If you are contacting MQA, I don’t think I’d do it by accusing them of that list of “lies”! If you contact them, be respectful and ask questions that can actually be answered. All of the things in your list have been discussed, and answered, many times already, and most of them have a long history of internet trolling.
I don’t feel like looking up the plot with the gap but the topic’s been discussed already and there was a plain answer for it.
Well I see it very different. Since all MQA files (all 24 bit) i have compared to either a 16/44 or an actual HiRes file has lost perspective, has skewed the tone balance. And worst of all MQA has a sound of its own, I am never in doubt of whether or not I’m listening to MQA encoded music.
The Thing about improving the master is by definition impossible, since it is the master. They can present it differently (which they do, but not as the artist intended) They claim to have lossless dynamic range of 17 bits (Bob’s words, not mine) sadly the embedding of info into the LSB’s raises the noise level by 12 dB. Meaning the calculation looks like this 17 bit dynamic range 102,34 dB - 12dB = 90,34 which is less than a CD (96 dB) Even the lowly MP3 only looses a few dB of dynamic range.
Chris…That van has got to go 
GAPGATE
Please tell me where I can find this plain answer. Since I’ve never seen it
Once again. 17b with noise shaping gets you an actual dynamic range of about 20b. MQA say they do not discuss bits because they always measure the dynamic range of the recording as part of the encoding, and set up their processing to accommodate it. So the final MQA has the same dynamic range (audible) as the original PCM would.
No, you do not lose 12 dB by embedding data into the LSB. This is a complete lack of understanding on your part of burying data into pseudo random noise.
[Måns Rullgård]
Note that the frequencies above the gap are a mirror image of those below it with a slight downward slant. This is because the weak upsampling filter used by MQA lets these images through rather than suppress them as it ought to. Nothing above that gap actually existed in the recording. It is all fake. The 2L recording originally has a high sample rate, and MQA preserves it reasonably well up to 44.1 kHz. If you did a playback and (analogue) capture through a certified MQA DAC at 176.4 kHz (or higher), you’d see similar mirroring in the 44.1-88.2 kHz range.
Not getting into discussions on measurements and their interpretations without accurate details about the measurement, source material, and methodology.
[Moderated]
The Nyquist “gap” is specific to 44.1/48 kHz MQA. This is due to the 88.2/96 kHz upsampling, since the MQA decoder always outputs at those rates.
AJ
Well the (let’s put this nicely) storytelling from MQA is astounding. What gets me chuckling is the fact that it seems that none of the MQA evangelists here seem to question the fact that loads of independent sources again and again debunk MQA’s claims.
[Moderated]
MQA say they do not discuss bits because they always measure the dynamic range of the recording as part of the encoding…really
No officer I couldn’t have been speeding because my average speed on this 200 mile journey is only 30 miles per hour 
If 17 bits with noise shaping gives You about 120 dB Dyn.Range Then 20 bits should give You a Dyn. Range of 141. 24 bits is then 169 ? Let me remind You that noise shaping is used in PCM/FLAC files as well. So then we’re back at the fact that 17 bits (MQA) is the same as 17 bits (PCM) except for the measured 12 dB penalty from the embedded data.
and they have to look at the dynamic range of the final format as well
It determines how lossy their compression should be.
This is very noticeable with mqa-cd.
Let’s start from a 24/352.8 source : After skipping 3/4 samples, they end up with a 24/88.2 pcm
And no matter what… version, the 16/44 or the 24/44 … “both mqas are lossless” according to the marketing boys 
Then why are there 2 possible versions of the same 24/352.8mqa? Why not make them all 16/44 and save even more bandwidth than a 24/44?
I compared 16/352.8 with 24/352.8 mqa and the 16 bit is audibly worse (of course it is).
It’s just impossible to start discussing with people that are unwilling to admit mqa is lossy, even when you show them proof like this that is undeniably showing that mqa is adjusting the compression rate of its encoder. Storing the same amount of hi-res data compressed in a) 7 bits or in b) 3 bits
and downscaling from 24bits to 17bits or even 13bits (half) to make room for that proves mqa IS lossy.
Where did you find 16/352.8 MQA? There are only 2-3 small labels I’m aware of doing 352.8 kHz and they are all 24b? How does a 16/352.8 MQA file decode? Or are you taking files from MQA-CD physical discs, in which case I don’t know how they are labelled since I have no experience with them.
Right. The mastering engineer comes in long after the artist(s) has left.
MQA-CD disks, which are 16/44.1, simply drop the ultrasonic frequencies, whereas, for regular MQA (24/48), the ultrasonic frequencies are separated and stuffed into 8 bits. The only thing these two different versions of MQA files/tracks have in common is that they both have the authentication control stream and both signal the MQA DAC to use the leaky MQA filters.
MQA-CD disks, which are 16/44.1, simply drop the ultrasonic frequencies,
You will need to read Bob Stuart’s discussion of MQA-CD disks at his blog. The description is admittedly vague, but makes it clear that they don’t simply drop the ultrasonic frequencies. Ultrasonic coding is definitely not as extensive as with high res on Tidal though.
for regular MQA (24/48), the ultrasonic frequencies are separated and stuffed into 8 bits.
I suggest reading their patents, there is considerably more to the encode/decode process, although you’re right that 24-48kHz ends up being packed into the low byte.
David Elias has released a MQA sampler in addition to his DSD music. This album has a couple of 352.8 MQA tracks.
with 16/352.8 mqa I mean a 24/352.8 mqa in a 16/44 wav file = mqa-cd
with 24/352.8 mqa I mean a 24/352.8 mqa in a 24/44 flac file = streaming mqa
It should have been clear reading my explanations above.
If it wasn’t it is now.
That’s interesting. I see the David Elliot MQA sampler is on Bandcamp, and the 352.8k is 24b. Great, the more the better. I didn’t know Bandcamp was hosting MQA although they have high res PCM.
No, it isn’t. You can’t have a 24/352.8 file in a 16/44 wav file because the bottom byte contains the encoded ultrasonic frequencies. Leaving out that byte makes the file undecodable. I’d still like to know where you’re getting a file like that since it’s not the released version.
For MQA cd the 24/352.8 is compressed into a 16/44 file and not into a 24/44 file.
Every single MQA-CD is 16 bit as is any normal CD.
I suggest you read the mqa-cd topic on Bob talks.
Have you seen their accounts , £10’s millions of loss against a few hundred thousand in revenue.
I’d hate to see your idea of failure…
I asked you earlier whether you were deriving files from MQA-CDs. Bob’s blog discussion says that MQA-CDs do some encoding of the ultrasonic region but not to the same extent as the full Tidal encodes. I don’t know how the dynamic range compares but he says that the encode/decode processing is different on MQA-CDs and that they handle it more efficiently. More information from them would be useful.
How are you doing the comparison? Have you ripped the MQA-CD files?
