Any one with experience of MQA CD

MQA CDs don't adhere to the Red Book standard, the MQA data is indeed dithered into the last 3 LSB...

 

OK, then it makes sense. I thought they comply with Red Book. So, I assume, they don't have the Compact Disc logo?

 

They actually do.

 

There has never been any consumer audio format which has a better chance at sounding audibly superior to a regular CD in listening tests than flipping a coin. So far, MQA CD does not appear to break this trend.

 

Then I am confused. They must have two layers.

 

I don’t believe that’s the case.

I also don’t believe that merely displaying the CD logo means that they are in compliance with the Redbook Standard: I could be mistaken, but I’m fairly sure that those awful copy protected CDs that came out some time ago also used the logo despite not adhering to the Standard.

However, I don’t have one in hand to verify so I may be misremembering.

 

I'm not sure that violates the Red Book standard. HDCD dithered extra dynamic range into 16-bit PCM and still carried the Compact Disc logo.

 

At least we know it won't cause any problems playing in any cd players so yea I guess you can display the logo, but the last 3 bits are actually data arranged as a dither that a regular PCM player will simply interpret as very low noise.

 

Thank you for taking the time and effort to do this test and this post. It represents a real act of kindness to yours friends on the Hoffman Forum. I am very impressed.

 

Thanks for the positive responses.

In fact, my friends and I always have a blast doing this sort of thing. Getting them to cooperate with straight faces is actually nearly impossible mind you, especially when Scotch and relentless kibbitzing are the general operating modes. Eventually, everybody settles down for long enough to make listening notes or selections for whatever it is I’m poking at.

So basically . . . we’ll use almost any excuse - picking up a new piece of gear (e.g., a MyTek Brooklyn DAC+ in this case), picking up more music (e.g., the UHQ/MQA CDs from CDJapan) - as a reason to get together. Life is short. Might as well examine it along the way!

 

Just joining in on the chorus as others - great work Agitater!

This post demands more than attention on the forum! With permission, it's now also online on my blog with some context for readers:
Archimago's Musings: "MQA-CD x UHQCD" Listening Test by Agitater.

Indeed, getting guys together for listening sessions such as these really adds to the quality of an audiophile's life .

Life's absolutely way too short...

 

I worked as a tour guide in Northwestern Spain in the 90's while my home was (and still is) in Southern Spain. Mini Discs were a blessing to me, I could carry so much music with me on such small discs I could record myself from my original CD's, man, that was really something 25 years ago long before FLAC and mini SD cards.

 

And yet they were never telling the buying public that it had 1/5 the audio data and garbled underwater sound worse than mp3. The recordable MO disc was not their main hope, they wanted prerecorded stamped Minidiscs to be common. "It's digital". Same for Digital Compact Cassette, which used 4:1 primitive lossy compression. The ignorance of us before the internet.

Before the internet could communicate marketing mistruths:

 

That kind of sound quality you're talking about only happened with the first generations, yes, it was said by Sony that it used data compression that reduced the original 44.1/16 data to 1/5th of the original data, that data compression was called ATRAC, that means Adaptative Transform Acoustic Coding. By 1996 all the sound issues with ATRAC were sorted out. Ever went to a cinema that showed movies on Sony Dinamic Digital Sound? Did it sound bad? SDDS used ATRAC compression at the same data rate (per two channels) as Mini Disc.

 

DCC's sound quality also improved a lot on the later generation machines. MiniDisc and DCC were both rushed to market before they got all the bugs worked out because cassette tape sales had started to decline and they were both desperate to become the replacement for it. But DCC and MiniDisc ended up competing more with each other than with the cassette, which proved itself to still have a lot of life left.

 

Been reading some conflicting information on this. Is MQA data using 3 bits or only 1 bit on an MQA-CD?

Also, has anyone captured a recording of a decoded MQA CD to see what the spectrum looks like?

 

I don't think there is a clear answer yet.

Technically, I have not seen an MQA CD spectrum. It at least has to lose 1 bit just to cover the blue/green light, show the original recording rate, and pick a filter. If there is any other high frequency information it would lose more bits. So at best it is 15/44, decoded or undecoded. at worst it could be 12-13 bits undecoded and 13 bits decoded.

Technically, if high definition is defined as the bit depth, it fails being at best, less than CD.

On the high res front:
@Archimago did a test a year ago. Look him up.
CA is trying to a test now.

There are question in both tests whether they are valid or not, or if something else is causing a difference. Continued discussion at CA.

For what it's worth, I was 99% confident on one of Archimago's samples. The rest were a coin flip for me.
In the current CA test, (try it yourself) I'm 100% confident that there is an audible difference between the two test tracks. I'm still unclear on which track is which but hear a definite difference between them and have a preference. I don't know if that preference is toward MQA or not.

Sorry it doesn't answer your question but I don't think there is consensus, outside of audio rags masquerading as journalists. My advice is to ignore them and make up your own mind.

 

^^^ An eye-opening update from Audiophile Style (formerly Computer Audiophile) forum on decoded MQA-CD HF spectral content:

MQA is Vaporware

 

I'm not a pro either at audio engineering or digital signal processing but I don't see possible to recover frequencies above the 22.05 Khz that Red Book allows and this comes from a person that back in the day read about how HDCD worked, understood the process and didn't see it as snake oil. I later checked out the Independence Day soundtrack that is HDCD encoded and uses all its features and heard an actual improvement over the undecoded CD.

 

That spectrum graph looks a bit like Pioneer was doing with their "Legato Link Conversion" back in the '90s -- allowing some aliasing above Nyquist to get through the lowpass filter in the DAC, with the theory that the end result will have spectral content more similar to the original analog audio source than a DAC with a strict "brickwall" filter:

 

The decoded MQA-CD spectrum is right there in the post I referred to, BTW... According to Mans all MQA-CD 16/44 audio he looked at so far had either 1 or 2 bits utilized by MQA.

 

Translation, MQA CD is usually 15/44.1 or 14/44.1. The max it can be 15 bit and technically as low as 8 bit.
Decoding doesn't bring back any bits of resolution. It will still be 14 or 15 bits of resolution but with a higher frequency response, a guessed number (by MQA) out to 88khz or 96khz sampling frequency, which we see by the graphs, isn't necessarily a good thing.

 

I've always thought that MQA had perceptual encoding tricks built in to simulate actual Hi-Rez sound quality. Basically a lossy encoding with "fewer" gaps in the audible spectrum. It's clear to me a lot of thought has been put into the format, even if it isn't audio nirvana and the greatest thing since sliced bread.

 

Nor mine...

 

I remember this Pioneer's technology from the mid 1990's but I always thought it would do some kind of interpolation or a guess of how armonics would be beyond 22.05 Khz. There's one thing I don't understand, if at the A/D stage an anti-aliasing filter is used to block all frequencies over 22.05 Khz prior to actual conversion then the resulting digital signal doesn't have any frequency content above 22.05 Khz, if so, why filter it again at the D/A stage? This is something I could never understand well.