MQA is Vaporware

[mansr]

6 minutes ago, smartin said:

Yes, MQA is mathematically lossy, but why does that matter at all in and of itself.  There is nothing inherently wrong with that.  The only aspect of 'lossy' that should matter (IMHO, and perhaps I am missing something) is how it effects sound quality.

The problem with perceptual codecs such as MQA is that the audible effects of the losses are unpredictable in the general case. For this reason, codec development includes extensive listening tests across a wide selection of test cases. Nevertheless, a real-world input may trigger an undesired behaviour of the encoder resulting in unacceptable degradation. It's impossible to avoid with certainty.

 

Bandwidth limitation, on the other hand, is comparatively easy to evaluate. It is also something that is always required since no storage medium has unlimited capacity. For every system constructed, the chosen cut-off is a compromise between fidelity, electrical/mechanical constraints, and price. The bandwidth of analogue tape is directly proportional to the speed of the tape. 30 ips reel-to-reel tape has very good bandwidth but is quite bulky and expensive. The compact cassette was designed with an emphasis on size and cost, and consequently the bandwidth suffers. The same principles are at work for digital systems. When the CD was developed, the main goal was to be somewhat better than vinyl at a comparable price level, something the final product delivered. As technology progressed, higher data rates became feasible, and this was taken advantage of to produce the SACD and DVD-A formats. Throughout the history of audio reproduction, the fidelity has largely been dictated by the technology available at reasonable prices.

 

The perceptual codecs were invented in order to maximise the sound quality under tight data rate constraints. Without such techniques, 128 kbps (stereo) would only be of telephone quality, and most people would agree that, for all its flaws, mp3 sounds considerably better than that. As network speeds have increased, so has the need for compression decreased. We can now stream audio at CD quality and above without resorting to perceptual coding tricks, and storage has become so cheap as to be effectively unlimited. Indeed, the fidelity delivered by MQA can be achieved without the use of lossy coding. It is thus strange that a newly introduced format should should incorporate such methods.

 

From a perceptual coding perspective, MQA is confounding in other ways. Controlled listening tests generally agree that frequencies above 20 kHz are inaudible to humans. For a perceptual codec to nevertheless make an attempt at preserving them is thus somewhat bizarre. The MQA process as a whole is claimed to sound better than non-MQA hi-res audio. However, no studies have been published regarding the audibility of the compression step in isolation. Even an mp3 of a great mastering can sound superior to an uncompressed poor one.

[Ralf11]

6 hours ago, The Computer Audiophile said:

tons of speculation and fear mongering without many facts 

 

I blame html for that - things were fine before '94

 

Now, where's my cane?

[The Computer Audiophile]

56 minutes ago, Sal1950 said:

First MP3, next MQA

 

Gizmodo reported yesterday that Fraunhofer Institute for Integrated Circuits IIS, developer of the MP3 audio data-compression system, has terminated the licensing program that allows companies to create MP3 encoders and decoders. This is probably not sad news for many audiophiles, who disdain the format’s lower audio quality compared with uncompressed CD quality. But there’s no doubt that MP3 fundamentally changed the face of music distribution, which makes it important in the history of our hobby.

 

Everything is temporary, if you give it enough time :~)

 

 

 

 

1 hour ago, crenca said:

 

No, no "come on" here

 

Really, that is the thing about the ground you walk on - it effects everything because well, it is the ground you walk on.  Your question is too broad for a comment box - which part of everyting do you want to discuss - and why because it is already discussed endlessly here on almost every thread?

 

Behind the context of a subjectivist "why does it matter - I only care what it sounds like and if I can't here it then it does not matter" is an entire engineering world where math matters...

 

 

 

I think you and I could have a great time arguing about this over a couple beverages some day. Not that we don't have better things to do with our time, but it would be much better to have a discourse in person. 

 

On one hand you're arguing the consumer perspective and how this affects the consumer. Now you're using the engineering argument to combat a consumer concern. Hmmmm.

 

A consumer doesn't care if a recording has 140 bD of dynamic range versus 141 dB of dynamic range. I'm sure engineers care and the math matters, but it's irrelevant to most people. I care about what something sounds like. I want music to sound as close to the artist's intent as possible. Sure, the theoretical argument that 141 dB of dynamic range is better and should get me closer to the artist's intent is true, but it's dumb in real life. 

 

Sample rate conversion is lossy and so is MQA. Standard FLAC albums are the result of a lossy process. MQA albums are the result of a lossy process. Thus, I don't think using the lossy argument is really sound in the case against MQA.

 

To me the best case against MQA is possibility that undecoded MQA sounds worse than redbook and the fact that many devices won't support MQA decoding. That's only the best case against it, not the only case. 

 

I really wish I could setup a skype video call with Bob Stuart and people like mansr and miska, so we could attempt to get to the bottom of some of these claims. Without knowledgable people representing both sides, it's impossible to get anywhere. 

 

 

 

 

 

 

[daverich4]

6 hours ago, The Computer Audiophile said:

Thus the reason I asked for people to boil down exactly what they don't like.

 

Well, my reason is a little different than most of the responses I've seen here. Two reasons actually.

 

1. No fault of MQA but I have no interest in streaming. I don't listen to music that way. I might listen to one of my rigs for 3 or 4 hours one day and then not do it again for a couple of days. I probably buy 8 or 10 albums a month so Tidal would certainly be cheaper but I like having my own music and to counter Witch Doctor, no I'm not angry about spending all that money.

 

2. I don't think I hear a difference between Redbook and Hi Rez. I'm saying I don't, not you or anyone else. I will say however, that if something I want to own is available in 24/96, that's the way I buy it. Just because I guess. I have albums all the way up to DSD mostly out of curiosity but I personally can't tell the difference. Certainly, I have any number of Redbook recordings that sound as good or better to me than any of the higher resolution music that I have.

 

As far as MQA goes then, with no interest in streaming and the personal inability to distinguish higher resolutions, not for me.

 

My 2¢

[mansr]

45 minutes ago, The Computer Audiophile said:

Sample rate conversion is lossy and so is MQA. Standard FLAC albums are the result of a lossy process. MQA albums are the result of a lossy process. Thus, I don't think using the lossy argument is really sound in the case against MQA.

As I have tried repeatedly to explain, the losses from sample rate reduction (bandwidth limiting) are of a fundamentally different nature than those from a perceptual coding algorithm. They cannot be compared. Besides, labels that care about sound quality usually offer the actual studio masters, whatever resolution they happen to be. Do you believe transferring files over the internet degrades sound quality? If not, how can you be saying that FLACs of the studio masters are lossy?

[The Computer Audiophile]

5 minutes ago, mansr said:

As I have tried repeatedly to explain, the losses from sample rate reduction (bandwidth limiting) are of a fundamentally different nature than those from a perceptual coding algorithm. They cannot be compared. Besides, labels that care about sound quality usually offer the actual studio masters, whatever resolution they happen to be. Do you believe transferring files over the internet degrades sound quality? If not, how can you be saying that FLACs of the studio masters are lossy?

 

FLAC of studio masters aren't lossy. Of course I don't believe the internet degrades audio. 

[mansr]

32 minutes ago, The Computer Audiophile said:

FLAC of studio masters aren't lossy. Of course I don't believe the internet degrades audio. 

At least we agree on something. Two things, even.

[realhifi]

5 hours ago, crenca said:

 

 

No doubt true - but there customers will continue reading...  (even if just a handful).

 

 Customers?  Or do you mean "non" customers? ?

[realhifi]

4 hours ago, The Computer Audiophile said:

 

To the consumer it doesn't matter how something is lost. 

 

Depends on the "consumer" and if they even care if something is "lost" as much as "consumers" here do. 

[Don Hills]

3 hours ago, Sal1950 said:

First MP3, next MQA

 

Gizmodo reported yesterday that Fraunhofer Institute for Integrated Circuits IIS, developer of the MP3 audio data-compression system, has terminated the licensing program that allows companies to create MP3 encoders and decoders. ...

 

 

This is another example of misdirection, akin to MQA's marketing. What has really happened is that the patents for MP3 have expired, so you no longer need to purchase a license. Instead, Fraunhofer would like you to purchase a license for AAC, for which the patents are still current.  So far from being "the death of MP3", it heralds a new beginning...

[Jud]

1 minute ago, Don Hills said:

 

 

This is another example of misdirection, akin to MQA's marketing. What has really happened is that the patents for MP3 have expired, so you no longer need to purchase a license. Instead, Fraunhofer would like you to purchase a license for AAC, for which the patents are still current.  So far from being "the death of MP3", it heralds a new beginning...

 

No no, it is death!! like MQA!!! Ggaahhh!!!

[Jud]

1 hour ago, mansr said:

As I have tried repeatedly to explain, the losses from sample rate reduction (bandwidth limiting) are of a fundamentally different nature than those from a perceptual coding algorithm. They cannot be compared.

 

True, there's probably less research on the audible effects of sample rate reduction.

[mansr]

2 minutes ago, Jud said:

True, there's probably less research on the audible effects of sample rate reduction.

The research says as long as frequencies up to 20 kHz are intact, the change is imperceptible. But you already knew that.

[jhwalker]

2 hours ago, mansr said:

As I have tried repeatedly to explain, the losses from sample rate reduction (bandwidth limiting) are of a fundamentally different nature than those from a perceptual coding algorithm. They cannot be compared.

 

Of course!  One discards inaudible content in order to fit into a given size, and the other discards inaudible content in order to fit . . . oh, wait . . .

 

Kidding, but I don't see how they are "of a fundamentally different nature" - sample rate reduction cuts off all frequencies above a certain point (and requires filtering / resampling to do so), and perceptual coding is just somewhat more selective

[Jud]

2 minutes ago, mansr said:

The research says as long as frequencies up to 20 kHz are intact, the change is imperceptible. But you already knew that.

 

And here I thought there was research on time smearing, linear phase vs. minimum phase, apodizing filters, the effect of ultrasonics, etc.  No?

[Sal1950]

3 hours ago, Jud said:

 

No no, it is death!! like MQA!!! Ggaahhh!!!

From your lips to Gods ears.

[fung0]

12 hours ago, The Computer Audiophile said:

Everything is lossy compared the original performance. What matters is how much loss people find acceptable.

 

The only relevant definition of "lossy" in the current discussion is: that loss in fidelity relative to the working studio master which is inflicted purely for the purposes of reducing data size of the recording delivered to the consumer.

 

Why is this particular type of 'loss' so significant?

 

Ever since Thomas Edison, audiophiles have been demanding access to recordings of higher and higher fidelity. Music labels have delivered technological improvements in numerous lucrative stages - LP, tape, cassette, CD, SACD, FLAC, high-res FLAC...But we are now reaching a limit. There is very little further improvement that music publishers can offer. The next logical step, already underway, is delivering to consumers the actual studio master copy. That is, allowing them to purchase the highest-quality digital representation that exists of the original analog (real-world) performance.

 

Music publishers would rather cut off an appendage than do this. Not just because it means they'll never be able to sell us the same content yet another time. But, more fundamentally, because it means relinquishing their control, their position of superiority. Once consumers have the masters, the publishers are no longer the custodians of the one true Holy Grail.

 

Hence their eager adoption of MQA. In the nick of time, someone has developed a format that lets publishers proclaim that they're releasing something that "Sounds Just as Good!!" as the original studio master - but which is definitively not the original studio master. The distinction is very real, even if the human ear is incapable of distinguishing the 'lossy' MQA file from the original master. It's the only conceivable reason for MQA to be a 'lossy' format, and not simply an enhancement to 'lossless' PCM. The master remains the master, and the MQA file is emphatically not a bit-for-bit copy of it.

 

Once you see this logic, you can't help but realize that MQA is being embraced not as a parallel delivery format to high-resolution PCM, but as a replacement for it. An entirely needless replacement. There are no technical impediments any more. So why would a true audiophile choose a somewhat more-compact recording over one that is absolutely unequivocally guaranteed to contain every possible bit of the original studio recording? The only plausible reason for the adoption of MQA is to avoid giving us that choice.

 

People keep saying "Don't worry, our high-res PCM-FLAC files won't go away." But they will. That's the whole point.

[r0dd3r5]

28 minutes ago, fung0 said:

 

The only relevant definition of "lossy" in the current discussion is: that loss in fidelity relative to the working studio master which is inflicted purely for the purposes of reducing data size of the recording delivered to the consumer.

 

Why is this particular type of 'loss' so significant?

 

Ever since Thomas Edison, audiophiles have been demanding access to recordings of higher and higher fidelity. Music labels have delivered technological improvements in numerous lucrative stages - LP, tape, cassette, CD, SACD, FLAC, high-res FLAC...But we are now reaching a limit. There is very little further improvement that music publishers can offer. The next logical step, already underway, is delivering to consumers the actual studio master copy. That is, allowing them to purchase the highest-quality digital representation that exists of the original analog (real-world) performance.

 

Music publishers would rather cut off an appendage than do this. Not just because it means they'll never be able to sell us the same content yet another time. But, more fundamentally, because it means relinquishing their control, their position of superiority. Once consumers have the masters, the publishers are no longer the custodians of the one true Holy Grail.

 

Hence their eager adoption of MQA. In the nick of time, someone has developed a format that lets publishers proclaim that they're releasing something that "Sounds Just as Good!!" as the original studio master - but which is definitively not the original studio master. The distinction is very real, even if the human ear is incapable of distinguishing the 'lossy' MQA file from the original master. It's the only conceivable reason for MQA to be a 'lossy' format, and not simply an enhancement to 'lossless' PCM. The master remains the master, and the MQA file is emphatically not a bit-for-bit copy of it.

 

Once you see this logic, you can't help but realize that MQA is being embraced not as a parallel delivery format to high-resolution PCM, but as a replacement for it. An entirely needless replacement. There are no technical impediments any more. So why would a true audiophile choose a somewhat more-compact recording over one that is absolutely unequivocally guaranteed to contain every possible bit of the original studio recording? The only plausible reason for the adoption of MQA is to avoid giving us that choice.

 

People keep saying "Don't worry, our high-res PCM-FLAC files won't go away." But they will. That's the whole point.

+1

[mansr]

10 hours ago, Jud said:

And here I thought there was research on time smearing, linear phase vs. minimum phase, apodizing filters, the effect of ultrasonics, etc.  No?

Is there more than anecdotal evidence that any of those things are perceptible?

[adamdea]

We already have a very good, open source, widely distributed and well supported  system for producing perceptually indistinguishable, reduced size version of studio masters. It's called 16/44. It is marginally preferable to 256 Khz AAC as it can be further transcoded transparently. 

 

If people want something better than 2 channel 16/44 they would be better off working on some version of surround/HRTF or whatever.

[Sal1950]

7 hours ago, fung0 said:

So why would a true audiophile choose a somewhat more-compact recording over one that is absolutely unequivocally guaranteed to contain every possible bit of the original studio recording? The only plausible reason for the adoption of MQA is to avoid giving us that choice.

Your entire post is very well worded and right on the money!

Thank You

[rickca]

8 hours ago, fung0 said:

People keep saying "Don't worry, our high-res PCM-FLAC files won't go away." But they will. That's the whole point.

I completely agree.  The labels like MQA because MQA distribution files are readily identifiable as such.  To combat piracy, the labels will eventually withdraw authorization to distribute non-MQA files, whether it is streaming or downloads.  Then they will issue takedown requests for anything out there that isn't MQA.

 

The campaign claiming that MQA sounds just as good if not better is just propaganda to counter any resistance.  MQA has enlisted audio journalists to enthusiastically embrace MQA as a technical wonder.  If they don't, they risk losing their industry insider status ... no more nondisclosure agreements, no more software betas, no more invitations to launch events ... your career is over.

 

Of course, there's no proof that this is where we're headed, so the journalists just dismiss this as nonsense.  But it makes perfect sense.  The stakes are too high, there's so much money to be made.

[Jud]

6 hours ago, mansr said:

Is there more than anecdotal evidence that any of those things are perceptible?

 

I believe there is actual research, yes.  But life being busy at the moment, perhaps we could reach a resolution faster this way:

 

@Miska has said he can perceive differences in the sounds of his filters, which vary from each other in the sorts of characteristics I mentioned.  Do you believe him?  If so, these things are indeed perceptible.  Accuracy of those perceptions will then be a matter of training.

[The Computer Audiophile]

I don't disagree labels like MQA, but I can't figure out why they all wouldn't be onboard if it's as cut and dry as everyone thinks. 

[mansr]

Just now, The Computer Audiophile said:

I don't disagree labels like MQA, but I can't figure out why they all wouldn't be onboard if it's as cut and dry as everyone thinks. 

Maybe some of them see the downsides. Linn certainly seems to.