MQA is Vaporware

[asdf1000]

4 minutes ago, Sonicularity said:

 

I don't think @crenca will allow Roon to get away with any bias toward MQA when both Tidal and Qobuz subscriptions are enabled.  Roon has already admitted to a problem that they are working on to resolve that favors MQA in some instances where there is no reason to consider that format to be superior in quality.

 

I shared a public quote by Roon's COO about MQA:

 

"if MQA dominates the world of hi-res, then the fear is spot-on. A world in which all content higher resolution than 44.1/16 is encoded in MQA is bad for the future of audio. No company should have power like that."

 

https://community.roonlabs.com/t/mqa-and-digital-rights-management/63220/3

[Sonicularity]

My position has always been that hi-res adds no improvement to 44.1/16 sound quality for practically every situation with regards to source or playback equipment with exceedingly few exceptions, and none that should be significant.  However, with MQA, or some similar proprietary format with complete control, a considerable difference can be artificially implemented to control both quality and cost.  That is why I care.  

[asdf1000]

1 hour ago, Rt66indierock said:

And there isn't a market for streaming CD quality either.  

 

Disclaimer: I would love MQA to go away.

 

I believe from reports (might not be accurate) there are ~ 3million Tidal subscribers and ~10 million Deezer subcribers.

 

Out of 13 million there, would it be outrageous to guess 5 million may be paying for CD quality streaming?

 

Not suggesting 5 million would move across to Spotify, but if Spotify could get 5 million of their 100 million subscribers to pay double for CD quality (or better), it wouldn't interest them financially?

 

Is that what you're saying?

 

Note: Spotify were testing CD quality streaming less than 24 months ago....

 

https://www.theverge.com/2017/4/5/15168340/lossless-audio-music-compression-test-spotify-hi-fi-tidal

[sandyk]

1 hour ago, fas42 said:

IME, this is not the approach that gives the results I'm after -

 Frank

 Your requirements appear to be quite different to that of most members.

 

Alex

[firedog]

2 minutes ago, Em2016 said:

 

Disclaimer: I would love MQA to go away.

 

I believe from reports (might not be accurate) there are ~ 3million Tidal subscribers and ~10 million Deezer subcribers.

 

Out of 13 million there, would it be outrageous to guess 5 million may be paying for CD quality streaming?

 

Not suggesting 5 million would move across to Spotify, but if Spotify could get 5 million of their 100 million subscribers to pay double for CD quality (or better), it wouldn't interest them financially?

 

Is that what you're saying?

 

From reports, the actual number of subscribers paying for the CD quality streams is nowhere near that, and is a tiny percentage of overall subscribers. 

[sandyk]

6 minutes ago, Sonicularity said:

My position has always been that hi-res adds no improvement to 44.1/16 sound quality for practically every situation with regards to source or playback equipment with exceedingly few exceptions, and none that should be significant. 

 

  Your position is markedly different to that of the  large numbers of members who love their high res, including DSD .

[asdf1000]

3 minutes ago, firedog said:

From reports, the actual number of subscribers paying for the CD quality streams is nowhere near that, and is a tiny percentage of overall subscribers. 

 

Cool. Would love to see these numbers. Does @Rt66indierock know?

 

Not guesses (like I've done) but actual figure of total Tidal + Deezer users paying for CD quality (of total ~13 million Tidal + Deezer users)

[Sonicularity]

3 minutes ago, sandyk said:

 

  Your position is markedly different to that of the  large numbers of members who love their high res, including DSD .

 

Absolutely.  An exceedingly few numbers of exceptions.  I did mention that.

[fas42]

3 minutes ago, sandyk said:

 Frank

 Your requirements appear to be quite different to that of most members.

 

Alex

 

Alex, my requirements are that I hear the recording, and only the recording - there are those who use selected recordings to show off the 'qualities' of their playback rigs, but that's something that doesn't interest me, at all .

 

Connecting emotionally with the musical event captured is paramount, and if a setup can't do that, with "ordinary" recordings, then it's failed, for me. Adding distortion is not necessary to achieve this, but scrupulously worked on audible integrity is.

 

My attitude is that the recording is king, and the rig has to be fit, to be worthy of presenting it - this helps enormously in setting priorities, and making the right moves ... IME .

 

 

[Rt66indierock]

47 minutes ago, Em2016 said:

 

Disclaimer: I would love MQA to go away.

 

I believe from reports (might not be accurate) there are ~ 3million Tidal subscribers and ~10 million Deezer subcribers.

 

Out of 13 million there, would it be outrageous to guess 5 million may be paying for CD quality streaming?

 

Not suggesting 5 million would move across to Spotify, but if Spotify could get 5 million of their 100 million subscribers to pay double for CD quality (or better), it wouldn't interest them financially?

 

Is that what you're saying?

 

Note: Spotify were testing CD quality streaming less than 24 months ago....

 

https://www.theverge.com/2017/4/5/15168340/lossless-audio-music-compression-test-spotify-hi-fi-tidal

 

Yes it would be outrageous to guess 5 million are paying for CD quality streaming. Deezer's revenue is near $400 million dollars for 2018 with 7 million paid subscribers. Those numbers would indicate there are very few people paying for CD quality.

I got a Universal Royalty statement of Tidal's for 2017. The CD quality tier subscriber number was less than 200,000. And you have overstated Tidal's subscriber numbers based on their last financial statements.

 

Spotify has done enough market testing to know their subscriber base will not pay more for CD or higher quality. 

 

What I'm saying is nobody cares. If I'm wrong Marc Finer of the DEG among others would throw subscriber in numbers in my face. They haven't.

 

[asdf1000]

16 minutes ago, Rt66indierock said:

 

Yes it would be outrageous to guess 5 million are paying for CD quality streaming. Deezer's revenue is near $400 million dollars for 2018 with 7 million paid subscribers. Those numbers would indicate there are very few people paying for CD quality.

I got a Universal Royalty statement of Tidal's for 2017. The CD quality tier subscriber number was less than 200,000. And you have overstated Tidal's subscriber numbers based on their last financial statements.

 

Spotify has done enough market testing to know their subscriber base will not pay more for CD or higher quality. 

 

What I'm saying is nobody cares. If I'm wrong Marc Finer of the DEG among others would throw subscriber in numbers in my face. They haven't.

 

 

All noted. You speak with the same certainty as Rob Darling in the podcast.

 

We shall see which of you is correct and which has it all wrong...

 

Disclaimer: I hope MQA goes away.

[sandyk]

56 minutes ago, fas42 said:

Alex, my requirements are that I hear the recording, and only the recording - there are those who use selected recordings to show off the 'qualities' of their playback rigs, but that's something that doesn't interest me, at all .

 Frank

 Perhaps this thread isn't the appropriate place to go there again ?

 

Alex

[Rt66indierock]

6 minutes ago, Em2016 said:

 

All noted. You speak with the same certainty as Rob Darling in the podcast.

 

We shall see which of you is correct and which has it all wrong...

 

Disclaimer: I hope MQA goes away.

 

I speak as someone who has seen the revenue numbers and can compute average revenue per paid subscriber. Rob Darling is someone with something to sell. Roon isn't a value proposition with Mp3s.

 

For my analysis to be wrong average revenue per paid subscriber would have to grow at rate people would be talking about on the cable business channels. 

 

 

[Ran]

2 hours ago, Em2016 said:

@The Computer Audiophile an interesting chat here with Roon's Rob Darling...

 

Listen from 45mins... he says all the streaming services will have a high quality streaming tier and "it will end up being up being MQA..."

 

https://www.audiostream.com/content/rob-darling-roon-labs-audiostream-podcast-no6

 

 

 

Closed, proprietary, end to end, ecosystem, control, restrictions, etc.. I guess MQA model works for Roon as well. 

[Paul R]

5 minutes ago, Rt66indierock said:

 

I speak as someone who has seen the revenue numbers and can compute average revenue per paid subscriber. Rob Darling is someone with something to sell. Roon isn't a value proposition with Mp3s.

 

For my analysis to be wrong average revenue per paid subscriber would have to grow at rate people would be talking about on the cable business channels. 

 

 

 

So what do those numbers break down to in terms of Audiophiles, who are interested in higher quality, and just plain everyday music lovers, who are more interested in content than in quality?

[Rt66indierock]

12 minutes ago, Paul R said:

 

So what do those numbers break down to in terms of Audiophiles, who are interested in higher quality, and just plain everyday music lovers, who are more interested in content than in quality?

 

Take Deezer, according the CEO they are near $400 in revenue and he said they have 7 million paid subscribers. That is about $57 per paid subscriber. The student price is $60, the premium price is $100 to $120,  hi-fi is $240 and family with up to six accounts is $180.

 

My solid number on Tidal from had the Hi-Fi tier at about 171 thousand in 2017.  Some analysis I did when Deezer's 2018 numbers came out puts a best case for the HiFi tier at maybe twice that.  Most likely number for Deezer's Hi-Fi tier is closer to 200,000.

[Jud]

12 hours ago, Lee Scoggins said:

 

Actually I do.  I have been building predictive models since 1996 and I have won two competitions for machine learning applications, one with gated neural networks and one with cluster weighted modeling from MIT.  It's literally my job but now I have the luxury of leading data engineers and data scientists and finding more problems to solve.

 

Did any of your machine learning applications come up with bad digital audio filters?

 

Or, working in reverse, would you care to evaluate the success or failure of a years-long machine learning effort that resulted in a couple of dozen or fewer digital audio filters that demonstrably and measurably do not work as advertised?

[Jud]

6 hours ago, Paul R said:

It’s opinion. Just exactly as truthful as any other opinion. 

 

I do not agree with a lot of it, and I think some of it is just wrong - but so what? 

 

Sorry to say, but it does go beyond opinion. We see MQA marketing about "deblurring" repeated (notice no one in audio outside of MQA even uses the word) when it's known that MQA is incapable of doing that.

 

We have the response curves for all the MQA filters, and not a single one cuts at ringing frequencies. Period.

[Jud]

5 hours ago, fas42 said:

 

This is the bit that gets me. In the article, https://www.stereophile.com/content/mqa-questions-and-answers-tutorial-temporal-errors-audio, they say

 

 

Ummm, if you want gibberish, this to me is a perfect example, . Sciencey talk which contains nothing. I want to see a piece of real music recording, before and after MQA is applied, which clearly shows how the signal is altered - by getting a diff of the result - and an explanation for why the MQA variety is "more accurate" to what the microphones heard. Talking about theoretical, impulse responses does not compute - what is the ear hearing that makes the difference; with a genuine music signal?

 

Dispersion is related to group delay and is different than ringing.  Filters that are minimum phase are dispersive, meaning time to pass through the filter varies by frequency.  All you need to avoid dispersion is a linear phase filter, and there are plenty of those in the world.  (There may be other reasons for choosing a minimum phase filter instead.)

[Jud]

4 hours ago, Rt66indierock said:

 

Not me I told representatives from Warner, Universal, Sony and the DEG that you will find high resolution audio is a very hard sell.  Talk all you want the only people interested in high resolution audio are people selling it and the audio press.

 

A market for high resolution streaming does not exist today. And there isn't a market for streaming CD quality either.  

 

If they're looking for numbers, sure. If they're looking for a very little extra money from marketing to a niche, a parallel to what the hi res download services provide, then I can see it continuing.

[KeenObserver]

I think that is the bottom line.  People on AS and similar sights are looking for quality music.  And we think of ourselves as a large group.  However,  we are a small percentage of the music consuming public.  The majority of the music consuming public have no idea.  Someone comes up with something like MQA and they go: "Ya, OK".  They don't even realize that they are being screwed.  And Lee Scoggins appears to know what he is talking about, to them.  Sad.

[crenca]

On 3/1/2019 at 1:16 AM, Norton said:

 

In the UK we have an expression “Can’t see the wood* for the trees”.  Always amazes me that some of the people  who get het up about the potential of DRM in MQA seem to be enthusiastic,  unquestioning users of streaming services, where the issue is not so much the management of rights,  but whether the user has any rights to anything specific in the first place.  

 

AFAIK  no streaming service guarantees its paid-up users continued, let alone perpetual, access to any specific album or track; which (along with SQ) is why I’d only ever use a streaming service to supplement, rather than replace, my own library.

 

*forest.

 

I know you were responding to a post about access of albums/tracks in streaming vs. ownership.  I just want to point out there are other kinds of rights besides ownership in a digital ecosystem.  MQA's IP lock up of PCM sent to your DAC implies management of different digital rights than what you guys are talking about here.  Open formats do not imply unlimited rights/ownership, but they do imply other rights.  For example, the end users choice of hardware/software, decoding, DSP, filtering, etc.  One way of putting it is that they allow the end user freedom from an "end to end" prison of application.  There is a reason Bob S calls an non-MQA controlled dac a "legacy dac"...

[KeenObserver]

The majors were happy to take our money when they sold us millions and millions of CD's, that were advertised as "perfect sound forever".  However, the concept of "perfect sound forever" is anathema to the majors.  They always want us to believe that there is something better out there.

MQA is crippled for a reason.  In the future you can always hold out the carrot and say: " we have something better".

[KeenObserver]

The saddest part is that there are people who have faith in publications like TAS and Stereophile who are being led down the primrose path.

Like sheep to the slaughter.  To a world where MQA controls the entire music chain.

Sad, indeed.

[crenca]

On 2/28/2019 at 11:49 PM, Fokus said:

 

You are still clueless as to the deblurring thing.

 

You are clueless as to how to identify ADCs in given recordings.

 

You are clueless as to machine learning.

 

And you are clearly clueless as to how much time was invested to dream up MQA.

 

 

Lee's technobabble sales jargon use of the term "machine learning" reminds me, in an ancillary way, of Searle/chinese room argument: