

 

I look forward to reading Miska's (Jussi) technical evaluation when ready.

[barrows]

I think we need to remember that (in my understanding) there are two aspects to MQA processing:

 

1. The compression algorithm

 

2. The ADC/DAC compensation

 

All we are getting with the Tidal files is the compression, not the compensation. The compensation has to be applied in the hardware (DAC) because the compensation applied would have to be specific to each DAC design. Additionally, I believe (but am not sure on this) the ADC compensation has to occur in hardware where the analog source is first digitized by the ADC, either in initial recording or subsequent mastering from analog to digital. The improved sound quality claims appear to mainly stem from the compensation techniques, all the compression does is allow for a hi res file to be delivered in a smaller container (of course to the average consumer comparing this to an MP3 it might deliver improved sound quality, but to audiophiles who are already listening to an uncompressed hi res file this is a moot point).

Additionally, the MQA compression technique adds alias products, said to be inaudible of course, just as the psyychoacoustically derived MP3 algorithm is also claimed to be inaudible...

[barrows]

No I myself cannot, since I'm not technically savvy enough to be able to point out the correctness or incorrectness of his exact technical reasoning.

Having said that: can you yourself indicate if John Siau is technically more correct or less correct than Stereophile is in their analyses of MQA's working?

 

Furthermore: please allow me to be much more sceptical towards the VP of a DAC brand that's clearly not willing to embrace MQA for their own personal (and no doubt business/financial) reasons. For this exact reason I tend to trust Stereophile's opinion more, as they are not paid purely to represent a specific brand's own financial interest. Don't you think John (although being a digital expert) just might be a little bit biased on this topic..?

 

To me John's opinion on MQA is just as much worth as Bob's. Both are likely to be taken with a nice grain of salt.

 

Sent from my HTC One_M8 using Computer Audiophile mobile app

 

Personally, I am more likely to trust Siau, as he has not really steered me wrong. When Bob Stuart uses the term: "Master Quality Authenticated" I immediately know there is some bad smelling stuff going on, as MQA clearly bears very little resemblance to the actual master. I can purchase, for example, 24/176.4 HRx files which are exact bit perfect copies of the master from Reference Recordings, MQA is an entirely different thing.

[barrows]

Fair point. However I do trust Stereophile more than either Bob or John, for the reasons I have explained.

 

Do check out the Stereophile article on actual MQA bit depths and you will understand why.

 

Sent from my HTC One_M8 using Computer Audiophile mobile app

 

I have read everything at Stereophile and find that most of it appears to be a regurgitation of what Bob Stuart/Meridian are saying, which I find to be obfuscating marketing speak for the most part.

I do look forward to a real technical analysis of what MQA compression and compensation actually does by John Atkinson with actual measurements showing the spectrum analysis, jitter, alias products, etc. That would be interesting. Hopefully Jussi will be reporting on this in the near future, I know he has a vested interest with HQPlayer and a very high degree of understanding of digital filtering (which is what MQA does after all).

[barrows]

My understanding is that MQA really reduces 24bit resolution to 17bit, that makes the 1st block on Siau's diagram. When anyone speaks about higher 'apparent' resolution, it may be caused by some DSP, but such a 'resolution' cannot return information which was lost by reducing resolution from 24 to 17 bits. It's similar like when you reduce resolution of some picture and then you perform some sharpening algorithm on it. You may improve something for your eyes, but the low level detail is not reconstructed.

 

And there is the rub. I agree. On the one hand, for the average consumer, MQA can be, and may be a step up in sound quality, as they (may) get a version of hi res (sort of) when they are used to MP3 or (maybe at best) CD level quality. But for audiophiles it is a different thing, we are already listening to hi res (often) and are using DACs which are painstakingly designed specifically to retrieve the last bit of low level detail. Suddenly all the work of our favorite DAC developers on low level detail, and digital filter design is not really relevant anymore with MQA, and we will be settling for whatever MQA feels is appropriate.

[barrows]

Some healthy skepticism is always OK. Please note however that a pretty extensive technical analysis has already been done by Jon Iverson in the same article on several tracks, of which some were recorded by John Atkinson himself. Although I think you would like to see these analyses to be even more elaborated?

 

Sent from my HTC One_M8 using Computer Audiophile mobile app

 

I will go back and re-visit then, it has been awhile.

[barrows]

What is wrong about his statement? The MQA file is compressed in a 24/48 container, but contains (almost) all the data necessary to reconstruct the original 24/192 (or 24/384 or higher) file once it passes through the USB interface.

 

Yes, it's not 100% lossless, but it's pretty darned close (i.e., lossless to 24kHz, very nearly lossless to 48kHz, lossy but containing quite a bit of the original content up to 96kHz, etc.). The part that's missing is clearly inaudible (i.e., a lot of garbage below the noise floor and some of the content above 24kHz).

 

I understand skepticism, but there's quite a bit of rancor going around which is (in my opinion) overblown and misplaced.

 

Problem: The "inaudible" artifacts of which you speak have the probability to inter modulate with analog audio signal, causing audible artifacts. Digital designers do care what happens above beyond 24 kHz, for these distortions can, and often do, have an effect on audible frequencies.

[barrows]

Do writers and reviewers, in the audio industry, act under a code of ethics?.

 

At Stereophile writers are required to follow a set of ethical rules. Since Valin is still employed by TAS, it appears they enforce no such rules.

[barrows]

My reason to criticize MQA is because I am afraid that it may catch on. As long as other sources of hi resolution music files are available, I care not a lick about MQA, but if even a single recording which I want in a hi res version is only available in MQA, then I will be very, very upset. I think this outcome is very possible, so I will continue to try and point out the compromises which MQA makes.

[barrows]

What do you mean exactly?

 

Sent from my HTC One_M8 using Computer Audiophile mobile app

 

I am referring to some not so ethical stuff which was reported years ago. It does not need to be re-hashed here, i am sure if you are interested some Internet searching will find you the details.

[barrows]

While I do believe the above "fire wall" is exists in practice, and even that the editors/reviewers themselves depend on it, it the case of MQA it has been irrelevant. You ask for "proof" of their anti-consumer, pro-industry bias. The only "proof" I or any other consumer has is in their behavior and writings. They simply do not care to look at MQA from any other angle excepting the industries larger business problems (see Robert Harley's "Master Quality Authenticated (MQA): The View From 30,000 Feet") or from an insular, "sound quality is the ONLY criteria to judge anything" angle perhaps best expressed by John Atkinson himself:

 

"In almost 40 years of attending audio press events, only rarely have I come away feeling that I was present at the birth of a new world."

 

When I discussed this issue with John Atkinson himself on the Stereophile blog, he completely dismissed the idea that MQA (or something like it) has any serious ramifications for the consumer at all except sound quality.

 

Heck, I would give them their due if they had been anywhere correct about the SQ aspect but here in the real world it turns out that MQA is not the "birth of a new world" SQ wise. It is at best a modest (even important) tweak in some cases, irrelevant or worse for other recordings.

 

You can go on about "proof" (which is largly irrelavant/not valid when it comes to human motivation) all you want but the proof is in da puddin...

 

I would differ slightly in POV: The masses listen almost entirely to MP3 coded music, if MQA catches on for the masses, and replaces MP3, then that will likely be a wholesale step up in sound quality for the masses, and a game changer in that regard.

 

The problem I see, is that for audiophiles, we may lose access to straight, uncompressed non-DRMed high res files.

[barrows]

I agree with you on this, but as stated in a previous post I think it's highly unlikely that any music company will choose MQA as their sole format. Business-wise that would be a very stupid decision.

 

Sent from my HTC One_M8 using Computer Audiophile mobile app

 

Maybe... But consider this scenario: Record company guys are deciding whether to release a popular title as hi res (say DSD) for download, instead they release it as MQA. Right there, that is a huge loss for me if it is music I want to have in its best form. I see this decision as being quite likely, if it is not already being made.

[barrows]

I am loving my VPI Classic 4 turntable more and more. All this fear and loathing is just not worth it.

 

As for benefits of High Rez recordings/MQA? Only a small number of all recordings meet audiophile standards.

 

Only a very small number of recent recordings meet audiophile standards.

 

Therefore, why would anyone pay for a high rez download or stream of a poor recording?

 

Would not a CD or LP of a quality recording would be a reasonable alternative to a monthly charge for dreck?

 

Because sampling any recording at a higher rate makes it sound better. The digital filters applied can be much more benign with even 24/88.2 and 24/96, and this leads to much less digital artifacts (and hence less "digital" sounding in the old school use of that term). I am not advocating for MQA/streaming, but for real hi res when available.

[barrows]

I see a lot of posts about bitrates. Remember the old saying, garbage in garbage out. The algorithm used to convert analog to digital bits is 40 years old. Putting lipstick on a pig by increasing bit rates does not change that its a pig. There is a device called aftermasterpro that claims to change the algorithm used by "remastering" the signal and then sending it to your device:

 

A/B some tracks and decide for yourself:

 

 

 

Listen | AfterMaster | Audio Labs

 

What? The algorithm used is to convert to digital is constantly being changed, many ADCs use something very different from each other, not sure what you are talking about.