MQA Software / Hardware Decode Etc... Questions

[MikeJazz]

+1

Again, my only problem with MQA is IF MQA files sound worse than non-MQA files on non-MQA hardware, and we end up only with access to the MQA version. As long as we have a choice of versions, I don't have an issue with it.

 

+1. I expressed my concerns to Tidal support and they were unable to provide a clear answer...on the stream I find the same album twice. We know one is mqa but don't know which is which.

[mansr]

MQA has nothing to do with the decisions of record labels.

 

You are essentially trying to quarterback a game by watching it on TV. You only have some of the information, but you're talking like you have more than the team on the field.

 

Sorry, I don't know anything about baseball.

[mansr]

This is where the second leg of research should come in handy. You should be taking ABX tests to back up your claims of audibility.

 

Where did I make any claims about audibility?

 

I can claim all day long that DSD noise in the MHz range is "wrong" because there just shouldn't be noise, but if I can't hear it, my claims are kind of senseless.

 

The DSD noise is wrong. That's why DACs filter it out. If they didn't, it would ruin your tweeters or, failing that, cause very audible intermodulation products.

[mansr]

Your points would actually be accepted by more people if you didn't say things like "there's no reason for XYZ," when you really don't know if there is a reason.

 

When I say there is no reason, I mean there is no reason known to science and mathematics established 100 years ago and uncontested ever since. What's that saying about extraordinary claims? Oh yes, they demand extraordinary evidence. Hand-waving does not qualify as evidence, let alone extraordinary.

[The Computer Audiophile]

When I say there is no reason, I mean there is no reason known to science and mathematics established 100 years ago and uncontested ever since. What's that saying about extraordinary claims? Oh yes, they demand extraordinary evidence. Hand-waving does not qualify as evidence, let alone extraordinary.

So you're saying you are the best digital audio expert in the world or at least equal to the best. Nobody can know more than you.

[The Computer Audiophile]

Where did I make any claims about audibility?

 

If there are no audible effects, then why are you so against this? What's the point of saying something is vastly inferior technically, if it isn't inferior sonically? If audio file graphs were pictures to hang on a wall and you wanted it to look a certain way, then I get your point. But, we listen to them.

 

 

The DSD noise is wrong. That's why DACs filter it out. If they didn't, it would ruin your tweeters or, failing that, cause very audible intermodulation products.

Wow. I look at the noise as a design choice. There are very few good engineers who look at these types of design decisions with pros and cons as being wrong. Each decision has pros and cons with many competing influences.

[Jud]

When I say there is no reason, I mean there is no reason known to science and mathematics established 100 years ago and uncontested ever since.

 

Yes, but for people like me who don't have a background in that science and mathematics (thus what is quite obvious to you is completely obscure to us), it very much helps if you say things like you have in response to me, that MQA's filtering may under the right circumstances (content that triggers aliasing/imaging) produce audible distortion. (You mentioned NOS DACs as a point of comparison.)

[Samuel T Cogley]

So you're saying you are the best digital audio expert in the world or at least equal to the best. Nobody can know more than you.

 

Why not ask a math question instead of resorting to thinly veiled ad hominem?

[Jud]

Why not ask a math question instead of resorting to thinly veiled ad hominem?

 

Can you think of a good one?

[Samuel T Cogley]

Can you think of a good one?

 

Chris first :-)

[mansr]

So you're saying you are the best digital audio expert in the world or at least equal to the best. Nobody can know more than you.

 

Absolutely not. However, when one person (Bob Stuart) makes wild claims that contradict everything I've ever read on the topic, it makes me somewhat doubtful. Would a person such as Alan V. Oppenheim qualify as an expert in your eyes? He's just one of the authors represented on my bookshelf.

[mansr]

Yes, but for people like me who don't have a background in that science and mathematics (thus what is quite obvious to you is completely obscure to us), it very much helps if you say things like you have in response to me, that MQA's filtering may under the right circumstances (content that triggers aliasing/imaging) produce audible distortion. (You mentioned NOS DACs as a point of comparison.)

 

It's not really practical to repeat everything I've learned through years of studies each time I post something on a web forum.

[abrxx]

Absolutely not. However, when one person (Bob Stuart) makes wild claims that contradict everything I've ever read on the topic, it makes me somewhat doubtful. Would a person such as Alan V. Oppenheim qualify as an expert in your eyes? He's just one of the authors represented on my bookshelf.

Perhaps all that Chris is saying is that we should give Bob Stuart and Peter Craven the "benefit of doubt", since after all they are highly regarded digital audio engineers. Not gods, but not people without prior track records. Their papers and patents are very clear that certain amounts of aliasing is a trade-off to gain temporal precision, which they claim sounds better. This does come down to listening to the end product and seeing if the trade-off is acceptable to us. I can't believe for one second that Bob and Peter deliberately designed a system to sound worse for no good reason. Meridian (where MQA was developed) is after all an audiophile company.

[The Computer Audiophile]

Absolutely not. However, when one person (Bob Stuart) makes wild claims that contradict everything I've ever read on the topic, it makes me somewhat doubtful. Would a person such as Alan V. Oppenheim qualify as an expert in your eyes? He's just one of the authors represented on my bookshelf.

Skepticism is always cool with me, but you seem way beyond "somewhat doubtful."

 

A book on your shelf means absolutely nothing to me. The author is a smart guy, but I'm very doubtful he's make the bold statements you're making here. Perhaps it's because there are repercussions for him to say things unlike the anonymity of an online forum where one can push any agenda he wants.

[The Computer Audiophile]

Perhaps all that Chris is saying is that we should give Bob Stuart and Peter Craven the "benefit of doubt", since after all they are highly regarded digital audio engineers. Not gods, but not people without prior track records. Their papers and patents are very clear that certain amounts of aliasing is a trade-off to gain temporal precision, which they claim sounds better. This does come down to listening to the end product and seeing if the trade-off is acceptable to us. I can't believe for one second that Bob and Peter deliberately designed a system to sound worse for no good reason. Meridian (where MQA was developed) is after all an audiophile company.

 

Bingo. Even if people don't want to give the benefit of the doubt, at least argue both sides. At least take a guess as to why someone would make a design decision that appears different than the norm. Perhaps the view from the bleachers isn't as clear as the view from behind the plate. It just seems to me people are making wild accusations without all the information.

 

Speculation is totally fine, just state it as so.

[abrxx]

Regarding the Madonna album, I don't think this is a case of fake hi-rez. As in, this is not a rebook digital master simply up sampled. Its seems more likely that it is the original master (whether digital or analog tape) played back through an analog mixing desk, remastered in analog, and then the output of that desk captured at 192/24. Such a scenario is a valid hi-rez master, even if the input signal was not hi-rez. At least, so says Bob Ludwig of Gateway Mastering. (See the recent Montreal AES video on youtube: he doesn't refer directly to the Madonna album but he describes such a process as a valid use of hi-rez.)

 

Hi-rez to me seems like the early days of digital where various parts of the recording chain could be digital. Perhaps we need a similar provenance chain standard, like these sorts of notes:

 

- Recorded with high bandwidth microphones

- Mastered at 96/24

- Remastered at 192/24

 

Again, Bob Ludwig suggests such as thing, but in reality the labels are not going to listen...

[PeterSt]

As in, this is not a rebook digital master simply up sampled. Its seems more likely that it is the original master (whether digital or analog tape) played back through an analog mixing desk,

 

The "whether digital or analog tape" is wrong, because it is most clearly about a digital playback, which then is not upsampled but re-recorded in analog and all what happens is that from 22.05KHz up to 96KHz you caught noise and nothing else. And this is not digital noise, assumed that the source indeed has been properly filtered redbook (and I say it was).

So, fake all the way and consuming a lot of space for additional noise only.

 

Admitted, via MQA storage, the space is less.

[Samuel T Cogley]

Regarding the Madonna album, I don't think this is a case of fake hi-rez. As in, this is not a rebook digital master simply up sampled. Its seems more likely that it is the original master (whether digital or analog tape) played back through an analog mixing desk, remastered in analog, and then the output of that desk captured at 192/24. Such a scenario is a valid hi-rez master, even if the input signal was not hi-rez. At least, so says Bob Ludwig of Gateway Mastering. (See the recent Montreal AES video on youtube: he doesn't refer directly to the Madonna album but he describes such a process as a valid use of hi-rez.)

 

Hi-rez to me seems like the early days of digital where various parts of the recording chain could be digital. Perhaps we need a similar provenance chain standard, like these sorts of notes:

 

- Recorded with high bandwidth microphones

- Mastered at 96/24

- Remastered at 192/24

 

Again, Bob Ludwig suggests such as thing, but in reality the labels are not going to listen...

 

The difference is published provenance. Using the version of Rush's Moving Pictures available for download on HDTracks as an example, this title was originally mastered digitally in 16/48 and that digital master was recorded to analog tape. Those digital bits are long gone due to that legacy gear being out of production for so long. The label (perhaps at the behest of the artist?) made the provenance clear and the HDTracks version is 24/48.

 

The only reason we know that Like A Virgin came from a 14/44.1 (yes, fourteen bits) is that the gear it was mastered on was well known at the time (don't know the nomenclature, sorry) and a simple bandwidth analysis. Sampling that analog tape (similar to Moving Pictures) at a high sample rate (24/192) permits the label to sell the "audiophile" version of that title when the provenance does not justify the sample rate or the "audiophile" moniker. I've listened to the decoded MQA tracks. I can't imagine anyone calling that "audiophile grade".

 

The point of this is that some (maybe not all) labels will happily mislead their customers just to make an extra buck. I'm aware of the opinion that anyone expecting a Madonna release from that era to be "audiophile grade" deserves to be separated from their money, and I kind of agree. But MQA is not for knowledgeable audiophiles (I think that's been well established now). It's for the mass market, and it's being marketed in a way that claims it's bringing a level of sound quality to the masses that was previously only available to audiophiles. And no arguments here have dissuaded me that "Master Quality" is in no way an indication of objective quality. Audiophiles almost always rely on outside (boutique) mastering facilities to make what could objectively be called "master quality". MQA is not that.

[abrxx]

Thanks for all that info. But did you miss the thrust of my point? If industry leading mastering engineers tell us that its ok for redbook to be played back on an analog desk, remastered, and then resampled at a higher rate, who are we to tell them its fake? I think Bob Ludwig's words were along the line of "the higher sample rate accurately captures the decisions made at the [re]mastering stage".

 

The difference is published provenance. Using the version of Rush's Moving Pictures available for download on HDTracks as an example, this title was originally mastered digitally in 16/48 and that digital master was recorded to analog tape. Those digital bits are long gone due to that legacy gear being out of production for so long. The label (perhaps at the behest of the artist?) made the provenance clear and the HDTracks version is 24/48.

 

The only reason we know that Like A Virgin came from a 14/44.1 (yes, fourteen bits) is that the gear it was mastered on was well known at the time (don't know the nomenclature, sorry) and a simple bandwidth analysis. Sampling that analog tape (similar to Moving Pictures) at a high sample rate (24/192) permits the label to sell the "audiophile" version of that title when the provenance does not justify the sample rate or the "audiophile" moniker. I've listened to the decoded MQA tracks. I can't imagine anyone calling that "audiophile grade".

 

The point of this is that some (maybe not all) labels will happily mislead their customers just to make an extra buck. I'm aware of the opinion that anyone expecting a Madonna release from that era to be "audiophile grade" deserves to be separated from their money, and I kind of agree. But MQA is not for knowledgeable audiophiles (I think that's been well established now). It's for the mass market, and it's being marketed in a way that claims it's bringing a level of sound quality to the masses that was previously only available to audiophiles. And no arguments here have dissuaded me that "Master Quality" is in no way an indication of objective quality. Audiophiles almost always rely on outside (boutique) mastering facilities to make what could objectively be called "master quality". MQA is not that.

[Samuel T Cogley]

Thanks for all that info. But did you miss the thrust of my point? If industry leading mastering engineers tell us that its ok for redbook to be played back on an analog desk, remastered, and then resampled at a higher rate, who are we to tell them its fake? I think Bob Ludwig's words were along the line of "the higher sample rate accurately captures the decisions made at the [re]mastering stage".

 

Apples/oranges in my opinion. We're talking about a digital master that had to be captured to analog for the purposes of preservation. Not an "artistic decision" to add analog to the production chain. My understanding (admittedly anecdotal from some peripherally involved in the original project) is that the original CD of Like A Virgin was essentially a clone of that original 14/44.1 master. I'm happy to hear other information if you have it.

[abrxx]

Apples/oranges in my opinion. We're talking about a digital master that had to be captured to analog for the purposes of preservation. Not an "artistic decision" to add analog to the production chain. My understanding (admittedly anecdotal from some peripherally involved in the original project) is that the original CD of Like A Virgin was essentially a clone of that original 14/44.1 master. I'm happy to hear other information if you have it.

 

To be clear I am making some assumptions:

 

- that the Madonna track was remastered (no idea if it was, only that many other MQA tracks I have listened to have definitely been remastered)

 

- that the remastering was done through an analog console (since the existing archive master was analog)

 

- to accurately capture the remastering, a high rez sample rate was used.

 

Again, to be clear, all conjecture on my part. But IF that was the case, I am saying such a situation mirrors the case where digital assets are assembled in the analog domain and then resampled at a higher rate, something according to Bob Ludwig is not uncommon in his experience.

 

Such a scenario seems different from simply doing a digital upsample of a file and selling it at a hi-rez recording.

[Jud]

Perhaps all that Chris is saying is that we should give Bob Stuart and Peter Craven the "benefit of doubt", since after all they are highly regarded digital audio engineers. Not gods, but not people without prior track records. Their papers and patents are very clear that certain amounts of aliasing is a trade-off to gain temporal precision, which they claim sounds better. This does come down to listening to the end product and seeing if the trade-off is acceptable to us. I can't believe for one second that Bob and Peter deliberately designed a system to sound worse for no good reason. Meridian (where MQA was developed) is after all an audiophile company.

 

So here's the problem with that: Meridian's old filtering systems did the same thing (traded off some frequency domain distortion for gains in the time domain). No problem. With MQA, what's being done on top of that is to (1) take a problem many feel is not a very big one (streaming file size); (2) offer a proprietary solution when there are non-proprietary ones (e.g., compressed flac or alac); (3) where it's been demonstrated that the non-proprietary solutions permit higher resolution in less space than MQA's proprietary one.

 

 

In other words, no one has presently come forward with a sound mathematical/technical argument that MQA's format is proprietary due to *technical* necessity or advantage. Given this, folks have of course been making the argument it is proprietary in order to keep MQA's processing and/or the MQA musical content locked down, now or in the future. I certainly can't say the prospect of DRM would necessarily be *un*attractive to the labels; it could well have been a sales point for MQA.

[Jud]

To be clear I am making some assumptions:

 

- that the Madonna track was remastered (no idea if it was, only that many other MQA tracks I have listened to have definitely been remastered)

 

- that the remastering was done through an analog console (since the existing archive master was analog)

 

- to accurately capture the remastering, a high rez sample rate was used.

 

Again, to be clear, all conjecture on my part. But IF that was the case, I am saying such a situation mirrors the case where digital assets are assembled in the analog domain and then resampled at a higher rate, something according to Bob Ludwig is not uncommon in his experience.

 

Such a scenario seems different from simply doing a digital upsample of a file and selling it at a hi-rez recording.

 

But also provenance the consumer ought to know about, because it could arguably affect the sound quality at playback.

[abrxx]

But also provenance the consumer ought to know about, because it could arguably affect the sound quality at playback.

I totally agree. Sadly I don't think the labels care that much (if at all?). Perhaps MQA 2.0 can address this?

[Samuel T Cogley]

To be clear I am making some assumptions:

 

- that the Madonna track was remastered (no idea if it was, only that many other MQA tracks I have listened to have definitely been remastered)

 

- that the remastering was done through an analog console (since the existing archive master was analog)

 

- to accurately capture the remastering, a high rez sample rate was used.

 

Again, to be clear, all conjecture on my part. But IF that was the case, I am saying such a situation mirrors the case where digital assets are assembled in the analog domain and then resampled at a higher rate, something according to Bob Ludwig is not uncommon in his experience.

 

Such a scenario seems different from simply doing a digital upsample of a file and selling it at a hi-rez recording.

 

Perhaps where we disagree is that Warner did not publish the provenance of the source, and that the final HDTracks (and Tidal MQA) product sample rate is not justified by the actual provenance. I'm not disputing Mr. Ludwig's expertise, I just think that a more honest final consumer sample rate is 16/44.1.