MQA Deep Dive - I published music on tidal to test MQA

[earlevel]

They are *very* far from music. No one in this thread or elsewhere has produced a single piece of music that is similar to spectrum of those test tones or even close. The last sample was produced that sounded noise like still followed an exponential drop off and is unlikely to have much in ultrasonic range:

View attachment 133181

Remember, you can NOT play OP's test signal on your system without potentially damaging it. Your tweeter, nor your ears can handle high amplitude high frequencies. And there would be no reason for an artist to generate artificial sounds that have ultrasonic content let alone extreme amount of it. All music is listened to by humans which guarantees you that kind of drop off. I can't imagine anyone releasing content that instantly blows up the tweeter in your system, causes the amp to oscillate and you running away with a scream!

Sure, can someone do it on purpose like OP did as a prank? Yes. In that case what MQA does to it doesn't matter. The content itself is not high fidelity and no one would care one bit what the noise level underneath all of that "signal" would be.

OK, I may not have the time to come up with a song with naked white noise, but there are such songs, and why shouldn't there be. I can come up with relaxation-type audio, but we could say that isn't the material targeted by MQA. And if I came up with a few seconds of white noise in a song, we could argue that it wasn't sampled 96k and therefore doesn't have bandwidth to 44k, and that would probably be true due to the huge volume of CD-quality audio available. And we could go through the same thing with square waves—maybe there is some reverb, maybe it's not a forbidden level or frequency that makes MQA fail. There are just too many arguments as to why whatever material I come up with doesn't fit the "forbidden" status.

But that is my point. what are these levels, what is the criteria? To be clear, I'm not saying this is an utter failure of MQA, I'm saying this is something I don't know, something that makes me wonder. I don't know the criteria at which MQA is "perfectly well expected" to fail. Do you?

I'm not firing shots at MQA. I don't see how it succeeds, against streaming hi-res, considering it's significantly more intrusive, but I don't hate it—I just don't think its compression justifies the baggage. The fact it can't encode some types of audio signals doesn't help that feeling. But it does make me curious where the bounds are. Noise is used a lot in music. The fact that it most often slopes downward in spectrum—sorry, but saying that does not alleviate my concerns. I've certainly used unfiltered analog white noise blasts in music before—at what point does MQA fail? I don't think any of us have any idea. Why shouldn't I want to know?

(I have to make a little face to convey that I mean this in a friendly way, as this thread is a little overly contentious—I do appreciate that you're hanging in there and making your points )

 

[amirm]

Here we have a picture of the peak levels represented by an orange line with a slope of -1dB/kHz. According to it the max. allowed level at 24 kHz would be -30 dB. This clearly exceeds the allowed MQA level.

So does natural music already hurt the MQA conditions?

I am not following your argument. What natural music? Do you have a spectrum to share that shows only 30 dB reduction by ultrasonics?

 

[Mulder]

I can't follow your writing. If you have an MQA DAC, you can play normal non-MQA content on it so there is no limitation there. And you can use any player you want.

The amount of MQA, non-Tidal content is so small that it doesn't matter. There are only 600 CDs published in Japan. You can rip those and play them on any DAC or any player. Or you you can use Roon to decode MQA layer in them.

So there is nothing "practical" here. You are making a case out of exception out of exception to argue something.

I am talking about the combination of DSP and a MQA DAC. If you want to use DSP and take the full advantage of MQA, then ROON is the only option. I can’t use Jriver for example, or somerhing else. Exeption or not. Those who owns and wants to utilise a MQA DAC have to use ROON. If they also want to use DSP. There are no alternatives. This conclusion ought not to be difficult to follow. Then if it in your opinion do not matter because there are so few MQA coded records to download, then this is just your opinion. As it is my opinion that it matters.

 

[levimax]

I’m flabbergasted that MQA is called “broken” (or worse) in this thread. MQA sounds just fine. Anyone thinking that MQA sounds bad should listen to Blue Maqams by Anouar Brahem/Dave Holland/DeJohnette/Django Bates in MQA, on Tidal, and explain exactly how it’s “broken”. It’s not. It sounds stunning.

I don't think most people, even anti-MQA, go along with "broken" and I have no issue with MQA sound quality. My issue is even though MQA "sounds stunning" so does a hi-res FLAC file and you don't have to pay money and jump through hoops to play or EQ them.

 

[DimitryZ]

Generally I’m in full agreement with you and have defended most of your points (until bailing out due to thread SNR dropping through the floor), but I have a very small quibble here.

When I was a Tidal customer, I used my DAC’s streaming app (my DAC is MQA enabled and its streaming app can be given your login details, and then you can instruct it to pull an audio stream directly from Tidal or Qobuz) and I got the ”total solution”, as you call it, for MQA decoding without Roon or the Tidal app.

I’m flabbergasted that MQA is called “broken” (or worse) in this thread. MQA sounds just fine. Anyone thinking that MQA sounds bad should listen to Blue Maqams by Anouar Brahem/Dave Holland/DeJohnette/Django Bates in MQA, on Tidal, and explain exactly how it’s “broken”. It’s not. It sounds stunning.

I was listening to it yesterday! Beautiful album!

 

[symphara]

Could you use room EQ with this set up?

Of course, if I had a Room EQ gadget such as one of the Lyngdorf or high-end NAD amplifiers. I could even put the streamer into my AV receiver and run it through Audyssey.

 

[lucretius]

As I have explained multiple times, there are only two mainstream players that consume Tidal content with MQA and one is Tidal which doesn't have DSP, and the other is Roon that does

Audirvana 3.5 and Audirvana Studio also "consume" Tidal content and decode MQA (1st unfold). Similarly, USB Audio Player Pro (for Android) does the same.

 

[Raindog123]

The last sample was produced that sounded noise like still followed an exponential drop off and is unlikely to have much in ultrasonic range:

Remember, you can NOT play OP's test signal on your system without potentially damaging it. Your tweeter, nor your ears can handle high amplitude high frequencies. And there would be no reason for an artist to generate artificial sounds that have ultrasonic content let alone extreme amount of it. All music is listened to by humans which guarantees you that kind of drop off.

I still do not buy it… Forget OP or ultrasonics for a moment... But here is a nice constant-level 20-20k sweep…. Unpleasant, maybe. But still on Tidal:

https://tidal.com/track/29581922

 

[symphara]

I don't think most people, even anti-MQA, go along with "broken" and I have no issue with MQA sound quality. My issue is even though MQA "sounds stunning" so does a hi-res FLAC file and you don't have to pay money and jump through hoops to play or EQ them.

I didn’t pay anything extra for MQA, nor do I jump though hoops to play MQA. Playing MQA , PCM or DSD is identical through my streamer interface. If I wanted to EQ, having MQA would make no difference to what I’d need to do.

 

[amirm]

But that is my point. what are these levels, what is the criteria?

What is the criteria for design of this speaker and a million like it?

Isn't that the high frequencies are played at much lower levels and if you push them overwise, you will damage it?

Statistical solutions are very helpful in designing workable systems. Trying to make a solution 100% massively increases cost which is the problem with FLAC encoding high sample rate/high-res content. I have tracks that are nearly one gigabyte in size for just one song!

This is the cost of "encode anything."

No lossy encoder would exist ever if we applied the same standard. Corner cases can be terrible sounding. But to the extent that 99.999% of content does well, the job is done. If you want that last 0.001% then, the solution is not for you can seek out options like above.

 

[levimax]

Of course, if I had a Room EQ gadget such as one of the Lyngdorf or high-end NAD amplifiers. I could even put the streamer into my AV receiver and run it through Audyssey.

I don't think so, the only way to get "full MQA decode" is to use an MQA DAC and Roon (which cost over $100 per year) otherwise you are only getting the "first unfold".

 

[symphara]

I am talking about the combination of DSP and a MQA DAC. If you want to use DSP and take the full advantage of MQA, then ROON is the only option. I can’t use Jriver for example, or somerhing else. Exeption or not. Those who owns and wants to utilise a MQA DAC have to use ROON. If they also want to use DSP. There are no alternatives. This conclusion ought not to be difficult to follow. Then if it in your opinion do not matter because there are so few MQA coded records to download, then this is just your opinion. As it is my opinion that it matters.

But that’s not an issue with MQA per se, but with the DACs. Since MQA is a niche format, DACs barely support it, and few DACs support DSP/EQ, so it’s no surprise that you cannot find an MQA-capable DAC that can also DSP. However the solution DOES exist. You‘re of course free to add whatever DSP you want through an additional ADC/DAC step.

 

[amirm]

I still do not buy it… Forger OP for a moment, or ultrasonics... But here is a nice constant-level 20-20k sweep…. Unpleasant, maybe. But still on Tidal:

https://tidal.com/track/2958192

That link doesn't open. And the fact that it is 20 to 20K answers your question. It has no ultrasonics so MQA can just pass it through with zero trouble.

 

[symphara]

I don't think so, the only way to get "full MQA decode" is to use an MQA DAC and Roon (which cost over $100 per year) otherwise you are only getting the "first unfold".

I don’t believe that’s true, but I’m willing to be schooled on the issue. I think that my streamer/DAC (Teac NT-505) supports full MQA decoding.

 

[Brab]

What ARE the criteria.

 

[amirm]

Audirvana 3.5 and Audirvana Studio also "consume" Tidal content and decode MQA (1st unfold). Similarly, USB Audio Player Pro (for Android) does the same.

I specifically bolded mainstream. Did you not notice that? Regardless, what Roon has done with MQA+DSP, they can too. I am not a user of those so don't know if they do DSP at all for this to matter.

 

[amirm]

I don't think so, the only way to get "full MQA decode" is to use an MQA DAC and Roon (which cost over $100 per year) otherwise you are only getting the "first unfold".

Not at all. You can send the digital bitstream however you like to an MQA DAC and it will decode and play it. Roon is not needed.

 

[Raindog123]

I think bandsplitting is a very common technique in signal processing.

You are correct… It is common, as soon as it does not violate Shannon Limit - of only that much information fitting in that much BW…

But that’s not the point. My point is that such band-splitting of a PCM waveform, when based on the frequency band/bin dynamic range - while doable - is definitely beyond your average ‘let me draw you a triangle’ hand-waiving.

 

[UliBru]

I am not following your argument. What natural music? Do you have a spectrum to share that shows only 30 dB reduction by ultrasonics?

Bob Stuart's Hierarchical Approach AES paper states in chapter 3.1:

3.1 Spectral Content of Music
There is significant content above 20 kHz in many types of music, as an analysis of high-rate recordings summarized in Fig. 7 has revealed. One notable and common characteristic of musical instrument spectra is that the power declines, often significantly, with rising frequency.

And Fig.7 is titled by

Fig. 7. Peak spectral level gathered over a corpus of 96- and192-kHz recordings that have not clipped in the digital domain

So I do not know which music has been used for this picture but anyway the description sound logical (e.g. not clipped = no squarewave ). And I've learnt in this long thread that such papers are peer reviewed. So why shall I doubt?

 

[levimax]

I don’t believe that’s true, but I’m willing to be schooled on the issue. I think that my streamer/DAC (Teac NT-505) supports full MQA decoding.

I am not sure which streamer does the "full unfold" or not but the problem is EQ. You can't EQ before you send information to your MQA DAC because it changes the "bits" and the light won't come on and you won't get the "final unfold". Roon cut a deal with MQA where you can apply EQ to the MQA stream and then Roon "reauthorizes" the stream so the MQA light comes on and you get the "final unfold". If you don't care about the "final unfold" then it doesn't matter.