MQA - Hoisted on their own petard

JerryLove

JerryLove

Audioholic Samurai
How is digital tape 'analog?
The same way non-digital tape is analog.

They are both (essentially) identical media.

The difference is the logic used to write and read to them; and generally the physical read/write head.

the little magnetic paricles are aligned on a continuiously variable orientation in 360x360-degrees. Hence: analog.

Sampled audio accounts for signal amplitude AND duration- points on a curve, do not.
I don't think "duration" is really a thing here. Perhaps you should describe what you think it means in relationship to an audio wave.

There's no "duration", only a time interval between one reference point on the wave and the next.

You understand that this link you referenced shows sampled audio as points on a curve (the little blue dots in the picture),right?
 
Last edited:
highfigh

highfigh

Seriously, I have no life.
The same way non-digital tape is analog.

They are both (essentially) identical media.

The difference is the logic used to write and read to them; and generally the physical read/write head.

the little magnetic paricles are aligned on a continuiously variable orientation in 360x360-degrees. Hence: analog.


I don't think "duration" is really a thing here. Perhaps you should describe what you think it means in relationship to an audio wave.

There's no "duration", only a time interval between one reference point on the wave and the next.


You understand that this link you referenced shows sampled audio as points on a curve (the little blue dots in the picture),right?
Are you arguing for the sake of argument?

THE MEDIUM ISN'T THE POINT! IT'S WHAT'S RECORDED ONTO THE MEDIUM (AND HOW) THAT'S DIGITAL OR ANALOG!!!!!!!!!!!!!!!!!!!!!!!!!!!

No duration eh? Did time stop when going from one sample to the next? No. Is there a sampling rate? What is the reference for that? Time. You can't remove time from digital or analog.
 
Last edited:
JerryLove

JerryLove

Audioholic Samurai
All media is analog. Everything in the real world is analog (at least above a quantum level).

The question is whether the *information* is analog.
Are you arguing for the sake of argument?

THE MEDIUM ISN'T THE POINT! IT'S WHAT'S REDORDED ONTO THE MEDIUM (AND HOW) THAT'S DIGITAL OR ANALOG!!!!!!!!!!!!!!!!!!!!!!!!!!!
Gee. If only I had thought to say that exact thing.

Oh! I did say that exact thing already; and now you are typing it back to me in all caps as though I didn't and asking me if I'm just arguing for the sake of arguing?

I feel like you haven't bothered to read what I said. Certainly, you have not comprehended it.

No duration eh? Did time stop when going from one sample to the next? No. Is there a sampling rate? What is the reference for that? Time. You can't remove time from digital or analog.
So when I asked before "Perhaps you should describe what you think it means in relationship to an audio wave.", it was to try to understand what your claims were so that we could find consensus. You have made zero attempt to answer. Why is that? All that can do is force an argument from non-clarity.

Unless you just want to argue for the sake of arguing.

Of course there is a sampling rate. Of course the sampling rate is an interval between measurements.

In fact. I already said exactly that "There's no "duration", only a time interval between one reference point on the wave and the next."

The "time interval" referenced there is the "sampling rate". But unless the interval is variable: there's no need to encode it in anything other than tag information in the file header or footer. It's a given value and writing it over and over would be a waste of time.

Again: look at the link that *you* provided. Word search for "duration" on that page.

The scale is part of the spec. The sampling rate is part of the spec. Where these are variables, they would only need to be declared once. To actually digitize a waveform requires only a single value representing the Y value of the wave repeated over and over as we move along the X axis.

That's what the ADC does. The DAC then takes these points, applies "maths" to make a curve between them, and reproduce the original wave.

So again: why are you arguing by asserting something I already said as though it's a disagreement with me.

While I'm asking questions: would you care to answer my point about a "digital watch display" not meeting your definition of the word "digital" because it uses digits that aren't "0 and 1"? I mean: you brought up that argument, without explaining how it was germane to the large point, and you did so in a rather patronizing manner... like you just wanted to have an argument or something.
 
Last edited:
Bucknekked

Bucknekked

Audioholic Samurai
I don't have a leg to stand on here. I'm just a guy who listens to music and tries to enjoy the chatter and talk about how we make it at home. At the risk of sounding too simplistic, I think the discussion of what's analog and what's digital has reached a point of diminishing returns. There's some conflation going on and that's probably not helpful to the overall discussion.

If I may suggest that because something has discrete values, i.e. digits, it is not what is normally considered "digital". As well, giving some event or measurement a discrete value, like a "6", doesn't make it "digital" either.
One may argue from a definition point of view that since I have 5 digits on my hand, my hand is now digital. While it may fit a definition, it doesn't fit common usage.

The most common definition and usage for analog and digital in the audio world may not fit a precise Websters description, I would wager (someone mentioned a wagering school) most folks know what digital media is. Most common folks would also be able to guess that a vinyl record is an example of analog media. How precisely we can define the record as being analog is one of those exercizes that put us all to sleep in school.

This thread started as a discussion about the relative merits, or lack thereof, for the MQA technology inside Tidal. Does MQA and its associated tech provide an audible benefit? Is that audible benefit worth 2x what other services cost? Somehow we got off the track and in to a discussion with diminishing returns about what's digital and what's analog.

If you line up 100 of the AH forum members and asked if a vinyl record is analog, I think an astonishing number would get the concept right. If you ask 100 of the AH forum members if the stream of information coming off a CD platter is digital, I think an astonishing number would get the concept right.

I apologize in advance for the inconvenience of not having a bigger magnifying glass to see how many analog angels can dance on the head of a digital pin.
 
JerryLove

JerryLove

Audioholic Samurai
I agree with 90%... My contention would be with "If I may suggest that because something has discrete values, i.e. digits, it is not what is normally considered "digital". As well, giving some event or measurement a discrete value, like a "6", doesn't make it "digital" either. "

Describing something with a digit is literally the most digital thing I can think of.

Forget everything else and ask yourself what properties of digital vs analog things are.

Reproduce-ability. Digital content can be perfectly copied. Analog copies are always approximations. This is why analog suffers from generational loss and digital doesn't.

Error tolerance. Digital data can be completely corrupted by small changes in the data ("bar" vs "bear"). It tends to suffer catastrophic failure. Analog tends to fail incrementally.

Comprehension: Generally digital data must be understood by something to be used. It (generally) lacks a physical presence. Hence the DAC, or some reading/conversion device with logic is required. Analog can be transferred/changed through purely physical means.

And, yes, descretion. This is an area where highfigh and I do seem agree. Digital data is discreet (broken into small chunks) while analog data is continuous.

Circling back: It's not that the record is analog; it's that the grooves in the record encode in analog. I could put puts in a record and read them as 0 and 1 and have a digital record. Conversely: I could put analog data on a CD (indeed: early laser disk did use analog audio).

But for the most common datatypes (and speaking of audio)
Wax cylinder: analog
Record: analog
8-track: analog
Reel-to-reel: analog
cassette tape: analog
CD: digital
Everything that you have on your PC in whatever form: digital
Minidisc: digital
DAT: digital
DVD/HD-DVD/BR: digital.

Yes. We do all know. :)
 
lovinthehd

lovinthehd

Audioholic Jedi
I agree with 90%... My contention would be with "If I may suggest that because something has discrete values, i.e. digits, it is not what is normally considered "digital". As well, giving some event or measurement a discrete value, like a "6", doesn't make it "digital" either. "

Describing something with a digit is literally the most digital thing I can think of.

Forget everything else and ask yourself what properties of digital vs analog things are.

Reproduce-ability. Digital content can be perfectly copied. Analog copies are always approximations. This is why analog suffers from generational loss and digital doesn't.

Error tolerance. Digital data can be completely corrupted by small changes in the data ("bar" vs "bear"). It tends to suffer catastrophic failure. Analog tends to fail incrementally.

Comprehension: Generally digital data must be understood by something to be used. It (generally) lacks a physical presence. Hence the DAC, or some reading/conversion device with logic is required. Analog can be transferred/changed through purely physical means.

And, yes, descretion. This is an area where highfigh and I do seem agree. Digital data is discreet (broken into small chunks) while analog data is continuous.

Circling back: It's not that the record is analog; it's that the grooves in the record encode in analog. I could put puts in a record and read them as 0 and 1 and have a digital record. Conversely: I could put analog data on a CD (indeed: early laser disk did use analog audio).

But for the most common datatypes (and speaking of audio)
Wax cylinder: analog
Record: analog
8-track: analog
Reel-to-reel: analog
cassette tape: analog
CD: digital
Everything that you have on your PC in whatever form: digital
Minidisc: digital
DAT: digital
DVD/HD-DVD/BR: digital.

Yes. We do all know. :)
I like to distinguish non-electrical analog from the slightly less analog electrical type :) Old victrola style records/players were also such....
Analog purity! LOL
 
Doge

Doge

Junior Audioholic
Magnetic tape still exists- the master for vinyl was always magnetic tape before digital came along and the fidelity of tape can be better than that of vinyl. Recording studios are using analog magnetic tape much more often, lately.
I didn't know that. How are they dealing with the hiss?
 
Doge

Doge

Junior Audioholic
...

The DAC uses these points to recreate a continuous analog wave.
I know it is not as simple as this, but give me 3 points and I can give you a perfect circle that goes through them.

With all the starts/stops of different sounds, I agree, even if you can create the perfect note /SIN wave from a few points, you won't get the exact starting point.

Just that it seems all well beyond anyone's ability to hear it.
 
lovinthehd

lovinthehd

Audioholic Jedi
I know it is not as simple as this, but give me 3 points and I can give you a perfect circle that goes through them.

With all the starts/stops of different sounds, I agree, even if you can create the perfect note /SIN wave from a few points, you won't get the exact starting point.

Just that it seems all well beyond anyone's ability to hear it.
If you sample sound 44,100 times per second just how much off from the "starting point" could you get?
 
Doge

Doge

Junior Audioholic
If you sample sound 44,100 times per second just how much off from the "starting point" could you get?
1/44,100 of a second. There is more there that I don't know, and it is clear to me the human can perceive stuff that to me is amazing. But, that is kinda my point. The digital is beyond perception ability, for most of us...me.
What I have noticed is the high sample rate recordings more often sound better. DSD sounds better to me. I guess, that is because they just care more about the recording when they bother to release in some high sample rate format.
 
highfigh

highfigh

Seriously, I have no life.
I didn't know that. How are they dealing with the hiss?
They made sure the average VU level of the signal was high enough to mask the hiss and if not, they often used compression/expansion to lower the effective noise floor. Some studios used Dolby, some used dbx, some used another method of noise reduction. If the track was fairly quiet, they might record it hot to maximize the S/N ratio, then lower the level to match that of the rest of the tracks. If it's meant to be in the background, the noise from that track will be inaudible.

You might be interested in a show called 'Sound City'- it's about a recording facility of the same name, in Van Nuys, CA.
 
Doge

Doge

Junior Audioholic
My OP was kinda lost. This is all interesting. The math bravado is not. I am much more interested in what people hear.

Was it a good MQA marketing move to make an issue of something that could be so easily replicated by just more bandwidth (Qobuz). Maybe I just don't know what others can stream. I can get 2G Internet. I don't as I find 1/4 of that is fine - for everything. If you like having your own media, separate issue if you want it stored as analog or digital. But if you stream...

Did MQA and Tidal do it right - or did they actually push many to less folded formats - Qobuz and DSD downloads?
OR
Why use MQA vs a much larger file that still fits in the pipe just fine of non-lossy music.
 
Doge

Doge

Junior Audioholic
I think TIDAL sounds better. So, my marketing comments stand, but I renewed. I'll run both a couple months.
 
newsletter

  • RBHsound.com
  • BlueJeansCable.com
  • SVS Sound Subwoofers
  • Experience the Martin Logan Montis
Top