MQA and DRM

[crenca]

Wow, you are so caught up in your anti-MQA crusade that you can't take a step back to see lossless is not as objective as many think. Grouping MP3 and MQA into the same lossy bucket doesn't make you look smarter and doesn't make your argument against MQA look any better. The ability to realize nuances and actually discuss what is lost when something is lossy, is common sense. Maybe common sense isn't that common.

 

Nope. MP3 and in MQA are in the same lossy bucket because that is where they belong mathematically which is binary which is black and white. That is common sense but what is uncommon is the fundamental confusion between mathematics and pshycho-aucustic subjective "sounds like" evaulation...

[labjr]

It depends on who you believe. MQA suggests its technology has made the crown jewels better by eliminating problems with the crown jewels. I suppose down the road something could make the crown jewels even better than MQA, etc...

 

However, one could argue that the MQA version is now the crown jewels. But, MQA isn't an archival format if one wants to store absolutely everything (including the good and bad).

 

it's a strange situation.

 

 

OK, I'll back away from this conversation so I don't appear to be a shill for MQA.

 

By them saying both of those things, tells me they're trying to fool everyone into believing the format is best for everyone just because they say so. And IMO their ultimate goal is to convince the record industry to make MQA the only format available because otherwise the (now worthless because MQA makes them better) crown jewels would be vulnerable if we don't keep the out of the public.

 

It's a crock.

[smartin]

Nope. MP3 and in MQA are in the same lossy bucket because that is where they belong mathematically which is binary which is black and white. That is common sense but what is uncommon is the fundamental confusion between mathematics and pshycho-aucustic subjective "sounds like" evaulation...

 

Obviously true from a linguistic standpoint, but Chris also has a point that the terms lossy and lossless are more typically used interchangeably describing both the math and acoustic quality. I've often heard Pandora's free quality described as being more lossy than the paid quality (or gratingly 'less lossless'). That is incorrect in describing the absolute classification of the compression algorithm, but is accurate when describing the accuracy of the mathematical reproduction and its acoustic quality. While it was clear to me that mansr and others were referring to the classification of the algorithm, and thus it being a binary descriptor, I dare say at least some people reading this will infer about the other way of using the terms.

 

That of course has no bearing on MQA claiming to be lossless (as a clarification, is it still lossy when converting a 16/44 original to 24/44 MQA??), but impact how some may interpret all the discussion of 'lossy'.

 

Anyway, going to go spend some time with my bitch (she's a Wheaten Terrier).

[bogi]

Lossless shouldn't be considered against a distribution format. Apple distributes 256 kbps AAC. In many cases this AAC file is prepared by the mastering engineer and delivered to Apple in this format (despite Apple recommending everyone supply 24/96 to them). The distribution format is often lossy, thus why I don't recommend using it as the standard by which lossless is figured.

Lossy and lossless audio formats aren't about master or distribution. You don't find any mention of master or distribution here: https://en.wikipedia.org/wiki/Audio_file_format

Lossy and lossless are about processing raw audio data to store it in a particular audio format (encoding) or about retrieving raw audio data from already encoded content (decoding). Regardless of the origin of input data.

[labjr]

The thing is, Bob Stuart knew what he was doing when he chose those words.

 

Sent from my XT1528 using Tapatalk

[mansr]

In my professional realm, we deal in "mathematically lossless" vs. "visually lossless". Sample rate in PCM is roughly equivalent (this is by no means a consensus) to video resolution.

 

In video, consumer formats are always lossy because of the huge file sizes.

 

What you're arguing here is "audibly lossless", which is subjective.

 

Exactly, and "psychoacoustically transparent" is the correct term to describe a coding system with imperceptible losses.

 

I'd love to have lossless video too. Even Blu-ray occasionally has visible compression artefacts. Unfortunately, the data rates still make lossless unpractical for general use.

 

In the 90s, I was content with mp3 since that was what was feasible at the time. For audio, there is no longer any reason to use lossy compression. Technologically speaking, MQA is a step back.

[crenca]

Obviously true from a linguistic standpoint, but Chris also has a point that the terms lossy and lossless are more typically used interchangeably describing both the math and acoustic quality. I've often heard Pandora's free quality described as being more lossy than the paid quality (or gratingly 'less lossless'). That is incorrect in describing the absolute classification of the compression algorithm, but is accurate when describing the accuracy of the mathematical reproduction and its acoustic quality. While it was clear to me that mansr and others were referring to the classification of the algorithm, and thus it being a binary descriptor, I dare say at least some people reading this will infer about the other way of using the terms.

 

That of course has no bearing on MQA claiming to be lossless (as a clarification, is it still lossy when converting a 16/44 original to 24/44 MQA??), but impact how some may interpret all the discussion of 'lossy'.

 

Anyway, going to go spend some time with my bitch (she's a Wheaten Terrier).

 

The thing is, Bob Stuart knew what he was doing when he chose those words.

 

Sent from my XT1528 using Tapatalk

 

From a linguistic standpoint because of the mathematical, digital, truth of the matter. If you start "subjectivising" domains that are obviously "binary" and "black and white" like mathematics, well you might as well give up all hope of communication - really, 1+1=3 in such a world. digitally encoded music IS simply software, which IS simply mathematics, which IS what a computer (or DAC, etc.) does - it computes (it is a calculating machine). Either that information adds up (lossless) or it does not (lossy).

 

Yes, psychoacoustically I have heard 320 mp3's that sound better than 24/96. Yes, MQA is claiming that their lossy encoding "sounds better" than (well, everything else). Two separate issues, but they are related as well - Bob picks and chooses in a "just so" way - how convient.

 

Yes, I have great respect for Bob as a showman and a salesman - he knows what he is doing...

[MikeJazz]

Give me a break. The world isn't binary.

 

Chris, you need to accept the basics...

https://en.wikipedia.org/wiki/Lossless_compression

 

MQA not only is not lossless, is also not accepted as an archive format.

It's a "delivery" format...

[kumakuma]

Chris, you need to accept the basics...

https://en.wikipedia.org/wiki/Lossless_compression

 

MQA not only is not lossless, is also not accepted as an archive format.

It's a "delivery" format...

 

And it uses DRM...

[scintilla]

Exactly, and "psychoacoustically transparent" is the correct term to describe a coding system with imperceptible losses.

 

I've been away a few days and all this verbosity! I've read the claims of the patent application so far. So, you've been a solid all the way through this discussion, but two points: 1. A patentee can be her own lexicographer, even when the definition goes against plain-meaning or well-understood practice when the definition is explicitly defined in the spec. 2. Lossless includes also a channel-capacity definition, not just a perceptual coding equivalent. Here, they make explicit that the 'lossless' portion is the baseband 24/44 or 24/48 signal, but that it is obtained by noise-shaping and dithering to a 16/44 or 16/48 signal that supplementally includes a secondary signal in the bottom 8 bits of the channel; it's all about the definition.

 

Also, claim 14 explicitly claims a reversible watermark but the spec is silent as to application. So, I stand corrected. There is a clear DRM-element contemplated here. I still think it is used just for the authentication of the hardware element for end-to-end authentication. But I fully accept that it is not a conspiracy theory. No reason to claim it if it doesn't have economic value. And you know what that means... Cheers, people. Doing well. Now get back to reverse engineering.��

[audiventory]

Nope. MP3 and in MQA are in the same lossy bucket because that is where they belong mathematically which is binary which is black and white. That is common sense but what is uncommon is the fundamental confusion between mathematics and pshycho-aucustic subjective "sounds like" evaulation...

 

I always meant and mean "lossless" as "binary identical". "Lossy" is not negative or "sound worse". Because border between "audible" and "non-audible" loses is subjective.

[The Computer Audiophile]

Chris, you need to accept the basics...

https://en.wikipedia.org/wiki/Lossless_compression

 

MQA not only is not lossless, is also not accepted as an archive format.

It's a "delivery" format...

 

When the MQA version of a 16/44.1 track is larger than the original, we must ask what has been lost and take that into consideration.

 

I don't believe MQA should be in the same bucket as MP3. That's all I'm saying.

[mansr]

I don't believe MQA should be in the same bucket as MP3. That's all I'm saying.

 

You are entitled to believe whatever you want. You are not entitled to be right.

[The Computer Audiophile]

You are entitled to believe whatever you want. You are not entitled to be right.

 

Neither are you.

[mansr]

Neither are you.

The facts are on my side.

[audiventory]

In my opinion, MQA have other concept than mp3 in:

 

1. Folding full spectrum vs. compressing full band.

 

2. MQA optimized to deivery high resolution audio vs. mp3 for 44/48 kHz (may there new expansions, but I don't know about it).

 

3. MQA's compatibility to standard playback vs. mp3's obligatory using of decoder.

 

MQA and mp3 is different ideologies for me.

[realhifi]

I've been away a few days and all this verbosity! I've read the claims of the patent application so far. So, you've been a solid all the way through this discussion, but two points: 1. A patentee can be her own lexicographer, even when the definition goes against plain-meaning or well-understood practice when the definition is explicitly defined in the spec. 2. Lossless includes also a channel-capacity definition, not just a perceptual coding equivalent. Here, they make explicit that the 'lossless' portion is the baseband 24/44 or 24/48 signal, but that it is obtained by noise-shaping and dithering to a 16/44 or 16/48 signal that supplementally includes a secondary signal in the bottom 8 bits of the channel; it's all about the definition.

 

Also, claim 14 explicitly claims a reversible watermark but the spec is silent as to application. So, I stand corrected. There is a clear DRM-element contemplated here. I still think it is used just for the authentication of the hardware element for end-to-end authentication. But I fully accept that it is not a conspiracy theory. No reason to claim it if it doesn't have economic value. And you know what that means... Cheers, people. Doing well. Now get back to reverse engineering.��

 

Now THAT is some useful information. Thank you.

 

PS. If it is DRM and it's used in streaming I have no problem with that. The idea that a customer should be able to download files for "off-line" listening from a streaming service has always seemed odd to me. It implies that the listener then "owns" that file more or less so putting parameters on how the user can use that file seems reasonable to me.

[The Computer Audiophile]

The facts are on my side.

 

In your black and white world.

[firedog]

Gotta say I think the accurate way to describe MQA would probably be "transparent"; meaning it supposedly sounds as if it was technically lossless, even when it bit terms it isn't.

 

It's true, I think, that MQA is misleading us here, because when you read their somewhat more detailed descriptions of what it is/does, they are clearly talking about their claim that it is acoustically transparent to hi-res originals, even though in a literal sense it is mathematically lossy. Claiming in one place that it is "transparent", but in others that it is "lossless" isn't the most honest approach.

[The Computer Audiophile]

Gotta say I think the accurate way to describe MQA would probably be "transparent"; meaning it supposedly sounds as if it was technically lossless, even when it bit terms it isn't.

 

It's true, I think, that MQA is misleading us here, because when you read their somewhat more detailed descriptions of what it is/does, they are clearly talking about their claim that it is acoustically transparent to hi-res originals, even though in a literal sense it is mathematically lossy. Claiming in one place that it is "transparent", but in others that it is "lossless" isn't the most honest approach.

 

I agree.

[crenca]

Also, claim 14 explicitly claims a reversible watermark but the spec is silent as to application. So, I stand corrected. There is a clear DRM-element contemplated here. I still think it is used just for the authentication of the hardware element for end-to-end authentication. But I fully accept that it is not a conspiracy theory. No reason to claim it if it doesn't have economic value. And you know what that means... Cheers, people. Doing well. Now get back to reverse engineering.��

 

When you say "silent at to application", you mean high level "what is this for" and not implementation details (how it technically "phones home", etc.) correct?

[crenca]

In your black and white world.

 

I gotta do a thumbs down on this - math is math and is "black and white". The digital/mathematical definition of "lossless" is what what everyone means by it (MQA try as it might ain't going to change that). For years, when discussing these things this was agreed upon even while on the other hand the perceptual/psychoacoustic side was also discussed without having to confuse these aspects with the term "lossless".

 

Chris, when your in a hole the first thing to do is stop digging...

[scintilla]

Yes. In a spec, you usually do not provide much detail as to implementation, since you want to cover as much under the "doctrine of equivalents" as possible. I might have written using methods well-known in the art as will be appreciated by the skilled practitioner.

[crenca]

...I might have written using methods well-known in the art as will be appreciated by the skilled practitioner.

 

Ah, the well known dark arts of DRM...

[The Computer Audiophile]

I gotta do a thumbs down on this - math is math and is "black and white". The digital/mathematical definition of "lossless" is what what everyone means by it (MQA try as it might ain't going to change that). For years, when discussing these things this was agreed upon even while on the other hand the perceptual/psychoacoustic side was also discussed without having to confuse these aspects with the term "lossless".

 

Chris, when your in a hole the first thing to do is stop digging...

 

I'm not in a hole.

 

This is about much more than a static definition. The apple cart has been upset. People want to put things into categories that make them easier to understand for themselves and to push their own agenda. I'm pretty sure you all can see the difference between MQA and MP3. They aren't even close. Yet, you like to inject the word lossy whenever someone says anything positive about MQA. It's your effort to persuade people into your way of thinking. I get it. Do what you want.

 

However, it seems a bit obtuse to group MP3 and MQA in the same bucket. Thus, why I don't think the traditional terms lossy and lossless apply. You are using old school terminology to describe something that's a bit different. MQA is a combination of lossless and lossy, and it should be discussed with nuance.

 

In addition, it seems like people railing against MQA don't actually think any of this matters, they just want to see it fail. The same people who argue that high resolution can't be distinguished from redbook, are the same people saying MQA is terrible because it measures a certain way. In essence you are asking for subjective tests when it fits your narrative, but now you want to eschew subjective tests because so far they have refuted your MQA is terrible narrative.

 

MQA isn't lossless, I've said it a number of times. But, it isn't like an old school compression scheme. Grouping the two together is digging a deeper hole that goes against your crusade.