MQA: A Review of controversies, concerns, and cautions

[RichB]

What exactly is the benefit of preserving ultrasonics?
We cannot hear them. Never, ever.

Let's accept, as a given, that there are those that can detect the difference between a 44.1 and 96k recording derived from the same master on "True" audiophile systems.
Therefore, the difference must be in the audible range.
If you don't believe that, there is not point in further discussion.

Again, assuming it is heard, it is either more accurate or less accurate. This cannot be denied.

Case 1: The reproduction is more accurate.
The DAC and filtering performed better with the 96K signal.
Ultrasonics are not heard, so that's about it.

Case 2: The reproduction is less accurate.
The presence of ultrasonics could cause the processor, amplifier, and speaker to modulate distortion in the audible range.
This is not unreasonable.

Case 3: I Like MQA
OK. Magic rocks. I have nothing to say if we get to case 3.

IMO, the most likely cause for MQA differentness is the use a proprietary reconstruction filter.
That could well be audible and preferred but you don't need MQA for an alternate reconstruction filter.
The Marantz AV8805 has one of those, and if you take away the hand-waving, it is broken.

MQA is not required to obtain and alternate filter, fast slow, minimum or linear phase. Actually, it removes that option.
Why would anyone want this restriction?

MQA is in a 44.1 container so it does not benefit from higher sample rates (If that's a thing).

Perhaps there will be even better DACs that magically transform CD quality into near HD Audio.
In that case, MQA please don't F up the source file (even if it is streamed)

- Rich

 

[Soniclife]

Would MQA do the same for my *****?
S

Why does the forum stick little stars for letters?

Only if you let Bob shrink it first, but some will be lost forever in that process.

 

[RichB]

Is it confirmed that Roon will integrate with Amazon lossless?

- Rich

 

[Krunok]

Perhaps there will be even better DACs that magically transform CD quality into near HD Audio.
In that case, MQA please don't F up the source file (even if it is streamed)

- Rich

It won't be, as it can't be - DACs are already reconstructing every bit of information from CD quality files so there is nothing left to "transform".

 

[DonH56]

The arguments I have heard for ultrasonics:

1. Allows greater timing accuracy (a myth in practice for those not understanding how sampling works);
2. Ultrasonic sounds exist in nature and can produce undertones (beat frequencies) that fall into the audible range (at least somewhat plausible if unlikely to matter in the real world); and,
3. Somehow we hear ultrasonic sounds at some (perhaps subconscious) level and thus they are vital to full-range reproduction (uhhhh...)

These are reasons I have heard for ultrasonic reproduction independent of the technical arguments for oversampling to relieve the anti-alias and image filter requirements. Getting old, probably forgot a few.

 

[RichB]

The arguments I have heard for ultrasonics:

1. Allows greater timing accuracy (a myth in practice for those not understanding how sampling works);
2. Ultrasonic sounds exist in nature and can produce undertones (beat frequencies) that fall into the audible range (at least somewhat plausible if unlikely to matter in the real world); and,
3. Somehow we hear ultrasonic sounds at some (perhaps subconscious) level and thus they are vital to full-range reproduction (uhhhh...)

These are reasons I have heard for ultrasonic reproduction independent of the technical arguments for oversampling to relieve the anti-alias and image filter requirements. Getting old, probably forgot a few.

1. It seems to me that could only be true for high-sample rate, not 44.1 with origami.

2. Then those beat frequencies would be encoded in the audible range. I don't believe the mic knows to exclude a beat frequency.

3. Yes, I have felt that argument in my bones.

I can switch the filters on the Oppo UDP-205 and hear a difference. I find the most likely cause for MQA audible difference, is the reconstruction filter, ignoring bias, of course.

As I said, assuming HD Audio has a benefit, MQA is not HD Audio.
It is merely 44.1/13 bit source with lossey ultrasonics, plus a funky filter.
They have expert marketing that has successfully targeted a gullible industry.

- Rich

 

[Eirikur]

Does the higher frequency rate at 88.2 kHz or above provide any benefit like reduced quantization errors or greater resilience from decoding errors?

Error resilience is typically a data feature of the a physical medium like CD (using Reed-Solomon error correction codes) or transmission protocol like TCP, usually detection + resend or forward error correction in TCP-IR.
In most cases the audio data container (e.g. APE or FLAC) is only equipped to detect errors, and this is regardless of the specific content.

 

[mansr]

Maybe we should take Amir more seriously. After all, he has far more experience than most of us in unsuccessfully marketing an unnecessary audio format.

 

[Thomas savage]

Would MQA do the same for my *****?
S

Why does the forum stick little stars for letters?

Maybe if MQA audio fails Bob can develop a MQA cream for our members ...

 

[firedog]

Is it confirmed that Roon will integrate with Amazon lossless?

- Rich

No. At the Roon forum they’ve only confirmed that Amazon is aware of Roon and the integration issue. Nothing more. They suggested people contact Amazon Music and request integration. They said that a large number of requests will influence Amazon.

 

[RichB]

No. At the Roon forum they’ve only confirmed that Amazon is aware of Roon and the integration issue. Nothing more. They suggested people contact Amazon Music and request integration. They said that a large number of requests will influence Amazon.

So if I buy it, it will come.

- Rich

 

[HammerSandwich]

I have a question about MQA's success so far: Why have labels adopted it?

It seems that MQA can offer the labels:

1. High-res as a marketing point.
2. Charging more for high-res.
3. High(er)-res in smaller files.
4. Deblurring.
5. Promises that end users won't have access to full digital data, aka the crown jewels.

Have I missed anything?

1 & 2 could work the same with FLAC or whatever. I'll grant that the "authentic!" light is dead-simple for any end user to understand, so there may be a psychological marketing factor here, even if we already know it's bogus. (Thanks, @mansr!) Otherwise, I don't see what makes MQA distinct from other high-res in these areas.

3 could affect Tidal, but I again fail to see how it's an issue for a label. Already mentioned comparisons indicate that the actual savings are insignificant, and anyone who questions the cost of sending a byte should look at what GOG & Steam sell for a few bucks. (Spoiler: multi-GB games that you may download repeatedly, without restriction.)

Regarding 4, however - and however well - this actually works, it doesn't appear to have much promotion outside of the audiophile press.

So, those first points don't seem to offer much. There are a relative handful of big spenders who might matter to a Chesky or 2L, but they're not affecting, say, Beyonce sales. Normal people select their music by who performed it, not if it has an MQA logo. Because MQA encoding obviously increases costs, how much additional profit can any of these areas offer a major label? Is it truly worth the extra licensing & effort versus simply selling 24/96 FLAC?

Which leaves 5.

So I'm left a bit suspicious that the labels responded to a DRMish sales pitch. Sure, they've been knocked off the DRM horse, but that doesn't mean they wouldn't like to climb back on. And they have a long track record of bad behavior (e.g., royalty payments) that helps me to question their current motives. The loudness war sure as hell shows that they're not offering MQA or other high-res just to provide better sound quality at the same profits...

I understand @amirm's point that MQA files can be copied freely, and I'll even grant that current players don't include a dormant downsampler. Mind you, I don't actually know that buyers doesn't receive files with unique watermarks or that a downsampler isn't hidden in the players, but I'll concede both for discussion.

Even granting those points, what is MQA's upside for the labels? If it's not DRM as an actual copy-restriction mechanism, protecting the last few dB of the crown jewels with a proprietary wrapper still smells like DRM with a veneer of "New & Improved!" marketing. And, if it's not a "DRM Light" package, what enticed the labels to buy in? Are the majors offering all of their MQA tracks in 24/96 FLAC as well? For a not-higher price? THAT would convince me that there's not a shady angle being played here. Otherwise, I'm sad to say that I just haven't seen the logic of it.

 

[HammerSandwich]

Maybe if MQA audio fails Bob can develop a MQA cream for our members ...

Well, that was a savage pun indeed.

 

[DonH56]

1. It seems to me that could only be true for high-sample rate, not 44.1 with origami.

2. Then those beat frequencies would be encoded in the audible range. I don't believe the mic knows to exclude a beat frequency.

3. Yes, I have felt that argument in my bones.

I can switch the filters on the Oppo UDP-205 and hear a difference. I find the most likely cause for MQA audible difference, is the reconstruction filter, ignoring bias, of course.

As I said, assuming HD Audio has a benefit, MQA is not HD Audio.
It is merely 44.1/13 bit source with lossey ultrasonics, plus a funky filter.
They have expert marketing that has successfully targeted a gullible industry.

- Rich

To be clear I do not buy any of them personally. Note (1) is a myth, stated in the post; that is not the way sampling works. I agree on you with (2). And I "hear" the argument but don't buy it for (3). Switching filters has more to do with in-band response than ultrasonic impact IMO. Of course, "ultrasonic" is a rapidly decreasing number for me...

I should have stated those are the arguments I have heard, not arguments I believe to be true or even factual.

 

[restorer-john]

Only if you buy the special MQA unfolder.

 

[Wombat]

Yes , plus this idea because it probably won't break through we don't need to worry about it , well 99% of folks don't buy audiophile DACs so in effect there's nothing to worry about if the odd beautique audio company makes a crap one and sells it for a small fortune..

Why worry as most don't even know what a DAC is let alone care about how it measures and obviously won't even be aware of these boutique products let alone buy one. They are oblivious, listening to audio coming out their phones, TVs ,Amazon echos etc .

It's all only relevant to a very very small percentage of people, so let's all go home and find something else to do .

Bizarre.

That can be said about audio forums as well.

 

[Wombat]

As long as you don't buy this one: https://www.amazon.co.uk/Fairytales-Original-Master-MQA-VINYL/dp/B0789G9WT9

Appropriately named record.

 

[digicidal]

What exactly is the benefit of preserving ultrasonics?
We cannot hear them. Never, ever.

With the level of manufacturing excellence available today... how will speaker builders survive if we don't melt a large quantity of tweeters over the next few decades? Besides you're missing all that air... you can hear the air after all. You did know that didn't you?

Every time I read a review about hearing the air in a recording I'm reminded of the times I could taste and smell colors - but I don't recommend LSD as a long-term solution for understanding audiophilia.

 

[Blumlein 88]

With the level of manufacturing excellence available today... how will speaker builders survive if we don't melt a large quantity of tweeters over the next few decades? Besides you're missing all that air... you can hear the air after all. You did know that didn't you?

Every time I read a review about hearing the air in a recording I'm reminded of the times I could taste and smell colors - but I don't recommend LSD as a long-term solution for understanding audiophilia.

I'm convinced that the only explanation for some of the ideas, designs and performance of some high end product is LSD.

 

[digicidal]

I'm convinced that the only explanation for some of the ideas, designs and performance of some high end product is LSD.

Maybe that's the secret code in MQA orgami... M-1=L, Q+2=S, A+3=D... First fold is lower rate (-1), Second is higher (2) and Third is highest (3). Now if I only had a bunch of string, photos, and pushpins I could map out everything and expose the conspiracy to the world!