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

[KeithPhantom]

Nevermind, found a good explanation. Thanks though.

 

[mhardy6647]

Cynicism is like entropy. It can never be reduced, only redistributed.

Worse than that: If it's really like entropy -- then it must always increase, eh?

 

[mansr]

Worse than that: If it's really like entropy -- then it must always increase, eh?

In total, yes, but not necessarily uniformly.

 

[John Atkinson]

This is from page 2 of Keith Howard's 2006 article in Stereophile in your link. It showed a complete misunderstanding on the sampling theory (i.e. how band-limiting works). Try band-limit an impulse to 22.05 kHz, and look at it in an oscilloscope. (Or look at the impulse response of an anti-alias filter.) When sampled, it will not give you only one single pulse. When you have only one single pulse in the digital samples, it means and it is the sinc waveform.

Of course. I have not said otherwise and neither did Keith Howard. As you write, the single high sample is "mapping" the coefficients of the filter. In my 2018 article I showed how three A/D converters with different antialiasing filters coped with a shaped analog impulse that had content up to 60kHz and rolled off above that frequency.

John Atkinson
Technical Editor, Stereophile

 

[Eskamobob1]

"Quality" is a subjective thing. And application is clearly music, not test signals.

Wait, what? What is the point of this site again? Because I sure as hell thought it was objective measurement of the quality of audio equipment to weed out snakeoil in the audio industry. Should I just go start reading $2000 power cable reviews if subjectivity is what matters, not objective measurements of marketing claims? Because that is the exact line that power cable selling scam artists use to justify their products

 

[Raindog123]

Analog [bandpass] filters are used in RF communication systems all the time. But their role there is to assist the frequency up (down) conversion - to filter out “wrong“ images (aliases) and reveal the “good, used” ones. And yes, to still filter out the out-of-band noise. Often at the frequencies well above the max frequency of the DAC/DSP and sometimes in multiple stages... Audio digitalization/restoration does not have the baseband frequency upconversion need thus problem.

 

[sterkoff]

Wait, what? What is the point of this site again? Because I sure as hell thought it was objective measurement of the quality of audio equipment to weed out snakeoil in the audio industry. Should I just go start reading $2000 power cable reviews if subjectivity is what matters, not objective measurements of marketing claims? Because that is the exact line that power cable selling scam artists use to justify their products

'Quality' is a subjective term as used by amirm in that example. The actual definition of 'quality' includes things like it has a 'peculiar or essential character' or a 'like social status'. Nothing in the definition of 'quality' would lead me to believe that it denotes any objectively definable trait. To me, quality is a personally identifiable term that is better applied to a product after all measurements have been done and I can then use that data to determine the level of 'quality' I perceive a product to have based on all my (edit: pre-determined parameters). I would say that this website is about "the objective measurement of audio equipment'. The rest is up to you...

 

[Raindog123]

Quality' is...

For formally-trained engineers, “quality“ is a set of metrics capturing how a product meets its requirements. Like in “quality control” or “quality assurance”. And those metrics can be, eg, performance, reliability or longevity; but they are always tied to a set of predefined, pre-agreed (!) “requirements”.

 

[Eskamobob1]

For formally-trained engineers, “quality“ is a set of metrics capturing how a product meets its requirements. Like in “quality control” or “quality assurance”. And those metrics can be, eg, performance, reliability or longevity; but they are always tied to a set of predefined, pre-agreed (!) “requirements”.

Lol. Thank you. While how quality is measured may have some subjective traits to it in how you weight things, quality its self is not subjective in engineering. "Does it pass these metrics? No? Well it does not meet quality standards"

 

[ezra_s]

Wow, apologies, I hadn't seen this awesome thread and I posted the video in a new thread. Thanks to the moderators and apologies for double-posting! (I searched for MQA but didn't find it) Now on to read the whole thread!

And Thanks @GoldenOne for the video, it was very interesting and revealing.

 

[mieswall]

It is true that filter implementations measure differently in the audible band.
The answer is to master in HD-Audio, high bitrate and bit-depth. Then optionally deliver at 44.1.

If HD-Audio sound better leave it alone.
Don't lower the sample rate, lower the dynamic range, apply a poor digital filter, and turn on a little blue light.
Many audiophiles chose different filters, SOX and such, who is MQA to limit this choice, especially, when their filter measures poorly.
Slow filters are not phase coherent so it is a bit rich to force the very problem, MQA proports to fix.
I am sorry, there is sufficient evidence that conclude, the MQA is predatory and lies to make money.
Pointing to other truths does not lend veracity to their huckstering.

If Audiophiles are access to both an MQA (down-sampled master) and the actual HD-Audio master, which one would they chose?
For an answer, I went to HDTracks, I found 3 tracks that seem to be samples.
This sometimes offers multiple resolutions and different price points, but for some reason MQA is not one.

- Rich

Thank you for answering, interesting. I'm using 'lowpass filters' to leave aside those judgments, a bit premature before clearing some issues first.

- Masters in HD-audio. That's right, I agree, and I understand that's exactly what they do when processing historical analog masters. When original masters are already in digital HD, MQA does a different process (although they say better results are achieved from analog).

- They don't lower the sample rate, they just package it with (only the) significant information captured at oversampling stored under the noise floor of the first 44 or 48K (that's why by definition a orthodox "lossyness" test is flawed, since it is trying to compare pears with apples: two completely different noise-floor bands, among other things). When you unfold the file, you restore that higher sample rate. Generally 96K o to 192K, I've been listening some of 352K.

- About lowering the dynamic range: in fact they use something that resemble (or perhaps is) a variable bit depth of progressive lower bitrate the higher the frequency. I tried to explain the reason in previous posts: at higher frequencies music has much lower amplitudes (Meridian shows this in sets of statistic analysis of real music, although the same could be inferred from the amplitude patterns of harmonics of every instrument). If you still use 24bit there, you are wasting storage space (that space is instead used for the folds mentioned before) . The fact that the square wave test of GoldenEar shows those big ripples anomalities confirm this reduced bitdepth in highs almost to perfection; otherwise, the only other possible explanation would be a disastrous frequency response, that I doubt any professional audio engineer would do.
Unless, of course, your signal is full of white noise, square waves, etc, that, with very few exceptions, has no resemblance with real music which was the initial premise. Again, two other reasons why those tests are flawed: a) not intended (at least in the automated process for file uploads that GoldenEar used -surely studio processing is more sophisticated-) to process information outside the threshold of music; b) that space (probably void in PCM) is instead replaced with info from upper folds that were not in the same place in the input file. No way, then, you can match the input and output trying to find "losslessness" in them.

- Why an MQA instead of a plain HD audio file? Because the first one is compact enough to be streamed easily (we may question if in 2021 this is still necessary, but that's another problem). And because MQA has processed the HD file for a much better impulse response. John Atkinson has just posted a link of a Stereophile article about this. There are some other explanations about the differences of a MQA file and a plain HD file of the same sampling that can be found in MQA papers and articles, but I don't understand them well enough to talk about.

 

[RichB]

- Masters in HD-audio. That's right, I agree, and I understand that's exactly what they do when processing historical analog masters. When original masters are already in digital HD, MQA does a different process (although they say better results are achieved from analog).

MQA may start with an digital HD master and serious people are not buying their improvement.

- They don't lower the sample rate, they just package it with (only the) significant information captured at oversampling stored under the noise floor of the first 44 or 48K (that's why by definition a orthodox "lossyness" test is flawed, since it is trying to compare pears with apples: two completely different noise-floor bands, among other things). When you unfold the file, you restore that higher sample rate. Generally 96K o to 192K, I've been listening some of 352K.

This is mostly false. A LPCM of FLAC lossless file has a sample rate.
In this area, MQA is not entitled to their own truth.

- About lowering the dynamic range: in fact they use something that resemble (or perhaps is) a variable bit depth of progressive lower bitrate the higher the frequency. I tried to explain the reason in previous posts: at higher frequencies music has much lower amplitudes (Meridian shows this in sets of statistic analysis of real music, although the same could be inferred from the amplitude patterns of harmonics of every instrument). If you still use 24bit there, you are wasting storage space (that space is instead used for the folds mentioned before) . The fact that the square wave test of GoldenEar shows those big ripples anomalities confirm this reduced bitdepth in highs almost to perfection; otherwise, the only other possible explanation would be a disastrous frequency response, that I doubt any professional audio engineer would do.
Unless, of course, your signal is full of white noise, square waves, etc, that, with very few exceptions, has no resemblance with real music which was the initial premise. Again, two other reasons why those tests are flawed: a) not intended (at least in the automated process for file uploads that GoldenEar used -surely studio processing is more sophisticated-) to process information outside the threshold of music; b) that space (probably void in PCM) is instead replaced with info from upper folds that were not in the same place in the input file. No way, then, you can match the input and output trying to find "losslessness" in them.

Nope, a file does not have a variable bit rate, no matter how hands are waved.

- Why an MQA instead of a plain HD audio file? Because the first one is compact enough to be streamed easily (we may question if in 2021 this is still necessary, but that's another problem). And because MQA has processed the HD file for a much better impulse response. John Atkinson has just posted a link of a Stereophile article about this. There are some other explanations about the differences of a MQA file and a plain HD file of the same sampling that can be found in MQA papers and articles, but I don't understand them well enough to talk about.

If this was an issue for most, there would be no Netflix, HBO Max, etc.
Archimago published some data where a 96kHz/18 bit sample achieves similar file sizes without any of the MQA deleterious effects.
I am sure most most customers would pick native HD-Audio over MQA when given the choice.

- Rich

 

[RichB]

MQA in Broadcast | Bob Talks ... and tails wag

In all areas of sound, the key benefits of MQA include:

Efficient modern coding with a clearer sound than normal digital using matching deblur technology in the encoder and decoder.

Clearly mostly false, since multi-track recordings may not be using the same encoding.
Encoded audio files are always decoded by matching decoding.
MQA attempts to make proprietary currently open HD-Audio formats.

The ability to transmit the full gamut of high-resolution sound with no wastage of data, that is, in a smaller right-sized stream.

I'd hate to fill my landfills with wasted data. There are cases where 44.1/16 have been MQA encoded creating wrong-sized data.

A flexible and consumer-friendly ‘last mile’ architecture for listening at home or on the go.

This is a key feature? I am not sure this is even a coherent statement.
What the hell is "last mile" technology?

Authentication – an MQA light or indicator to confirm that the listener has the exact result intended; that the audio had not been modified on the journey to the listener.

Neil Young does not agree. He demanded that Tidal remove his work precisely because they had altered his masters.

Provenance – the additional Blue light indicator is very helpful for artists, engineers and labels to confirm that the sound uses the highest quality source available.

This little blue light indicated that up to 70% of the file is intact.
I have downloaded numerous tracks from HDTracks, no bits are missing or altered.

- Rich

 

[MarioBoi]

DMS (from Abyss headphones) seems to imply that Amir has changed his tune on this topic because the MQA bashing negatively impacts the companies that (allegedly) pay him.

https://twitter.com/i/web/status/1384569886081536005

Edit: It seems crinacle also shares the same view as he retweeted this

 

[BinkieHuckerback]

The reason not to bother with 'Twatter'.

 

[ezra_s]

DMS (from Abyss headphones) seems to imply that Amir has changed his tune on this topic because the MQA bashing negatively impacts the companies that (allegedly) pay him.

https://twitter.com/i/web/status/1384569886081536005

If he usually reasons like that to simple and easy to understand sentences.. that tells much about what can you expect from the guy.

 

[LoyK]

I want to thank @GoldenOne for this really awesome work!

Some thoughts from my side on MQA:

• When I look at some of the thread here for DACs with MQA support it is really striking how much confusion this format brings:
  • People don´t know if they are getting all the unfolds
  • People don´t know what connection they can use
  • People don´t know which version they are listening to (different samplerates displayed etc.)
  • People don´t know that EQ/DSP breaks the second unfold
• It is just overcomplicating things and I don´t see any reason for it
  • Audio should be easy and fun
• DSP tailored to your speakers/room/headphones will always sound better than that built-in upsampling
• Some lies of their marketing like knowing the ADC for recording and then using the inverse for decoding are just...no comment...
• What @GoldenOne said.

 

[mkawa]

DMS (from Abyss headphones) seems to imply that Amir has changed his tune on this topic because the MQA bashing negatively impacts the companies that (allegedly) pay him.

https://twitter.com/i/web/status/1384569886081536005

sore behinds from getting their poorly measuring headphones measured well.

 

[Chrispy]

DMS (from Abyss headphones) seems to imply that Amir has changed his tune on this topic because the MQA bashing negatively impacts the companies that (allegedly) pay him.

https://twitter.com/i/web/status/1384569886081536005

Maybe he just forgot the T on rolling....

 

[RayDunzl]

I liked satellite radio when I'm on the road,didn't have to look for stations every 30 miles,just set it and forget it.

I liked local stations when I was on the road.

Didn't get the same canned stuff from two stations a thousand miles apart.