advertisement


MQA fracas at RMAF 2018

Status
Not open for further replies.
Jim Audiomisc said:
I'm not really the person for that in this case because my concerns and arguments are about the engineering and science of what their patents say they do. Their *adverts* are something else if they make claims of a more sweeping kind. And I'm not trying to 'stop' MQA because my concerns are:

1) To ensure people are well informed about the technical side. So they can make informed choices.

2) Know there are alternatives which may well make more sense or be preferred

3) We keep *choice*. My only worry here is the risk we may get into an 'MQA only' or 'MQA tax' situation in the future. But we aren't there at present.

Objecting to the ASA may well be a route others think useful to take. Personally, I think they are chocolate teapots, and not equipped to assess any 'technical' argument. You only have to look at other old rulings WRT other issues to see this. But if others want to raise objections with the ASA I'm happy if things I've said help.

That said, I don't actually recall seeing any detaied adverts for MQA. Just patents, 'papers'. etc. i.e. not paid-for advertising in UK magazines or newspapers that makes claims I could say were 'false'.
Click to expand...

I wrote to the ASA once. Me and about 120 other people.

The complaint was about a Euro football campaign run by Heineken that had some bloke crossing the world just in time to sit down in Wembley Stadium for the start of the match, he then pulls out a bottle of beer and I presume his team wins. So he wasn't English. Anyway, it is against the law to drink beer in a football stadium, had he done so he would have been arrested.

Heineken claimed that it was a generic stadium, but that didn't work as the competition was the Euros in the UK and the law applies throughout. They lost, were fined and the ads were banned several months after the competition had ended.

Most advertising goes through copyright and legal verification. MQA would be stupid not to, but you never know ...
 
Tin said:
Well, the thing is that I haven't seen a local commercial for MQA, so there is nothing to complain about.
Click to expand...
then google "MQA" and see if they are giving out the same lies where you are. If they are, go to your authorities.
 
Jim Audiomisc said:
I agree with the first comment. I also still hold Bob Stuart and Peter Craven in high regard as they've done a lot of good work and produced some excellent products. (I still love the 500 series DAC for example, and regret it won't do high-rate/res.)

The problem is that a good 'track record' doesn't ensure you can't make mistakes or go off in the wrong direction if you get too excited by your own ideas.

The 'ringing' problem for playback essentially vanishes once you use higher sampling rates. Because;

a) there is 3/5ths of SFA up in the region where any 'ringing' is likely to arise with high rate material.

b) you probably can't hear it anyway - either because of yer ears, or the speakers, etc, you use.

And as has been pointed out, real music isn't a series of isolated impulse functions which are a mathematical fiction anyway in the real world.

For me, the real advantage of high sample rate is to shove these problems well above where you'd hear them. And my personal view is that 96k is fine and gives bags of 'elbow room'. Then add noise shaping if you want smaller files, etc. Jobs a good'un.
Click to expand...
I took the first comment from one of your previous posts, as I'm more of a digital age person, which gives me enough confidence about the other points I made.
DRM is old fashioned. Young people don't mind paying for content, they just want easy and convenient access, which internet provides.

As most music isn't available in HiRes anyway, for the majority of people the real choice is between 16/44.1 FLAC, 320K OGG Vorbis or MQA. There will be only a very small group of people who can stream 320K, but cannot stream 16/44.1 FLAC at approx 800K. The number of people who consider that to be a problem will be even smaller. And as MQA has DRM that takes up precious real estate, it isn't even capable of sitting comfortably in the middle.

Anyway, people with internet speeds below 320K are out of luck no matter what, and the amount of people with access of 100Mbit, or more, is increasing every day.
 
sls4321 said:
I wrote to the ASA once. Me and about 120 other people.

The complaint was about a Euro football campaign run by Heineken that had some bloke crossing the world just in time to sit down in Wembley Stadium for the start of the match, he then pulls out a bottle of beer and I presume his team wins. So he wasn't English. Anyway, it is against the law to drink beer in a football stadium, had he done so he would have been arrested.

Heineken claimed that it was a generic stadium, but that didn't work as the competition was the Euros in the UK and the law applies throughout. They lost, were fined and the ads were banned several months after the competition had ended.

Most advertising goes through copyright and legal verification. MQA would be stupid not to, but you never know ...
Click to expand...

MQA are certainly lying in their claim that "MQA reveals every detail of the original recording." But they seem to have lost all contact with the concept of "legal, decent honest and truthful". I think that they are now just "desperate".
 
Tin said:
I'm not a native speaker, but 'Cracked the audiophile code' and 'as good as reel' nor 'as good as life' didn't particularly strike me as being 'much negative observation'. I also wasn't too sure about your admiration of the infamous 3.7 compression.
But I'm sure I interpreted all that incorrectly.
Live and learn, right?

I would also like to dispute your latest addition as MQA being 'clearly better than CD-quality PCM'. I'm sure it sounds comparitively better on your DAC where PCM is being crippled by the MQA filters being used all the time.
For the fast majority of tracks, you know, the unauthenticated ones that are being automatically transformed to MQA, the artifacts and worse S/N will decrease the quality arguably below CD quality.
But you knew this already, because we explained this a couple of times before.



Lastly, I'm going to take away your giddyness about the fantastic 3.7 compression factor of MQA.
You can download Lame (*), a nice Open Source program, or use any alternative you like for this test.
Grab one of your 24/192 WAV files, and scale it back to 20/96 FLAC.
Now compare the file sizes. You'll notice that MQA didn't do anything remotely spectacular there either.
So why aren't there 20/96 FLAC files you ask? Well, the 'solving a non existent problem' is the obvious answer. Having said that, 20/96 wouldn't be lossy, blurred and have a worse S/N compared to CD. So while as pointless as MQA, it still is superior.
Bummer, right?

*) I can't wait for your comment on the name.
Click to expand...
You actually have paid remarkably little attention to my posts.

I performed a set of listening tests, where i compared several permutations of MQA listening configuration and found it was severely lacking as a universal delivery codec with non-MQA enabled DACs.

The compression factor is from a comprehensive anti-MQA white paper referenced multiple times by your side. They reported 3.7 for MQA, 1.7 for entropy tricked out FLAC. The description of "better than CD streaming" is also from the same paper.

If you dont acte pay attention to the conversation, you cant contribute to the debate. Please up your game.
 
Nic Robinson said:
[Moderating]
This thread is generating a number of complaints about personal insults from one individual. Any more and that individual will find himself outside of this thread. Final warning (and a couple of posts tidied up).
Click to expand...
I have never even attempted to complain to you over multiple and vicous insults against me essentially for the duration of this thread.
 
Tin said:
Impressive, who helped you with this? Is this the same Dimitry?
I'm also very curious on why this is relevant.
MQA adds -unnecesary- ringing artifacts, while solving a non-existent real world problem.
Click to expand...
My MIT education and 30+ years of real world engineering experience "helped" my write this.
 
Jim Audiomisc said:
Fact: I haven't said MQA is a "scam".

Fact: Other methods can compress the data to the same extent *without* changing the musical content or needing any change to the equipment required by the end-user.
Click to expand...
I apologize for ascribing this assessment to you, it is a result of trying to have multiple conversations.

Part of your current statement is contradictory, by definition. Compression always changes musical content, the only question is how and to what extent.

I have also found in my listening tests, that MQA absolutely requires full decode for good performance. Today this means buting a new DAC. However, many audiophiles dont mind.
 
I
Shadders said:
This makes no sense. Why would you create a CD with impulses to see if there is ringing ? No one listens to impulse responses. An impulse response is a signal used only in the design of a filter, and not in its use. It does not occur on CD's or in nature. It is a "false" signal in music. It just does not occur.

Again, there is no pre-ringing in CD's. A CD is a band limited signal. The mastering process will not produce CD's with ringing. If ringing was there, you would see it in the spectral plot - as per Hifi News who provide spectral plots of downloads that they review.


The ADC's of today have such a high sample rate (at least 192kHz output sample rate) that there is no energy at or near the nyquist rate (assumes filter cut off near here), hence there is NO ringing.

There is no ringing - you keep on making false claims about CD and the mastering process. Why ?

When you state on complex sounds the ripple will unfold to the surrounding waves, this is another false statement. A Linear Time Invariant system, which is what LPCM, and CD is, does not behave like this.

So, not only have you made a false statement about ringing being on CD's, you have also made another false statement about ringing affecting the surrounding wave like a ripple.

The ringing event is finite and is half the length of the filter, and when it occurs is always small in amplitude. The only time it occurs is when the filter processes energy at the cut off frequency. All CD's and the mastering process ensures that this does not occur.


Regardless if the ringing is a transient phenomenon, it is ALWAYS present in the spectrum if it is present in the time domain. That is how FFT's work. The filter ringing is at the cut off frequency of the filter, so if it is present, you will see a blip on the spectrum at this frequency, it is NOT embedded in the remaining spectrum. Stating that ringing is invisible is a false statement.

Also, you made the claim that CD's have ringing, and did not provide evidence of which CD's do have ringing. Now you are stating that ringing is embedded in the CD which cannot be seen in the frequency domain, but can be seen in the time domain. You are continually making false statements, and changing them when challenged, and never provide evidence. Why ?

Regards,
Shadders.
Click to expand...
I think you are deeply confused about your own argument.

Band limited signal, again by definition has ringing. I have, again and again, offered you a perfectly valid explanation that what you reallt meant is that it is either low in amplitude or high in frequency and, therefore, inaudible. You have continued to argue that basic math doesnt exist and filters dont ring. Thank God you have come around ro embrace the rational view.

As to the visibility of filter and ADC ringing on a cumulative spectral power plot...i am not sure, but i think that it wont show up, for the following reasons:

1. It is much lower in magnitude
2. It is not at a single sinewave frequency, even for a single filter.
3. There are multiple filters employed in typical music production process

As for your pount that a modern CD production CAN avoid this issue - absolutely correct. However,, older masters can have this issue and careless production even today can have it as well.
 
davidsrsb said:
I was looking at the spectrum of some later Hendrix recordings and they are follow a smooth curve to 60dB down at 15kHz and brickwalled just above that. Electronically generated and digitally recorded music goes up further, but rarely very far unless as a result of extreme compression and digital clipping. Does faithful delivery of ultrasonics help much with the latter case?
Click to expand...
Only in facilitating the use of a better reconstruction filter on playback, it would seem.
 
Emlin said:
To counter a couple of your points:

MQA home page says: "MQA reveals every detail of the original recording". Are you willing to go along with that? If so, you disagree with your own statement that you want to ensure people are well informed about the technical side. So they can make informed choices.

If you allow MQA to spew out their guff without challenging it, you are complicit with their duplicity.

The ASA route may not work, but it'll only take a few moments of your time. And you have nothing to lose apart from those few moments. So why not?
Click to expand...

Because I regard it as a waste of time. I've watched others use the ASA route in the past and get nowhere.

Take the example you give. MQA might claim that the aliased folding and folding *does* recover every detail. Note that didn't say that it also added aliases that weren't in the source, only that it recovered details. An IT engineer knowledgable about audio might see the snag, but then the suits who *work in the ad biz* make the decisions being clueless about the engineering or theory.

Now emphasise the "every". I can't check that because they won't give us the source material to check. How do you show they were wrong without the source material. (Which MQA could claim they don't have to show anyway, because it belongs to the media company.)

Slide sideways and the claim may be said to mean "all that is audible" which then leads to arguments about what one person may hear when someone else doesn't, etc.

etc, etc...

Suits can run rings round the ASA, particularly if they can find an 'expert' to write some impressive 'reports' for them.

I've seen this all done in the past in one form or another in one case or another about other claims.

And *I've not seen the MQA ads*. If you have, you can complain and see how you get on. By all means reference this thread and pages I've written, etc.

However I suggest you look first into earlier examples. e.g. when Uncle Russ was taken to the ASA over claims about his mains cables. That is a textbook example of the problem with the ASA. Looking into these would probably save you time and stress.

I'm quite happy to discuss the technical side and let people make their *own* decisions on this. And as I've said before, I don't mind of some people 'prefer' MQA and get it. My central concern there is to avoid us reaching an 'MQA or nothing' situation.

And if people want the audio mags to speak out, write to them. That's probably a much better way forwards than the ASA.
 
Emlin said:
MQA are certainly lying in their claim that "MQA reveals every detail of the original recording." But they seem to have lost all contact with the concept of "legal, decent honest and truthful". I think that they are now just "desperate".
Click to expand...

I looked at the MQA site, looked at the press releases and saw the most recent from Pioneer. I then emailed Pioneer asking them how they can say such things. I've asked them before on something and they replied quickly. Let's see.

It is curious that Pioneer are upgrading their most popular streamers, the N70 and N50, for MQA for free as a software update. It seems MQA are giving away decoding to get traction. The N70 is also a very good device. More MQA desperation?
 
DimitryZ said:
I

I think you are deeply confused about your own argument.

Band limited signal, again by definition has ringing. I have, again and again, offered you a perfectly valid explanation that what you reallt meant is that it is either low in amplitude or high in frequency and, therefore, inaudible. You have continued to argue that basic math doesnt exist and filters dont ring. Thank God you have come around ro embrace the rational view.

As to the visibility of filter and ADC ringing on a cumulative spectral power plot...i am not sure, but i think that it wont show up, for the following reasons:

1. It is much lower in magnitude
2. It is not at a single sinewave frequency, even for a single filter.
3. There are multiple filters employed in typical music production process
Click to expand...
Hi,
OK - a band limited signal does NOT have ringing if it is constructed properly, or does not feature energy at the filter cut off frequency. The ADC's used today, have a high enough sample rate for the input signal that the filter does not ring, since there is no energy present at the filter cut off frequency.

Your last three points, indicate that you may need to examine ADC's - as their sample rate is so high, that there is no energy at their filter cut off frequency. Hence, there is no ringing.

Regards,
Shadders.
 
To win an argument that MQA doesn’t reveal every detail of the original recording you would need to produce an original recording and its MQA derivative and point to some detail that was audible in the original recording, but not in the MQA version.

Anyone care to do that?
 
AndyU said:
To win an argument that MQA doesn’t reveal every detail of the original recording you would need to produce an original recording and its MQA derivative and point to some detail that was audible in the original recording, but not in the MQA version.

Anyone care to do that?
Click to expand...
Hi,
MQA would have to agree to a null test between the master and their MQA file. They will not even agree to providing test tones MQA encoded files - so they will not allow a null test.

Regards,
Shadders.
 
Status
Not open for further replies.


advertisement


Back
Top