MQA Deep Dive - I published music on tidal to test MQA

[Zensō]

Apple,Amazon and all the rest are gonna use what the Record Labels supply them with.

Apple, Spotify, and Amazon certainly have a say in the matter.

 

[dmac6419]

Apple, Spotify, and Amazon certainly have a say in the matter.

Ok if you say so,but it's not their content.

 

[PO3c]

Ultimately, I believe this will be decided by Apple, Spotify, and perhaps Amazon. They will either accept or reject MQA files if/when they are submitted by Warner et al. It’ll be interesting to see what Spotify does when they roll out their high-res tier later this year.

Or maybe associate them as lossy with their low cost tier where they belong. Together with a WARNING: "This format has proved to struggle with simple sine waves and might produce artefacts on you hifi system. Only stream very optimed audio files that comply with MQA Ltd. Electronic music or podcast might not comply with these rules as they might suprice the MQA encoder."

 

[RichB]

They sure as heck do. How do I play DTS or Dolby stream *without* a decoder like I can with MQA? They are totally locked down formats and have mandatory decoder or you are out of luck. With MQA, I can play the baseline format with no need for any decoder.

Yes, forced royalties are not good but both DTS and Dolby added value.
Neither DTS or Dolby pretended that lossy was the true master.
Both embraces lossless and neither went beyond the terms lossy and lossless.

Yes we are burdened with licensing and royalties, but not degraded sound.
Let me know when Dolby and DTS start deblurring my audio, I suppose that will be supported as well.
It is not I who is inconsistent or pretending here.

- Rich

 

[sandymc]

They say MQA will infect 16-bit audio by omission, Tidal does not say it offers CD quality sound, it offers bitrates, Tidal does not offer Hi-Res audio it offers Master quality.
It is important to pay close attention while the goalposts are shifted, it starts with language.

We can look to Star Wars for answers.
When Luke finds out Darth Vader and Obi-Wan explains, telling Luke that his father was dead was "in a sense" true.
Here they redefine the meaning of "dead"
And of course, Darth Vader's (clearly an MQA influencer) pronouncement: Pray that I do not alter it further.

- Rich

Bob as Darth Vader? I like it

 

[AdamG]

To verify - the $23.99AUD price was the always the price for Tidal's MQA offering in AU. They've only just intoduced a "hifi" only offering for a cheaper price (they've basically gone from 2 plans, to 3). Definitely a reaction to this report, but worked out well for me - I've just downgraded to the "hifi" only version. Hopefully soon they start sending the original flacs.

Welcome Aboard @lordvader.

 

[RichB]

I already have Qobuz,Amazon HD and Tidal, $12.99 is misleading they what their money upfront $149

They are not misleading, many subscriptions provide discounts for yearly or even multi-year subscriptions.
Take out your calculator and decide what's best for you. I buy ATV+, Disney+, HBOMax with longer subscriptions to save money.
That is the actual price but it comes with a commitment.

- Rich

 

[RichB]

That's not the gist of it. OP brought up Dolby:

I explained how the analogy was wrong and that there is alternative to Dolby. Then folks tried to defend Dolby and DTS as being better than MQA and I explained the opposite is true.

This aside, as I have explained, your position makes no sense. Either you are for ALL open formats or not. You can't be selective in wanting high-res audio open, but perfectly fine with other closed audio formats. I don't know how you rationalize it in your mind, much less say I am not making sense.

The MQA position that embraces adding proprietary formats that may in fact reduce open access to 44.1/16 bit audio is inconsistent with audio science.
As a result, posts supporting it are falling outside the audible range of many rational minds

- Rich

 

[jensgk]

This aside, as I have explained, your position makes no sense.

You do not decide that.

Either you are for ALL open formats or not. You can't be selective in wanting high-res audio open, but perfectly fine with other closed audio formats. I don't know how you rationalize it in your mind, much less say I am not making sense.

I can be just as selective as I want. I can be against broccoli and for tomatoes.

Other closed audio formats does not affect me or my gear, so I can have a don't care position.

MQA does affect me, my gear and my relationship with Tidal, so therefore I fight MQA. That is very logical.

I can pick my own fights as selectively as I want.

 

[RichB]

Love you Amir, but this doesn't really make any sense. The TrueHD claim about "harshness" with 48k recording is of course ridiculous, as is the claim that an apodizing filter increases playback quality. But whole-integer upsampling, in this case simply doubling every sample to make a 48k original into a 96k file, while pointless, is still lossless. "Lossless" does not mean "the data is never altered in any way." "Lossless" means, "the original data is not altered in such a way that it cannot be fully recovered or reconstructed." By your logic here, every Delta-Sigma DAC ever made (and every oversampling R2R DAC as well) is "lossy," which is at best meaningless and at worst demonstrably false based on any meaningful sense of the concept of "lossless" in digital audio.

MQA, by contrast, takes any PCM original with a sample rate over 96k and destructively downsamples it, throwing out half (or 3/4 in the case of a 352.8k or 384k original) of the samples before encoding. And don't even bother responding that "no one needs 192k and 384k sample rates anyway" - of course they don't. But nor do they need a 96k sample rate either. That's not the point.

The point is that MQA is lossy (and not just in the way I've noted above - the encoding itself is lossy) - yet they continue to lie and claim otherwise. You know it's lossy, @John Atkinson knows it's lossy, and I would say that at this point it's not credible for anyone to try to claim that Bob Stuart doesn't know it too. Yet here is the current MQA web site on this topic:

View attachment 126838

By any reasonable interpretation, this is a lie. This is fraud.

Statements like "MQA delivers clearer sound" can be put into the bucket of misleading, evidence-free audiophile PR BS. Even "Master Quality," by the lenient legal standards established in the U.S., can be considered advertising talk and not fraud. But "Is MQA lossless? - Yes" is different. It's a lie. And "a lossless file is just a digital container... what really matters is the content!" is a willfully fraudulent statement in furtherance of that lie. The FLAC container can losslessly contain the MQA data they've stuffed in it, but the MQA data itself is not a lossless encoding of the original PCM data. You know, it, Atkinson knows it, and Bob Stuart knows it.

If you had a DAC in for review that uses a new DAC chip which claimed to perform unique forms of oversampling and ultrasonic processing, and your testing revealed that DAC actually altered the digital data so that its output did not match the input (beyond the slight amounts of noise, distortion, and jitter one gets with even the best DACs), you would go ballistic and call out the DAC and its maker for either fraud or a broken implementation. That's what MQA is: a fraudulent, broken implementation of lossless high-res. Its adoption rate and partial resemblance to other, past formats and schemes does not change that fact.

I sincerely hope you're correct that the format is going nowhere. But folks appear to have found evidence that MQA-encoded content has already infected the subscription (and perhaps download-for-purchase) music pool, with unflagged and unidentified MQA-encoded material showing up on multiple services. This is cause for concern. More broadly, as many of us have said over and over - and over - again, the lack of MQA's success thus far should not be equated with its aspirations and business model, both of which are predatory and therefore ipso facto worthy of sustained critique and opposition.

An excellent summary of the state of affairs.

- Rich

 

[mansr]

Recycling the same [MQA] marketing pamphlet again, and again, and again is hardly the way.

It has been said if a lie is repeated often enough, people will eventually believe it. This appears to be their strategy.

 

[Raindog123]

The dangerous products have come and gone and you seem to be OK with them such as Blu-ray format. Do you know a legal method to copy one? I know how to do that with MQA. Do you know how to play them without a bunch of proprietary technology with high licensing fees? I assure you MQA costs a fraction of these. So let's not be a part-time vegetarian. You have let the biggest dogs out of the gate and are worrying about the little ones...

In my eyes, the flaw of this argument is that that it is 'academic'. I do not give a rat's fart about Blu-ray. My big box of disks collects dust in the garage. But I do care about hi-fi audio streaming: Some of it is personal 'pleasure' and some psycho-socio-economical desires of 'the best', of 'respect to me as a customer', of '[pay-for-value] fairness'. All these are subjective, of course. But we do pick our battles per our personal priorities.

 

[voodooless]

But whole-integer upsampling, in this case simply doubling every sample to make a 48k original into a 96k file, while pointless, is still lossless. "Lossless" does not mean "the data is never altered in any way." "Lossless" means, "the original data is not altered in such a way that it cannot be fully recovered or reconstructed."

But the apodizing filter is not lossless. You cannot reconstruct the original half-rate signal, at least not bit-perfect. There will always be rounding losses involved.

 

[ebslo]

I think people are completely missing what MQA is doing as a technology solution. They created a way to perceptually encode ultrasonic and > 16 bit depth in music. They could have released this as a new format by itself but instead, choose to add a bonus: to encode the new information in a, in-the-clear baseband "16 bit" format. That format is designed to be "good enough." Not lossless but good enough.

It seems there is general agreement on most of this (excepting some pedantry regarding the meaning of perceptual encoding of ultrasonics). Just to clarify a point, you are not claiming the "new information" in the "16 bit" format includes "> 16 bit depth", that's only for the "24 bit" encoding, right? So the "new information" in the "16 bit" case is limited to ultrasoncis, right?. What I, and it seems others, are missing is why we would want low-level information in the audible band replaced by information in an inaudible band in a 16 bit format.

 

[RichB]

"Perceptual coding of ultrasonics" is jumping the shark btw.

It is jumping two sharks.

The only way to combat such absurd claims is with measurements, some of which we have seen here.
But these concerns are batted aside as invalid tests, while MQA actively blocks valid tests.

There is no evidence or blind tests supporting their marketing, some of which is a bold faced lie, others hyperbole.
Why are not the ones making this claims asked to prove them?
The onus of proof is on the one making the claims, not on those debunking them.

Some are accused of inconstant thinking, being argumentative, and exaggerating the threat.
Why not try a different approach, support with evidence, the claims made by MQA?

- Rich

 

[AdamG]

It’s the “Darth Pan” principle. Trust and Believe in the Force of Magic.

 

[tmtomh]

But the apodizing filter is not lossless. You cannot reconstruct the original half-rate signal, at least not bit-perfect. There will always be rounding losses involved.

Sure, but with respect that's not quite the point. Amir did not mention TrueHD's apodizing filter to make his claim that TrueHD is not lossless. He instead quoted and bolded TrueHD's 48k-96k upsampling to make his claim that TrueHD is not lossless. And for that reason his claim is not valid.

He mentioned the apodizing filter to rhetorically ask if folks believe that TrueHD's claim that the filter improves playback quality is valid. And Amir is right - that claim is not valid. But that fact means nothing with respect to whether or not MQA is lossless (it's not), or with respect to whether TrueHD is lossless (it is).

A digital reconstruction filter, apodizing or not, is applied as TrueHD implies - during playback. It is not part of the digital data. A file/encoding format like TrueHD - or MQA - that uses a particular type of digital reconstruction filter will contain a flag or instructions that will trigger the use of that filter by the playback equipment. But in such a scheme the reconstruction filter's effects are not hard-coded into the music file itself. The choice of reconstruction filter has nothing to do with whether or not the digital music file's format or content is lossless or not.

(Of course, any digital reconstruction filter applied along the way, during the production/creation of the file, will indeed have its effects permanently encoded into the file, like if there is a digital-to-analogue-to-digital step in a recording, mixing or mastering chain. But that's not the scenario being discussed here.)

 

[HoweSound]

Bob as Darth Vader? I like it

"Luke - I am your father Master..."

 

[voodooless]

Sure, but with respect that's not quite the point. Amir did not mention TrueHD's apodizing filter to make his claim that TrueHD is not lossless. He instead quoted and bolded TrueHD's 48k-96k upsampling to make his claim that TrueHD is not lossless. And for that reason his claim is not valid.

In that respect, you are quite correct.

A digital reconstruction filter, apodizing or not, is applied as TrueHD implies - during playback.

But it's not! It's done as a production step. See here and here for more info:

The upsampling takes place before Dolby TrueHD encoding

 

[dmac6419]

They are not misleading, many subscriptions provide discounts for yearly or even multi-year subscriptions.
Take out your calculator and decide what's best for you. I buy ATV+, Disney+, HBOMax with longer subscriptions to save money.
That is the actual price but it comes with a commitment.

- Rich

If you want the $12.99 price,you have to pay for the whole year in advance, Tidal, Amazon HD have discounts just the same