Is there more to Audio Measurements?

Status
Not open for further replies.
As always, logical fallacies crop up. There is no ideology in science, however, the same is not true for those who make unsubstantiated and unsustainable claims, such as there may some aspect of audio signals which are yet to be measured.
One is supported and the other is almost voodoo.

You couldn't be more wrong!

And what is "unsubstantiated claims" - been in a room with a bunch of scientists and knocking ideas around, sometimes jokingly, and never heard any of them or any other scientist say "unsubstantiated claims" - and you forget that often an important role is intuition.

I have dealt with scientists from the ANU and Monash University - and I see that you are in Melbourne, so Monash should be very familiar to you. I am working on a subject that is getting traction, but never would if it was relegated to a mere "unsubstantiated claim" and I got news for you, this is not how this works.

I tell you what really counts: RESULTS!

But give and take, reasonable time is also granted, patience is a virtue, but not with social media where premature pouncing is now a blood sport. Where is the science in that?

There are many competing ideas in science, there are even scientists who say that they don't like such-and-such science (like QM). No ideology? Where were you when "Climate Gate" exploded and the public was horrified that scientists were sending heated emails to each other: Well, welcome to the real world of science and like just about any other endeavour, there are not always black and white positions. Besides, science is not about finding absolute truth, it is about observation and accumulating knowledge over time and not in a nanosecond.

Science has limits, a few things that science does not do:

Science doesn't make moral judgments.

Science doesn't make aesthetic judgments.

Science doesn't tell you how to use scientific knowledge.

Science does not make quick off-the-cuff judgements.

In fact, social media and science does not apparently mix?

And what does that have with DIY? This domain is about DIY and not about preaching one version of science against another.

And what is a logical fallacy, but a contrivance - a poor substitute for what is just a disagreement. This is a misuse of the term - a cheap weapon in the armoury of a keyboard warrior.
 
Yes, & I disagree with Jakob - most music we replay isn't a recording of a live event, it's a studio creation so knowing how it should sound is unknowable but there is a resolution to this 'cone of confusion' as Toole called it

Therefore we don't judge it as such - we judge it according to our internal auditory models of sound objects & their behavior which we have learned & refined through many years of exposure to how sound objects & sound streams behave in the natural world. The closer this match, the more convincing the illusion, the more realistic it is perceived to be & as a result, we often connect with it in a more emotional way. Our higher cognitive abilities have more processing time available because we are not expending our limited processing resources to reconciling the unnatural ambiguities in the sound stream.

So we are judging the naturalness of the sounds themselves & the correct interrelationship between the various sounds - the patterns & sometimes the statistics within the sound stream.

When a replay system allows you to hear the inner detail of gentle brushing of cymbals rather than it sounding like white noise, when a system shows the inner detail of applause - you know that it is reproducing the low level detail needed for this naturalness to be perceived
For the last time, cut the verbosity and stop conflating "preference" with "difference".

While the preference for a certain audio setup is certainly up for discussion (implying there is no evidence that a measured "accuracy" or "fidelity" is equivalent to "better sound"), the ability to detect differences between two audio entities in a controlled environment is a completely different kettle of fish.

It may be very well that our stupid and rudimentary Fourier analyzer located between the ears sends dopamine signals when detecting certain harmonic distortions, or whatever other artifacts that would otherwise qualify as "inaccuracies", I don't know. But glossing about one's "preference" is a useless exercise; there are no two identical brains, hence one's "preference" can barely be extrapolated to others. Preference controlled tests can be (and are readily) designed, but they are very expensive to conduct and let's face it: nobody in the leftovers of what used to be the Audio Industry has any interest to conduct any such tests. It's much cheaper and effective to adopt snake oil marketing tactics.
 
The illusion is frankly pants with 2 channels. Inner detail be damned. If you go to a live acoustic event sound is all around you.

Oh dear, Alan Blumlein & many others wouldn't agree with you about it being pants - of course it's not up to the live experience - that's the meaning of an illusion, it's not the real thing.

Inner detail is what differentiates better replay systems which are perceived as more realistic (given the limitations of 2 channels & the illusion created)

We quiet happily look at TV, cinema & know it's not the real thing, just an illusion & yet we are able to judge the differences in quality between these illusions. The better the illusion the more we are freed to engage with the content rather than our conscious attention being drawn to screen anomalies.

I would cite the banding often seen in TV reproduction of what should be uniform blackness as one such anomaly that draws our attention away from the content & focuses it momentarily to a flaw in the reproduction. We know this is wrong, not because we were there when the scene was shot, but because it doesn't fit with our internal model of how blackness is perceived.

Inner detail, in the visual world, certainly has a bearing on how realistic we view an image to be, especially a moving image (audio is a moving stream of sounds)
 
For the last time, cut the verbosity and stop conflating "preference" with "difference".
What you call verbosity is explanation - something you would be well to practice in your replies.

Stop confusing others with your confusion.

It's how auditory perception works - nothing to do with preference.
If you knew anything about auditory perception, you would know this & wouldn't be so confused

While the preference for a certain audio setup is certainly up for discussion (implying there is no evidence that a measured "accuracy" or "fidelity" is equivalent to "better sound"), the ability to detect differences between two audio entities in a controlled environment is a completely different kettle of fish.


It may be very well that our stupid and rudimentary Fourier analyzer located between the ears sends dopamine signals when detecting certain harmonic distortions, or whatever other artifacts that would otherwise qualify as "inaccuracies", I don't know. But glossing about one's "preference" is a useless exercise; there are no two identical brains, hence one's "preference" can barely be extrapolated to others. Preference controlled tests can be (and are readily) designed, but they are very expensive to conduct and let's face it: nobody in the leftovers of what used to be the Audio Industry has any interest to conduct any such tests. It's much cheaper and effective to adopt snake oil marketing tactics.
You built the "preference" strawman & then tilt against it
Unfortunately, even your strawman argument holds no water - Harmon's speaker preference tests show that people coalesce to more accurate sound (according to the particular off-axis measurement they correlated it with)
 
@scottjoplin,

I take your point, but that's unknowable, granted we could make a guess based on our own experience

It is unknowable to us mere consumers of prerecorded content, but isn´t that generally true for the so-called input signal?
In theory one could to visit the recording facility but in practice that never happens to the mere consumer.

It is a hypothesis that a total linear processing of an (arbitrarily) choosen signal would per se lead to an accurate representation of the sound event that was recorded.

@00940,

Is it really ? The different parts of a system fulfill different roles and should be assessed for their ability to fulfill their particular role.

Some gear (DACs, amps) are defined by how faithful their output is to the input. That's a relatively straightforward question of measuring electrical signals. In that field, differences between competent gear today are at the level of a fly's fart.

That´s a hypothesis but is it true?
First of all (see my answer to scottjoplin) it is based on a premise that will be true in some cases but will not true in other.
And it is based on the premise that the usually available subset of measurement reflects the components behaviour in reality to a sufficient degree.

When this alleged "level of a fly´fart" is factually reached is still debateable as it was asserted already around 1978 but wasn´t true back then.

And what about all the used gear that already exists? Does it vanish or is it already vanished?
Afaik vintage audio gear is in use for quite long time spans.

Some other gear (DSP, speakers, room treatment, the whole recording/mixing process,...) are defined by their ability to make us believe in the recreation of the musical event. Considering all the compromises involved and all the different possible approaches, that's where better weighted measurements and a better understanding of the brain auditory functions would help.

Sure and there is ongoing reasearch in these areas.

imo, a lot of the hifi industry is busy hyping the small differences to be found with the first kind of gear while the big gains in realistic sound reproduction should rather be pursued with the second kind.

You were right if your premise(s) were true, but is it?
As said before, human listeners represent nonlinear systems so it is sometimes quite difficult to predict how they respond to certain aberrations

@mmerrill99,

Yes, & I disagree with Jakob - most music we replay isn't a recording of a live event, it's a studio creation so knowing how it should sound is unknowable but there is a resolution to this 'cone of confusion' as Toole called it

It might be a "studio creation", but i´d say (beside some realisations of art) there nevertheless was a sound event that was in the end distributed as recorded content.
And of course could standardization of production and reproduction solve the cone of confusion problem, but what about all the records done in the past and until the ultimate goal is reached?

Therefore we don't judge it as such - we judge it according to our internal auditory models of sound objects & their behavior which we have learned & refined through many years of exposure to how sound objects & sound streams behave in the natural world. The closer this match, the more convincing the illusion, the more realistic it is perceived to be & as a result, we often connect with it in a more emotional way. Our higher cognitive abilities have more processing time available because we are not expending our limited processing resources to reconciling the unnatural ambiguities in the sound stream.

That seems to be a highly probable hypothesis..
 
<snip>

While the preference for a certain audio setup is certainly up for discussion (implying there is no evidence that a measured "accuracy" or "fidelity" is equivalent to "better sound"), the ability to detect differences between two audio entities in a controlled environment is a completely different kettle of fish.

"Different kettle of fish" ? Yes and No.
Yes, because tests for difference are usually different
No, because an established preference needs a detected difference, but you have to check for possible segmentation.
...... there are no two identical brains, hence one's "preference" can barely be extrapolated to others.

Doesn´t that hold true too in the case of differences?

Preference controlled tests can be (and are readily) designed, but they are very expensive to conduct and let's face it: <snip>

Could you explain why preference tests are more expensive to conduct than difference tests?
 
The illusion is frankly pants with 2 channels. Inner detail be damned. If you go to a live acoustic event sound is all around you.

It can be done with two channels, just that very few have heard it, and I mean very few!

And I am talking two discrete channels, no processing, just as it comes off the recording. You will not believe it until you have heard it, so no need to utter the 'prove it' argument. But once heard, you crave it.
 
- the complexity of the actual music signal we listen to Vs the test signals used for measurements

This is one of those "intuitive analogies" that are no more accurate than saying the flow of electrical current is like water flowing down a creek

- the limitations of the measurements themselves

From source to final transducer the information is contained in the electrical signals. Adequate measurement technology exists and claims of "hidden" or unmeasureable information have no basis rather than idle speculation by people that generally don't understand the technology. As for differences in speaker technology or cone treatments, I think that is for another discussion.

- the judgement criteria used to dismiss what's audible/not audible in the measurements

The crux of the matter, even the most benign processing (such as a little a equalization) will leave large input to output differences when measured and compared in the most SIMPLISTIC sense.

- the lack of correlation between measurements & auditory perception

It's the interpretation and analysis of the measurements not the measurements themselves. In each case propose a hypothesis and some experiments. This is hard as Dave's work on speaker IM distortion showed. There is little reward in this and I don't see a large unserved market of folks ready to pay money for this "accuracy" whatever it is.
 
It can be done with two channels, just that very few have heard it, and I mean very few!
And I am talking two discrete channels, no processing, just as it comes off the recording. You will not believe it until you have heard it, so no need to utter the 'prove it' argument. But once heard, you crave it.
Yes, in theory it is perfectly possible and indeed expected.
In practice not so simple, but doable and I have 3D from my current (and earlier) systems.
There are a few elementary things to get right, including factors such as absolute acoustic polarity, channel matching and direction of cables.

Dan.
 
Last edited:
When this alleged "level of a fly´fart" is factually reached is still debateable as it was asserted already around 1978 but wasn´t true back then.

And what about all the used gear that already exists? Does it vanish or is it already vanished?
Afaik vintage audio gear is in use for quite long time spans.

Well, it was optimistic in 1978, 40 years ago. A lot of things happened since though and the margin for progression gets smaller and smaller.

I don't get the point about vintage gear in this context ? Some of it is pretty good, some pretty bad. A good occasion to learn from the past and investigate why some designs enjoy such sustained popularity.
 
Doesn´t that hold true too in the case of differences?

No. Different brain segments are involved, in case of "difference" there's no "pleasure" involved. Neurochemical stimulation of pleasure centers is different than rationalizing a black/white decision.

Could you explain why preference tests are more expensive to conduct than difference tests?
At least because, as you just said, a reliable "difference" test is required before the "preference" test. Doubles the cost, also I'm suspecting (intuitively, don't have any direct experience) a reliable "preference" test requires a much larger sample to draw any statistical valid conclusion.
 
Last edited:
No. Different brain segments are involved, in case of "difference" there's no "pleasure" involved. Neurochemical stimulation of pleasure centers is different than rationalizing a black/white decision.

Not really. Have you read Kahneman? Haidt? It is more the exception than the rule for people to be capable of open-minded objectivity. However, belief that one is or can be uninfluenced by hidden mental processes is ubiquitous.
 
What is widely criticized by the other camp is the "difference" tests are delivering unjustified null results. Do you have any open source references that shows "hidden mental processes" are responsible for hiding "differences" in a well designed test (and I mean, WELL DESIGNED, with controls, training, etc... to Mr. Jakob's satisfaction, whatever that means).
 
There are at least 2 big lies about audio measurements.

1. Stereo. No one measures stereo speakers. Always mono. This is absolutely a fraud. We are always listening to comb filtered music from stereo speakers, but we pretend to ignore it, because it is a fundamental defect of our audio system.

2. Room. This is by far the most important thing when we playback music at home, and it can be measured, but most of us pretend to forget about it, and keep measuring audio gadgets which effect is so tiny compared to room itself.

Headphones? well, I do not like them...
 
Status
Not open for further replies.