• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

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

Status
Not open for further replies.

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,388
Location
Seattle Area
Bob Stuart's Hierarchical Approach AES paper states in chapter 3.1:

And Fig.7 is titled by

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?
Ah, now I see. I have to congratulate you as you got them there! The triangle as drawn can't encode that sample of music in Figure 7. Further, figure 7 is inconsistent with Figure 10 further down in the paper on statistics of 100,000 tracks (later updated to millions in Bob's blog). This doesn't help OP at all since even that worst case set in Fig 7 drops with 1/f but it is a hole in MQA saying that triangle is enough.

The only thing I can think of that would get them out of jail would be that triangle being dynamic and content dependent. It is hard to read that into Bob's blog or the paper so I am not going to give them the benefit of doubt.

I will have to see if there is a way to contact Bob and ask him about this.

Net, net, very good point and good eyes you have! :)
 

ebslo

Senior Member
Forum Donor
Joined
Jan 27, 2021
Messages
324
Likes
413
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.
Is the risk of potential damage from the OP's test signal you reference due solely to the ultrasonic content? Does "high frequencies" here refer only ultrasonics, or does your warning apply equally to the upper range of the audible band?
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,388
Location
Seattle Area
Sorry I left out EQ as that was what I was responding to. As far as I know if you want both EQ and "final unfold" the only solution is Roon and a MQA DAC.
I don't know why this keeps getting brought up. There are not that many design wins for MQA because it is a tiny fish. So you have fewer choices of MQA playback to start with and only one has DSP. This is perfectly fine for those of us who use Roon anyway. If it isn't, then MQA is not a solution for you, mostly because you can't play it in many other places anyway.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,368
Likes
234,388
Location
Seattle Area
Is the risk of potential damage from the OP's test signal you reference due solely to the ultrasonic content? Does "high frequencies" here refer only ultrasonics, or does your warning apply equally to the upper range of the audible band?
Upper range as well. I wear hearing protection in my speaker tests because the high frequencies can be so piercing in a linear sweep. The bass is no big deal.
 

ebslo

Senior Member
Forum Donor
Joined
Jan 27, 2021
Messages
324
Likes
413
Upper range as well. I wear hearing protection in my speaker tests because the high frequencies can be so piercing in a linear sweep. The bass is no big deal.
Ah, hearing protection. I should probably utilize that the next time I follow your Room Measurement Tutorial. Fortunately the tweeters have survived thus far.
 

voodooless

Grand Contributor
Forum Donor
Joined
Jun 16, 2020
Messages
10,226
Likes
17,805
Location
Netherlands
The only thing I can think of that would get them out of jail would be that triangle being dynamic and content dependent.

I’d actually bet that the triangle really only exists as a marketing and explanation tool, and that the encoder really doesn’t care about how high the amplitude is, as I’ve eluded to before. It can encode just so much music information and then starts leaving information out. What exactly that is remains a mystery until we can do some more tests.
 

voodooless

Grand Contributor
Forum Donor
Joined
Jun 16, 2020
Messages
10,226
Likes
17,805
Location
Netherlands
Upper range as well. I wear hearing protection in my speaker tests because the high frequencies can be so piercing in a linear sweep. The bass is no big deal.

Note that within limits for most conventional tweeters there is little damage to be feared. Impedance above 20 kHz rises quickly, so power falls accordingly. The more exotic designs such as planars are probably more susceptible since they often have a flat impedance curve. I’d still be careful though.
 

UliBru

Active Member
Technical Expert
Joined
Jul 10, 2019
Messages
123
Likes
337
I’d actually bet that the triangle really only exists as a marketing and explanation tool, and that the encoder really doesn’t care about how high the amplitude is, as I’ve eluded to before. It can encode just so much information and then starts leaving data out. What exactly that is remains a mystery until we can do some more tests.
What does that mean : MQA protects the Orange triangle very strictly.
??
 

earlevel

Addicted to Fun and Learning
Joined
Nov 18, 2020
Messages
545
Likes
776
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
I don't think this is a pertinent view. I'm well aware of the average spectrum, I post it myself earlier. But there are sections that are far different from the steep spectral rolloff you get with the long-term peak plot like you are showing. 20k is down only about 10-20 dB at high volume in much of the song, if you account for your speakers rolling the very lows or diverting to a sub.

Here's an arbitrary 50 seconds, to match your plot length, I just stretched a selection to get the time, some short term sections are worse:

Screen Shot 2021-06-01 at 2.16.28 PM.png


I know this is getting OT, but it's misleading to show that this song in particular in all it's ear-splitting grit, actually rolls off as if it's a typical Beatles tune. No, it really does have an unusual amount of HF trash, just like it sounds.
 
Last edited:

UliBru

Active Member
Technical Expert
Joined
Jul 10, 2019
Messages
123
Likes
337
Ah, now I see. I have to congratulate you as you got them there! The triangle as drawn can't encode that sample of music in Figure 7. Further, figure 7 is inconsistent with Figure 10 further down in the paper on statistics of 100,000 tracks (later updated to millions in Bob's blog). This doesn't help OP at all since even that worst case set in Fig 7 drops with 1/f but it is a hole in MQA saying that triangle is enough.
Thanks.

Caution: 1/f results in a falling line -6dB/octave or -20dB/decade in a chart with logarithmic frequency axis. The triangle picture uses a linear frequency axis. So the orange line is not 1/f !
 

UliBru

Active Member
Technical Expert
Joined
Jul 10, 2019
Messages
123
Likes
337
Another observation and conclusion:
MQA is THE solution against loudness war ;). Because heavily clipped tracks contain square waves which cannot be MQA encoded. So the mastering is forced to avoid clipping, otherwise the track may not find its way to Tidal.
 

raistlin65

Major Contributor
Forum Donor
Joined
Nov 13, 2019
Messages
2,279
Likes
3,421
Location
Grand Rapids, MI
Somewhat "open" version of iOS exists in the form of Android.

A version of iOS?

The Android kernel was built on Linux, not iOS. MacOSX was built on BSD UNIX. Neither of those share a code base. iOS is Unix-like, but was not based on either Linux or Unix code.

Or are you saying that because iOS came first? Even though there are some GUI and functionality differences which now iOS has incorporated from Android, which make them similar?

Just seems strangely over-simplistic to me. Like saying Windows 98 was Apple OS-like. But maybe that's the way you guys talked about Windows at Microsoft???

Don't see Apple users dumping it in favor of Android. They want and need what Apple provides to them, and close said eyes on the fact that it is the most closed, regressive computer operating system in the history of operating systems! Same with MQA. Folks who want and like it, will use it even though it is "closed."

Your previous whataboutism attack on this topic was also against me. So now you are talking only about other people? I have no idea how many people in this ASR discussion who have brought up the issue of closed formats are also Apple users. Do you? Or just wildly speculating? Maybe none of them are? And no. Sorry, since it won't fit your narrative. I'm not an Apple user.

But let's look at your false whataboutism equivalency. MQA doesn't really offer anything over FLAC for consumers. So it's hardly the same thing as someone using Apple who likes the full user experience and could value that. The only people who would be upset if Tidal replaced MQA with FLAC are the people who like the pretty light, or people who incorrectly believe there is an audible advantage. And MQA, of course.

Somehow this makes you very upset that we say such things. You said recently,

As to closed formats, you all consume it a million times a day so don't come complaining about that.

How are people complaining to you when they are criticizing MQA??? If you aren't trying to be an MQA proxy or are not otherwise invested in MQA, what's it to you? After all, you have previously stated MQA probably won't last. This shouldn't be that personal to you. If it is personal to you, then disclose your bias.

In my case, I use it because it came for free to me one day. I am not going to NOT use it "just because." Once in a while a file I play is MQA and that is that.

You have previously asserted more than once that MQA is free to you and users. Given how untenable that claim is to anyone who understands the licensing, at least you have backed off the free to users part.

But how do you get it for free when there are MQA licensing fees?

Since you haven't supported this claim, I'll just speculate.

1) You don't understand how the MQA encoding and decoding licensing costs could be passed onto consumers. This seems very unlikely that you would not understand.

2) You received some combination of an encoder, decoder, and/or MQA music files from MQA so that you could evaluate it during MQA's development. Or maybe you are part owner in MDA Limited once they split off from Merdian?

3)???? Please feel free to fill this in with a reasonable explanation to support your claims that MQA is free.

Or, you could admit that MQA is not free to consumers and quit spreading misinformation by implying it is/can be.
 

Dennis_FL

Addicted to Fun and Learning
Forum Donor
Joined
Feb 21, 2020
Messages
525
Likes
413
I don't know why this keeps getting brought up. There are not that many design wins for MQA because it is a tiny fish. So you have fewer choices of MQA playback to start with and only one has DSP. This is perfectly fine for those of us who use Roon anyway. If it isn't, then MQA is not a solution for you, mostly because you can't play it in many other places anyway.

I just tried playing the same classical track with Amazon UltraHigh, JRiver DSF, and Tidal Master. Interesting that when I use Roon, my DAC display has MQA and when I just use the Tidal APP, The display has PCM. Not sure what is happening…there is a Tidal APP toggle option to send MQA direct and it seems to be converting instead. I have to fiddle more.
 

DimitryZ

Addicted to Fun and Learning
Forum Donor
Joined
May 30, 2021
Messages
667
Likes
342
Location
Waltham, MA, USA
A version of iOS?

The Android kernel was built on Linux, not iOS. MacOSX was built on BSD UNIX. Neither of those share a code base. iOS is Unix-like, but was not based on either Linux or Unix code.

Or are you saying that because iOS came first? Even though there are some GUI and functionality differences which now iOS has incorporated from Android, which make them similar?

Just seems strangely over-simplistic to me. Like saying Windows 98 was Apple OS-like. But maybe that's the way you guys talked about Windows at Microsoft???



Your previous whataboutism attack on this topic was also against me. So now you are talking only about other people? I have no idea how many people in this ASR discussion who have brought up the issue of closed formats are also Apple users. Do you? Or just wildly speculating? Maybe none of them are? And no. Sorry, since it won't fit your narrative. I'm not an Apple user.

But let's look at your false whataboutism equivalency. MQA doesn't really offer anything over FLAC for consumers. So it's hardly the same thing as someone using Apple who likes the full user experience and could value that. The only people who would be upset if Tidal replaced MQA with FLAC are the people who like the pretty light, or people who incorrectly believe there is an audible advantage. And MQA, of course.

Somehow this makes you very upset that we say such things. You said recently,



How are people complaining to you when they are criticizing MQA??? If you aren't trying to be an MQA proxy or are not otherwise invested in MQA, what's it to you? After all, you have previously stated MQA probably won't last. This shouldn't be that personal to you. If it is personal to you, then disclose your bias.



You have previously asserted more than once that MQA is free to you and users. Given how untenable that claim is to anyone who understands the licensing, at least you have backed off the free to users part.

But how do you get it for free when there are MQA licensing fees?

Since you haven't supported this claim, I'll just speculate.

1) You don't understand how the MQA encoding and decoding licensing costs could be passed onto consumers. This seems very unlikely that you would not understand.

2) You received some combination of an encoder, decoder, and/or MQA music files from MQA so that you could evaluate it during MQA's development. Or maybe you are part owner in MDA Limited once they split off from Merdian?

3)???? Please feel free to fill this in with a reasonable explanation to support your claims that MQA is free.

Or, you could admit that MQA is not free to consumers and quit spreading misinformation by implying it is/can be.
I will answer only your statement about MQA not providing value for consumer. It certainly has provided lots of value to me over many years. And to many people like me.

If it didn't provide value it wouldn't be here.
 

UliBru

Active Member
Technical Expert
Joined
Jul 10, 2019
Messages
123
Likes
337
But we don’t know what it means. It’s just posed as fact. There is no elaboration, no explanation.. nothing.
Yes, it is posed as fact. And it is also implied in the argumentation that hurting the triangle requirement leads to a lossy encoding. So there must be a border (exact details unknown except the orange triangle). Staying within the border leads to transparency, crossing it (e.g. square wave, white noise ...) results in a lossy playback.

A clear specification would immediately result in the design of test signals close to the border.
 

earlevel

Addicted to Fun and Learning
Joined
Nov 18, 2020
Messages
545
Likes
776
Another observation and conclusion:
MQA is THE solution against loudness war ;). Because heavily clipped tracks contain square waves which cannot be MQA encoded. So the mastering is forced to avoid clipping, otherwise the track may not find its way to Tidal.
Well, people say this (square waves), but there is no reason for music to clip to square waves. Sine waves sure, but not music in general. Just pointing this out because 1) you can't make assumptions of the harmonic content of clipped music by thinking about square waves, and 2) you can't lump clipped music and square waves into the same issue that might be regarding the health of your speakers at a given volume, or encoding with a compression algorithm. And if the clipping is digital, you have aliasing so even with a simple input the spectrum becomes more complex.

As far as spectrum, not especially dangerous—people don't seem to fear sawtooth waves and the harmonics drop off at the same rate (just that square waves are missing the even harmonics).
 
Status
Not open for further replies.
Top Bottom