"Things that cannot be measured"

[StevenEleven]

And yet, musical instruments are tuned according to a precise mathematical relationship, each note having a well-defined frequency, also called pitch.

Sounds nice, neat and fine (my three suburban adjectives), but like most things that sound nice, neat and fine, it is perhaps not really true.

At the highest levels, musicians often manipulate pitch by ear in fine variations depending on the situation and their on-the-fly understanding of how it affects the audience and fits in the music. i.e., how it affects human perception. They might play a shade sharp to stand out a little, or adjust pitch to adapt to the kinds of instruments being played and how they best sound together. On a less elevated level, if you've ever tried to tune a guitar and really thought about what you were doing you may have learned the hard way that it's often a bundle of compromises up and down the fret-board. And as you may well know to some degree, pianos are tuned to kind fit in with other instruments and within their own harmonics six one and half-dozen the other, so to speak (i.e., "temperament tuning," as opposed to "just tuning") https://en.wikipedia.org/wiki/Musical_temperament. Some of this is discussed here. . . https://www.audiosciencereview.com/forum/index.php?threads/missing-fundamental.18930/

 

[threni]

I imagine someone about 100 years ago asking the same question, with a smug assurance that history has stopped right there.

But back then there might have been some differences to investigate.

Currently is it the case that you can present a pair of devices (amps, DACs) in a blind test, find that a meaningful number of people can repeatedly identify a difference, but be unable to determine through measurement what's causing the difference?

Without such a situation there's literally nothing to talk about.

 

[KSTR]

Currently is it the case that you can present a pair of devices (amps, DACs) in a blind test, find that a meaningful number of people can repeatedly identify a difference, but be unable to determine through measurement what's causing the difference?

Without such a situation there's literally nothing to talk about.

Full ack!
I'm really willing to throw a lot of effort into technically identifying and exposing root causes for different sound but first there must be established a fact of hearing.
I stopped counting how many times I've uselessly offered my help to people claiming audible differences in nailing it down if there is something to it...

 

[Blumlein 88]

Sounds nice, neat and fine (my three suburban adjectives), but like most things that sound nice, neat and fine, it is perhaps not really true.

At the highest levels, musicians often manipulate pitch by ear in fine variations depending on the situation and their on-the-fly understanding of how it affects the audience and fits in the music. i.e., how it affects human perception. They might play a shade sharp to stand out a little, or adjust pitch to adapt to the kinds of instruments being played and how they best sound together. On a less elevated level, if you've ever tried to tune a guitar and really thought about what you were doing you may have learned the hard way that it's often a bundle of compromises up and down the fret-board. And as you may well know to some degree, pianos are tuned to kind fit in with other instruments and within their own harmonics six one and half-dozen the other, so to speak (i.e., "temperament tuning," as opposed to "just tuning") https://en.wikipedia.org/wiki/Musical_temperament. Some of this is discussed here. . . https://www.audiosciencereview.com/forum/index.php?threads/missing-fundamental.18930/

I don't think these facts dispute what mansr has written. What you describe is one reason I prefer to record complete musical groups rather than each one individually. There is an overall gestalt to the whole group that is more organic feeling when they all play and hear each other together. I've even found with some smaller groups they need to be within so many feet of each other or the delay messes with their relative timing to each other. Yet those who are most attune to this in my experience are also sticklers about getting their instruments most carefully tuned just before we record something. They are tuning to very precise electronic tuners.

 

[KSTR]

Delay is really PITA. As a bassist, I once had my monitor failing on a big stage and the drummer was like 20ft away and I was always late wrt to the drummer (who had my signal on his monitors). He thought I was constantly trying to slow the tempo down during the rest of the song.
Delay also is the reason why a large orchestra never ever can sound tight with percussion involved, at least from the viewpoint of the individual musicians.

 

[StevenEleven]

I don't think these facts dispute what mansr has written. What you describe is one reason I prefer to record complete musical groups rather than each one individually. There is an overall gestalt to the whole group that is more organic feeling when they all play and hear each other together. I've even found with some smaller groups they need to be within so many feet of each other or the delay messes with their relative timing to each other. Yet those who are most attune to this in my experience are also sticklers about getting their instruments most carefully tuned just before we record something. They are tuning to very precise electronic tuners.

Thanks, Your observations about recording are quite interesting.

Perhaps I rambled on a bit too long. I suppose the salient and more concise point in response to mansr would be that In equal temperament tuning (including as characterized by the precise tuning frequencies often found on electronic tuners), the precise mathematical relationships he alludes to are intentionally compromised, for practical purposes. The Wikipedia page discusses it at length. No big deal. I just find it interesting. I’m on the same page really as the tenor of the thread. I’m about done with folks trying to identify or claiming unmeasurable but audible differences between DACs, amps, sources, etc., myself. I just get distracted by what I actually find interesting—music, speakers, rooms, etc.—and recordings!!!

 

[KSTR]

As for tuning during recording sessions, guitarists often "tune to the song's dominant chords" so that the chromatic tuning doesn't spoil things too much.
John Frusciante being one prominent example.
Also, good guitarists always tune a bit too low because then you have room to arrive at the correct perfect intervals by bending the strings as needed, this is notably very important with lots of distortion. Personally, as a guitarist, I'm the kind of guy who pulls/pushed the neck all the time to get some fluctuation (and it will help smear slight tuning issues ;-)

 

[mansr]

Pitch and pitch perception are different things. Pitch comes before pitch perception and is measurable.

Right. CMOT is doing the equivalent of defining length as how a line is perceived rather than how long it actually is, then claiming that rulers don't work.

 

[mansr]

Perhaps I rambled on a bit too long. I suppose the salient and more concise point in response to mansr would be that In equal temperament tuning (including as characterized by the precise tuning frequencies often found on electronic tuners), the precise mathematical relationships he alludes to are intentionally compromised, for practical purposes.

In equal temperament tuning, the frequency ratio of adjacent semitones is exactly the 12th root of 2. This doesn't quite match the frequencies used in just intonation or meantone temperament. It's a bit of a mess, but it's a well-defined mess.

 

[mansr]

You sure it wouldn't be 16 toe taps, toe taps would be in the ounce range?

That depends on whether it's the imperial or the colonial system.

 

[CMOT]

But perception cannot be measured, it is individual and variable with mood, drugs, fitness, age, hunger etc. So what we can do is measure signals for fidelity and leave perception out of it.

So what actually matters when listening to music is the fidelity of the signal. The perception is out of everyone's control, usually even the listener's.

But since we cannot possibly recreate the exact signal of the original performance, we make psychoacoustic assumptions in both what we consider to be important in the fidelity of the signal and in how we record and recreate that signal to begin with. Even things like the assumption of stereo. But wrt fidelity, we, for example, decide that accurate reproduction over, say 60kHz, is not important - not because it couldn't be, but because we can't hear such signal - it is a psychoacoustic model of the frequency range we can perceive. Similarly, we make assumptions about the appropriate sampling rate, number of bits, etc. for digital music. Or the range of loudness we need to reproduce accurate. These are the more obvious assumptions. My point is we decide what measurements of fidelity matter and to what degree because of the humans listening to these signals - so we can't leave perception out of it. And to return to the question that I originally responded to, someone asked what differences should we measure and why (paraphrase here). We can't really know which things we should measure and with what precision without taking the listener into account. If bats had audio systems, they would probably make somewhat different decisions in measurement. So we can't avoid the psychoacoustics of things. That doesn't mean we shouldn't use measurement of fidelity as the way to compare different pieces of audio gear, just that we should understand what those measurements might mean in terms of our actual listening (Amir makes this point all the time about certain measurements) and we should probably not worry too much about measuring too much of things that we couldn't possibly hear. Modulo that we still don't understand exactly how and what people can hear across the entire human population.

 

[CMOT]

Right. CMOT is doing the equivalent of defining length as how a line is perceived rather than how long it actually is, then claiming that rulers don't work.

Absolutely not. And that is a rhetorical paraphrase that isn't what I said. I am saying that there is a physical measurement and there is a perceptual measurement/experience. But we shouldn't confuse the two. We need to understand how the two relate, which is not always clear nor linear. So you can measure all the lines you want with rulers, but what matters for the perceiver is what they perceive and why. So if context leads to me perceiving two equal length lines to be different lengths (which does happen), a system designer better well take that into account if line length and how I respond to it is important in that system. Audio reproduction is the same way. Lets measure everything precisely and with fidelity. I 100% agree. But we should understand how the things we measure are perceived by listeners. It matters for what differences we allow or not and how we design and build our systems. Every system is compromise in some way - there is a noise floor, there is channel separation, there is sampling (sometimes) etc. Do these measured things matter and to what degree?

 

[NTK]

@CMOT There is no need to explain. You are just being trolled

 

[mansr]

So if context leads to me perceiving two equal length lines to be different lengths (which does happen), a system designer better well take that into account if line length and how I respond to it is important in that system.

Pictures of optical illusions work without the image coding and display systems taking into account anything about perception. Likewise, sound reproduction doesn't rely on perception. If a physical stimulus is recreated with sufficient accuracy, the perception will be the same. Precisely what will be perceived and why is not important. Of course, considering perceptual effects can allow us to substitute a simpler stimulus that still produces (almost) the same experience as the original. That is how colour displays can get away with using only red, green, and blue.

This silly argument started with you saying it is impossible to measure pitch, which is simply not true. Pitch is just another word for frequency. If you have a problem with that, you should write a letter of complaint to your lexicographer.

 

[Kal Rubinson]

Pitch and pitch perception are different things. Pitch comes before pitch perception and is measurable.
If I had better pitch perception I may have had a chance of singing in-tune.

If I had better pitch perception, I might not strike out so often.

 

[CMOT]

Pictures of optical illusions work without the image coding and display systems taking into account anything about perception. Likewise, sound reproduction doesn't rely on perception. If a physical stimulus is recreated with sufficient accuracy, the perception will be the same. Precisely what will be perceived and why is not important. Of course, considering perceptual effects can allow us to substitute a simpler stimulus that still produces (almost) the same experience as the original. That is how colour displays can get away with using only red, green, and blue.

This silly argument started with you saying it is impossible to measure pitch, which is simply not true. Pitch is just another word for frequency. If you have a problem with that, you should write a letter of complaint to your lexicographer.

Pictures of optical illusions work without the image coding and display systems taking into account anything about perception. Likewise, sound reproduction doesn't rely on perception. If a physical stimulus is recreated with sufficient accuracy, the perception will be the same. Precisely what will be perceived and why is not important. Of course, considering perceptual effects can allow us to substitute a simpler stimulus that still produces (almost) the same experience as the original. That is how colour displays can get away with using only red, green, and blue.

This silly argument started with you saying it is impossible to measure pitch, which is simply not true. Pitch is just another word for frequency. If you have a problem with that, you should write a letter of complaint to your lexicographer.

Please feel free to respond or state that "Pitch is just another word for frequency" on any psychology, cognitive science, or neuroscience exam. While you are at it, you can assert that color is just another word for wavelength (in part of the EM spectrum).

 

[mansr]

I see you have moved on from the absurd to the ridiculous.

 

[JustJones]

Why is this thread starting to remind me of "God of the gaps" arguments ID proponents bring up? We can measure what the human ear can hear. How each individuals brain perceives this information isn't measurable or really relevant as far as I can tell. We have the equipment to reproduce and measure sound humans can hear and can't hear. Is it possible in the future someone develops an instrument that is better at these measurements? Yes, does it matter since we can already reproduce and measure everything humans and dolphins can hear? No.

Isn't the pitch A above middle C 440hz? An arbitrary standard we use. Is there a pitch that can't be represented by frequency?

 

[Cheztumoé]

I am 76 years old with 8khz ears and I have been chasing sounds all my life.
My equipment is Topping E30 and Topping L30, Yamaha AS-1100 , Yamaha AS-592, Denon X1400, speakers Avid 105, Rectilinear III HighBoy , Def. Tech. BP 10 . I have a Onkyo C-7030 cd player and 1000s of cd which I never play as I dont hear much difference from streaming services. I have read articles in ASR that indicates that it is unlikely that I could hear the differences between a $200.00 and $1000. dacs . I have taken AB test and I can't hear the difference from 320kbps and flac. I have tried all the premium streaming services like Tidal HIFI, Deezer HIFI, Primephonic, Idagio and Amazon HD. I could not try Qobuz at it is not available in Canada. I have cancel all streaming services except Amazon Hd which I presently really enjoy.
While I read and like what people say on ASR , I listen to a lot of classical and jazz music on Amazon HD and keep the one which the sound I like the most. So for me now it's what I hear and what I like or not.

 

[magicscreen]

Sometimes worse measurements are better:
saturated tubes