DARKO: MQA: a non-hostile takeover?

[The Computer Audiophile]

Again, we're speaking from different experience levels. I have an advantage in that I have the reference of comparing the actual live microphone feeds with the recorded. If I as a producer determine my product is as accurate a representation of the analog microphone feed possible with the equipment available, why on earth would I be interested in some third party applying their sweetening process to my recording without my consent?

I hear what you're saying, but you are excluding all new technologies by believing you're at the pinnacle. One never knows how much better it could be until one hears it for the first time.

[tailspn]

Shouldn't we be aiming to reproduce the actual sound the musicians hear rather than the microphone feed? Keith Richards doesn't tune his guitar through monitors. He wants the sound of his amp.

 

Keith doesn't use microphones? But your point is well taken, microphone selection, and placement is the art in recording, with the objective of reproducing the sounds and enviorment the musicians produce. Unfortunately in recording, microphones are a necessary evil, each with their own characteristic coloration.

[The Computer Audiophile]

@tailspn - Maybe it's just me, but I'd be totally satisfied and supportive of your comments if you said you had tried MQA and don't like the sound or if you had tried it and said you like it but upon further investigation the business model isn't for you. It seems like right now you are relying on speculative information to form your opinion.

[The Computer Audiophile]

Keith doesn't use microphones? But your point is well taken, microphone selection, and placement is the art in recording, with the objective of reproducing the sounds and enviorment the musicians produce. Unfortunately in recording, microphones are a necessary evil, each with their own characteristic coloration.

You know you're taking my example and twisting it beyond what's reasonable. Of course mics are used, but my goal would be to reproduce the actual instruments, not the captured sound.

[tailspn]

I hear what you're saying, but you are excluding all new technologies by believing you're at the pinnacle. One never knows how much better it could be until one hears it for the first time.

 

Absolutely. That's mom and apple pie. Technology will improve the tools we use to record and reproduce music. But I never implied I was at the pinnacle. It's just what's currently available by using 5 channels with puristic mic techniques, DSD256 recording with no post processing, I obtained results very pleasing and believable to me. And a few others. More channels, higher DSD bit rates, who knows.

 

But the psychoacoustic manipulation of a recording after it's been produced is just not for me. That's not to say it may make many recordings sound more appealing. YMMV

[The Computer Audiophile]

Absolutely. That's mom and apple pie. Technology will improve the tools we use to record and reproduce music. But I never implied I was at the pinnacle. It's just what's currently available by using 5 channels with puristic mic techniques, DSD256 recording with no post processing, I obtained results very pleasing and believable to me. And a few others. More channels, higher DSD bit rates, who knows.

 

But the psychoacoustic manipulation of a recording after it's been produced is just not for me. That's not to say it may make many recordings sound more appealing. YMMV

I hear ya Tom :~)

[Axiom05]

Granted, and probably for the better given that which the majority of studios use for monitoring. But do you believe the actual digital content as heard in the final mastered studio production is the same as delivered to you? It should be.

 

Yes, I believe that the digital file that I am playing (assuming I purchased a copy of the digital master) is the same as the studio production master.

[crenca]

..and are comprehending the downsides of MQA (or any other product that wants to be an "end to end" walled garden/format takeover). I have yet to hear MQA, and while I will be at RMAF this year any demo there will not be an (even rudimentary) A/B so I don't expect to to be able to hear it for a while yet. However, I suspect that any improvement MQA brings will never offset the downsides. As many commentators here have noted, Chris's (and the rest of the effort by the Audiophile Press to rally behind MQA) "but you have not heard it" rings hollow to me given the real and significant dowsides of MQA (and anything like it). Despite what MQA supporters argue, one does not have to experience something to be reasonably certain its cost is too high. I don't need to experience radiation poisoning to understand the benefits of not living near a nuclear power plant. I don't need to experience a heart attack to understand the cost or be convinced that a little prevention (exercise, good diet, etc.) is prudent. I don't need to "hear" MQA to understand that a SQ tweak (even a significant if not overwhelming one) is not worth the cost of an end to end takeover of the entire music digital chain ("hostile" or not).

 

In the end, MQA strikes me as a very very expensive (the cost not being in $ terms, but control) DSP solution to a set of problems that can be "solved" in ways that are better from a SQ standpoint as well as a freedom-to-innovate point of view (by better recording/mixing techniques, high sample rate PCM/DSD, "open" products like HQPlayer, time aligned speakers, etc. etc.). I enjoy my DSP solutions (such as quality EQ, etc.) but I would give them up in a second if they were closed and applied in a non-transparent, "one size fits all" product like MQA.

 

Also, there is so much over the top exaggeration by the industry it is kind of pathetic. When I come home from my local symphony with the sound of "real" music fresh in my ears/brain, and a pop in the best recording I have of the same piece and listen to it in my best system (which currently is a head-fi rig) the word "cartoonish" simply does not sum up the difference. There are significant differences fer sur, but there is also a real fidelity and quality that is palatable and obvious - cartoons need not apply.

 

Perhaps MQA will be a niche product for those who listen to poorly recorded/highly processed/mixed music that can always benefit from an easy to implement "magic bullet" SQ/DSP solution and for whom an "end to end" takeover is neither here nor there. Thing is, most folks who would benefit are also the very ones who (as all in the industry will attest) do not care if it is 128mp3 through $2 playback chains and are all pirates anyways and thus deserve all the DRM they get - and we are back to the fact that MQA was never about SQ in the first place...

[Confused]

For years I have read stuff about 'hi-res' music files, and much about how 16/44.1 is in fact all you need, going 24bit will add nothing that the human ear and brain will detect. OK, fair enough. Then again, it has always puzzled me that we had 16/44.1 in the 80's, and now thanks to Apple, Spotify and co. 320kbs is the norm and even considered 'high quality', and frankly for 99.9% of the population this is of no concern whatsoever.

 

Of course in the meantime broadband speeds have been getting faster and faster, 4G replaces 3G on 'phones, so actually streaming 16/44.1 or even 24bit looks entirely viable. After all, many folk are streaming 1080i and 4K TV without issue (and with audio), this is an amount of data that makes a music file look utterly irrelevant. Plus, look at 16/44.1 or 24bit files, very high quality and playable almost universally on anything that will play electronic files, your 'phone, your car, your hifi, anything.

 

 

Then MQA arrives to solve this problem. Eh? What problem. Then MQA demands bespoke hardware to work properly. Ah yes, but it can correct for the errors in the recording studio ADC (not the biggest issue with most recordings in my experience) and it also takes account of psychoacoustic effects of the human brain. (which the theoretically 'perfect' reproduction of 24bit somehow misses)

 

I am a naturally cynical person, I'll be honest, but all this MQA stuff strikes me as being total and utter rubbish! Just my view, I am sure others will look at this differently. Then again, I do not think this is something I'll need to worry about too much personally. To this day, the vast majority of what I actually end up listening to is either 16/44.1 or 'legacy' vinyl. I think it will be a long time before anything is released on MQA that I will actually want to listen to, if MQA lasts that long.

[tailspn]

Yes, I believe that the digital file that I am playing (assuming I purchased a copy of the digital master) is the same as the studio production master.

 

That fact is a significant reversal on the labels part brought about by the download market for audiophiles in just the last few years. Prior to that, labels were very reluctant to offer actual edited masters for fear of giving up the crown jewels. Once the EM was released, the labels had no protection for future unauthorized releases or re releases. Most labels do not have the economic resources to enter into a suite to claim copyright.

 

This EM availability is a huge customer win about sound quality.

[beetlemania]

24bit will add nothing that the human ear and brain will detect.

 

This opinion will not likely find many adherents around here - certainly not me. You'll find a much friendlier audience over at Hydrogen Audio. But I otherwise share your cynical view of MQA. Even if it does sound better, the downsides are considerable.

[esldude]

Shouldn't we be aiming to reproduce the actual sound the musicians hear rather than the microphone feed? Keith Richards doesn't tune his guitar through monitors. He wants the sound of his amp.

Absolutely not. Not until we can buy seats among the band or orchestra. Musicians don't know how they sound to the live audience in attendance.

 

Sent from my Nexus 6P using Computer Audiophile mobile app

[The Computer Audiophile]

Absolutely not. Not until we can buy seats among the band or orchestra. Musicians don't know how they sound to the live audience in attendance.

 

Sent from my Nexus 6P using Computer Audiophile mobile app

I should have been clearer with my wording. I'd prefer to hear the sound of the musicians / instruments rather than the sound after it has gone through components and speakers. I don't want the musician's perspective

[Maldur]

But the psychoacoustic manipulation of a recording after it's been produced is just not for me.

 

Exactly. I think that this kind of manipulation with bandwith, bitdeph/dynamic range and/or lossy compression is just like a snake oil. MP3, BBE, MQA etc. is marketed with slogans like "better, clearer, more efficient". How BBE can restore high bandwith when source has no clue of existence of life after Nyquist point? This can't be real. MQA uses adaptive optimisation of bandwith and bitdeph, lossy compression and all this ise "better"? Better from what? MP3? Probably...

John Siau - https://benchmarkmedia.com/blogs/application_notes/163302855-is-mqa-doa

[The Computer Audiophile]

I've mentioned this once before and I think it bears repeating. Almost all the people I've talked to who don't like MQA or who have bad things to say about it, haven't signed an NDA with the company to find out the fine details. Most are speculating. Others have made statements about it that I believe are marketing based, such as "We don't support MQA because ..." On the other hand, those who have signed the NDA and are digging deep into the technology, with a very good understanding of digital filtering, have said it's really advanced and that it takes brilliant minds to come up with this stuff.

 

I really worry that MQA is technically beyond the reach of most peoples' understanding and is too confusing to both manufacturers and consumers. Thus, it won't succeed because of this difficulty and confusion.

 

If it doesn't succeed on its merits, I am totally fine with that. that's the way a somewhat free market works. I just hope it gets a fair shake. We all win if it gets a fair shake, wether or not it succeeds.

[Maldur]

Chris, why we need very advanced, advanced or not advanced filtering when PCM and DSD can offer "untouched" quality? With FLAC everyone can save network bandwith even more than MQA does (in similar quality standpoint)?

[The Computer Audiophile]

Chris, why we need very advanced, advanced or not advanced filtering when PCM and DSD can offer "untouched" quality? With FLAC everyone can save network bandwith even more than MQA does (in similar quality standpoint)?

Great question @Maldur - It's a very complicated subject. I suggest starting with this information from Stanford, An Introduction To Digital Filters With Audio Applications -> https://ccrma.stanford.edu/~jos/filters/

[Paul R]

I really worry that MQA is technically beyond the reach of most peoples' understanding and is too confusing to both manufacturers and consumers. Thus, it won't succeed because of this difficulty and confusion.

 

I doubt anyone able to navigate a computer and deal with computer audiophile subject is going to be intimidated by MQA. Even the math behind it isn't that difficult.

 

 

-Paul

[The Computer Audiophile]

I doubt anyone able to navigate a computer and deal with computer audiophile subject is going to be intimidated by MQA. Even the math behind it isn't that difficult.

 

 

-Paul

I'd love to agree with you, but that's just not what I'm hearing from people who do this for a living. One person who was talking to me about this, and who doesn't work for MQA, was talking to me about how he used to use several different filters during a single track, based on the content of the track. He said MQA doesn't do this, but does some stuff that is in the same realm of advanced filtering.

[Paul R]

I'd love to agree with you, but that's just not what I'm hearing from people who do this for a living. One person who was talking to me about this, and who doesn't work for MQA, was talking to me about how he used to use several different filters during a single track, based on the content of the track. He said MQA doesn't do this, but does some stuff that is in the same realm of advanced filtering.

 

How much of that is just learning something new vs. how complex it really is? MQA is still new, people have not had time to get comfortable with it yet, so that kind of opinion is very much to be expected.

 

In any case, it will all work out one way or another.

 

-Paul

[The Computer Audiophile]

How much of that is just learning something new vs. how complex it really is? MQA is still new, people have not had time to get comfortable with it yet, so that kind of opinion is very much to be expected.

 

In any case, it will all work out one way or another.

 

-Paul

I'm with you, it will all work out in the end. The sky isn't falling :~)

 

You'd be surprised at how little digital engineering knowledge some high end companies actually have. A few engineers do all the work for many companies.

[Miska]

I've mentioned this once before and I think it bears repeating. Almost all the people I've talked to who don't like MQA or who have bad things to say about it, haven't signed an NDA with the company to find out the fine details.

 

It is not necessary to get anything from the MQA company as such. It is enough to take original non-MQA version of the content and then the MQA encoded one and compare the two. MQA both through decoder and without decoder.

 

MQA is lossy process. The way it manages to put the entire 22.05 - 44.1 kHz band into few bits and fold it down into LSBs of the 0 - 22.05 is because it first puts everything through heavy low-pass filter cutting almost everything out above the 22.05 kHz. So there is almost no information left to be encoded and folded. Also for this reason, for example 352.8 kHz DXD input content is encoded to 88.2 kHz sampling rate. So it has only twice the bandwidth of the FLAC rate and even then the top octave is heavily lossy (almost empty).

 

Since transients specifically require wide bandwidth and contain lot of high frequency content - sharper the transient more HF content it contains, the MQA "de-blurring" filter actually softens the transients by removing large portion of the HF content.

 

Overall, if you take for example 88.2 kHz 24-bit PCM white noise. There is no way you can fold it into 44.1 kHz 24-bit PCM in a bit-perfect loss-less way. Pure noise is uncompressible and lower rate simply doesn't have enough bandwidth to contain the information. So by definition, from information theory point of view, the process must be lossy. The way MQA chose to do it, is to take "hires" part out of "hires".

[The Computer Audiophile]

Hi @Miska - Reverse engineering something is quite different from being on the inside and knowing the facts from those who invented the technology.

 

That said, I'm not discounting your statements.

[Miska]

Hi @Miska - Reverse engineering something is quite different from being on the inside and knowing the facts from those who invented the technology.

 

Sure, but proof is in the pudding regardless...

 

If I see X going in and Y coming out, I can conclude X != Y and that difference is Y - X. Even if there's a black box in the middle.

[Confused]

I've mentioned this once before and I think it bears repeating. Almost all the people I've talked to who don't like MQA or who have bad things to say about it, haven't signed an NDA with the company to find out the fine details. Most are speculating. Others have made statements about it that I believe are marketing based, such as "We don't support MQA because ..." On the other hand, those who have signed the NDA and are digging deep into the technology, with a very good understanding of digital filtering, have said it's really advanced and that it takes brilliant minds to come up with this stuff.

 

I really worry that MQA is technically beyond the reach of most peoples' understanding and is too confusing to both manufacturers and consumers. Thus, it won't succeed because of this difficulty and confusion.

 

If it doesn't succeed on its merits, I am totally fine with that. that's the way a somewhat free market works. I just hope it gets a fair shake. We all win if it gets a fair shake, wether or not it succeeds.

OK - I guess I have already shown my hand as being in the 'cynical' camp, but I am prepared to be completely open minded. In fact, I should be pretty much MQA's target audience, I still spend money on recorded music, and I do care about deeply about how well recorded & mastered the music is. Am I an expert in digital filtering? No I am not. So I take a look at MQA's own website. It tells me 'Conventional audio formats discard parts of the sound to keep file size down, but part of this lost detail is the subtle timing information that allows us to build a realistic 3D soundscape in our minds.' ....'With MQA, we go all the way back to the original master recording and capture the missing timing detail. We then use advanced digital processing to deliver it in a form that’s small enough to download or stream.' OK, fine. However, I am happy to purchase 24bit PCM files. Now my basic understanding is that many modern recordings are mastered in PCM, and that a 24bit file has had nothing disregarded to keep the file size down. So all I see in MQA's material are words that completely fail to convince me that MQA offers something that say a Linn 'Studio Master' 24bit file does not offer. I am not posting this as a clever definitive argument against MQA, I do not have a deep enough knowledge to claim that, but what I am saying is that even reading MQA's own website I end up with the feeling it is all a bit of a con. So what am I missing here? I should be someone who reads MQA's words and gets excited about the prospect of buying some music I would enjoy in this fine new format. But no, I read their own words and think it's a con! So if I have missed the point, which is entirely possible, it is for sure a case that MQA's message is certainly not working for me! As I say, I am completely open minded so more than happy to be convinced otherwise.