Why Do Many Reasonable People Doubt Science?

[firedog]

I'm interested to know it the members of the "dyads" had actual personal contact of some kind. If there was some, then a desire not to hurt the others feelings would probably come into play.

 

If there wasn't actual personal contact, then the results here are more disturbing to me.

 

Certainly another topical article.

 

The science of protecting people

[Teresa]

subjectivists care about what they hear, objectivists care about what other people hear.

 

Great observation and in my experience as well, oh so true. I'm saving your quote on my hard-drive and will attribute it to you when I quote it.

[mmerrill99]

Great observation and in my experience as well, oh so true. I'm saving your quote on my hard-drive and will attribute it to you when I quote it.

Yea, great quote & maybe it should be amended to "subjectivists care about what they hear, objectivists care TOO MUCH about what other people hear."

 

What is it that makes objectivists so concerned about what others hear (or even more correctly try to convince them that they DON'T hear something) - is it just a way of parading their "expertise".

 

To my mind, real expertise doesn't need an audience.

[Jud]

I'm interested to know it the members of the "dyads" had actual personal contact of some kind. If there was some, then a desire not to hurt the others feelings would probably come into play.

 

If there wasn't actual personal contact, then the results here are more disturbing to me.

 

They must not have offered enough money.

[Paul R]

Yeah, that's taken as written, actually, like it's part of the constitution. Nobody's suggesting blindly anything, I can't be expected to reiterate basic principles in every post though, just to accommodate every newcomer.

 

'Thoughtful analysis and questioning of their statements', however, is a hopelessly optimistic view of what being instructed by experts is like.

 

It's more like being out of your depth in a fast-flowing icy stream and struggling to keep your head above water.

 

These subjects (electronics) are so vast, that a university lecturer can overload any student's capacity at will. This isn't just true of electronics, in most disciplines, if they're any good at their jobs, and they haven't been told not to, they put you under pressure for your final year, to sort the wheat from the chaff.

 

So people who have passed through this process, although they're not identical to each other in capacity, are nevertheless trained to achieve a certain performance. This is just like the army. Going to the desert? Get desert training. Don't get arctic training (SNAFU). Course you do get a few REMFs. Everybody here, everybody I'm arguing with, seems to be saying 'I'm a citizen of the world, I'm a natural born survivor, after all I got here didn't I? I can survive anywhere without skills using my native intelligence and google'

 

Snort - even very smart and well trained people are stupid about some things, even in their field of expertise. Point out that you need to examine the time domain to some of the people here who are trained only to think in the frequency domain and watch them immediately react with refutations based in the frequency domain only, while claiming it is the same thing. (It isn't of course, it is two ways of looking at /expressing the same thing, where one way is a simplification. Not the same thing at all. The map is not the territory and all that...) Blind spots. Or doctors who smoke, if you want an example outside the audio world. How Stupid is that?

 

Depending upon experts for simple things one can learn for one's self is one of the more obvious ways a society starts to fall apart, and leads to the cult of "SCIENCE!" instead of a rational society that at least attemps to use the scientific method.

 

As to people going through the same training having generally about the same level of expertise? Nope. Not even close. That is where experience comes in.

 

Listen to experience first, then fit what you lean into what you know of theory. If one or the other disagrees, then ask questions and keep asking questions until you understand. That is also different from "get beat to death until you accept the consensus line..."

 

By the way, Chris Mooney is a smart guy, but delpnding upon the WP for science news and opinion? Think about that for a moment in relation to that article...

[SoundAndMotion]

Snort - even very smart and well trained people are stupid about some things, even in their field of expertise.

No doubt about that! But...

Point out that you need to examine the time domain to some of the people here who are trained only to think in the frequency domain and watch them immediately react with refutations based in the frequency domain only, while claiming it is the same thing. (It isn't of course, it is two ways of looking at /expressing the same thing, where one way is a simplification. Not the same thing at all. The map is not the territory and all that...)

I'm relatively new at CA and this sounds as though this may have been covered ad nauseum in other threads, but I'm curious how this has come up in audiophile threads. In my own work, I know exactly when I must work in the time domain and when I can use frequency domain methods, but please point me to a thread where it has been covered here. Which one is the simplification (i.e. which one is the map)? How so? (I'm happy to catch up on my own... just where?)

 

... the cult of "SCIENCE!"

You have mentioned "SCIENCE" before. What do you mean with the quotes and all caps?

Thanks.

[mayhem13]

No doubt about that! But...

 

I'm relatively new at CA and this sounds as though this may have been covered ad nauseum in other threads, but I'm curious how this has come up in audiophile threads. In my own work, I know exactly when I must work in the time domain and when I can use frequency domain methods, but please point me to a thread where it has been covered here. Which one is the simplification (i.e. which one is the map)? How so? (I'm happy to catch up on my own... just where?)

 

 

You have mentioned "SCIENCE" before. What do you mean with the quotes and all caps?

Thanks.

 

Care to elaborate exactly how someone removes 'time' from the frequency domain and vice versa?

[jabbr]

Certainly another topical article.

 

The science of protecting people

 

Nice. This phenomenon is very concerning but bears a little analysis (so we can prevent in the future).

 

The history of autism is distressing. So-called experts had blamed the parents, particularly mothers: History of Autism Blame | Refrigerator Mothers | POV | PBS. Asperger had initially thought the basis was more likely genetic, and this is the current likely etiology.

 

In a similar vein, the government has promoted corn products like margarine and subsidized the production of chemicals such as corn syrup resulting in the industrialization of food production.

 

We now know that both of these concepts are incorrect. In this light, however, it is not surprising that some people do not have complete faith in so-called 'SCIENCE'. The reality of the situation is that both of these prior beliefs were the result of bad science, but nonetheless there is a real fear of injesting chemicals (and there is another theory that autism itself is caused by environmental chemicals such as mercury etc.).

 

The problem is that people like Oprah, Jenny Mcarthy and "Dr. Oz" have a large audience on daytime TV and people believe them. If we required people to pass a basics of the scientific method test before being allowed to use a credit card, then advertising would be very very different.

 

Of course it is far more important for people to vaccinate their children, than to prevent some poor sucker spending $$$ on ethernet cables. In any case BAD science probably worse than complete subjectivity.

[mmerrill99]

Care to elaborate exactly how someone removes 'time' from the frequency domain and vice versa?

Time-and-frequency domains are interchangeable for linear systems.

When we talk about non-linear systems does this still hold?

[jabbr]

Snort - even very smart and well trained people are stupid about some things, even in their field of expertise. Point out that you need to examine the time domain to some of the people here who are trained only to think in the frequency domain and watch them immediately react with refutations based in the frequency domain only, while claiming it is the same thing. (It isn't of course, it is two ways of looking at /expressing the same thing, where one way is a simplification. Not the same thing at all.

 

Fourier analysis allows the conversion of a signal in the time domain (infinite) into a signal in the frequency domain (infinite). This mathematical conversion is always appropriate but not typically practical. When limiting the frequency domain, the so-called Discrete Fourier Transform (DCT), allows digital analysis. Analyzing signals by analyzing individual frequencies in an individual fashion is only appropriate for a subset of systems -- I have used the term: linear to describe such systems. Other systems might have factors which are multiplicative between coefficients of individual frequencies e.g. a[57khz]^2*b[5khz] etc. which thus are nonlinear.

So really the incorrect assumption is that all systems can be linearly analyzed via Fourier transform. Sort of like assuming that the Earth is the center of the solar system's gravitational field -- makes for weird orbits. If you assume that Mars orbits in an ellipse around the Earth you will be wrong.

 

The map is not the territory and all that...) Blind spots. Or doctors who smoke, if you want an example outside the audio world. How Stupid is that?

 

Unfortunately Nicotine is very addictive. Many people abuse substances in an attempt to self-medicate. I wonder what would happen to the rate of lung cancer if edibles became legal -- yet another example of bad science on the part of our government promoting policies that are more political than scientific!

 

Depending upon experts for simple things one can learn for one's self is one of the more obvious ways a society starts to fall apart, and leads to the cult of "SCIENCE!" instead of a rational society that at least attemps to use the scientific method.

 

In many things: Caveat emptor

[alfe]

Time-and-frequency domains are interchangeable for linear systems.

When we talk about non-linear systems does this still hold?

 

You just add space.

[mmerrill99]

You just add space.

Huh?? Sorry, I don't understand what you mean.

Jabbr has elucidated my point somewhat

[Jud]

Time-and-frequency domains are interchangeable for linear systems.

When we talk about non-linear systems does this still hold?

 

Depends what you mean by "non-linear," I suppose. In terms of pretty much everything audio-related, yes.

[Jud]

Huh?? Sorry, I don't understand what you mean

 

I believe alfe's post had an unshown winkie after it.

[jabbr]

Time-and-frequency domains are interchangeable for linear systems.

When we talk about non-linear systems does this still hold?

 

I hope my above response starts to clarify this. When analyzing anything you need to be careful about assumptions and that includes limiting either the time or frequency domains. In the same way that you *can* consider the Earth the center of the Universe, doing so will complicate the math needed to analyze interactions between the gravitational fields of say: Jupiter and Saturn.

 

When using discrete transforms (which is what we use in digital systems), we need to pay particular attention to the places where the series are truncated in both the time and frequency domains. From a practical point of view our electronics may "click" when changing songs and our digital filters may produce bad artifacts when abruptly truncating frequencies.

 

When a system is non-linear, the math used needs to be non-linear -- you can't blindly use linear analysis and expect acceptable results. In non-linear systems small coefficients may, for example, have very large effects on the outcome. This can cause serious digital artifacts due to bit depth -- 16 bits might normally be enough resolution to represent a signal but if there is a low level high frequency coefficient it might introduce unacceptable quantization error that would get greatly magnified in a non-linear system.

 

Bottom line answer is: yes you can analyze either way theoretically but there are important practical issues that may limit our ability to do so.

[alfe]

Huh?? Sorry, I don't understand what you mean.

Jabbr has elucidated my point somewhat

 

A non linear system involve time, frequency and spatio-temporal domains.

[Jud]

Fourier analysis allows the conversion of a signal in the time domain (infinite) into a signal in the frequency domain (infinite). This mathematical conversion is always appropriate but not typically practical. When limiting the frequency domain, the so-called Discrete Fourier Transform (DCT), allows digital analysis. Analyzing signals by analyzing individual frequencies in an individual fashion is only appropriate for a subset of systems -- I have used the term: linear to describe such systems. Other systems might have factors which are multiplicative between coefficients of individual frequencies e.g. a[57khz]^2*b[5khz] etc. which thus are nonlinear.

So really the incorrect assumption is that all systems can be linearly analyzed via Fourier transform.

 

Chapter from a free textbook: Characteristics of the Phase

[jabbr]

Depends what you mean by "non-linear," I suppose. In terms of pretty much everything audio-related, yes.

 

Is my above definition suitable?

 

 

Generally:

In physics and other sciences, a nonlinear system, in contrast to a linear system, is a system which does not satisfy the superposition principle – meaning that the output of a nonlinear system is not directly proportional to the input.

-- Nonlinear system - Wikipedia, the free encyclopedia

 

The salient point is that even if our ability to hear frequencie s > 13khz is significantly diminished and even if the signal does not contain much information in this bandwidth (low coefficient), in a non-linear system there still may be a vastly outsized contribution to some aspect of auditory perception. Moreover such systems cannot be properly analyzed by analyzing the response to individual frequencies -- you will entirely miss any coefficients which are the multiplicative result of multiple simultaneous frequency components.

 

Its really a huge difference.

 

Now of course we do everything to cause our electronics to be linear -- that's a perfect amp.

[mmerrill99]

Depends what you mean by "non-linear," I suppose. In terms of pretty much everything audio-related, yes.

The digital filters that you say you hear the effects of - are these linear systems?

Dither, quantisation noise, DAC noise modulation, DAC noise shaping, ........

What about microphones & speakers?

Finally, is Auditory perception linear?

[Paul R]

No doubt about that! But...

 

I'm relatively new at CA and this sounds as though this may have been covered ad nauseum in other threads, but I'm curious how this has come up in audiophile threads. In my own work, I know exactly when I must work in the time domain and when I can use frequency domain methods, but please point me to a thread where it has been covered here. Which one is the simplification (i.e. which one is the map)? How so? (I'm happy to catch up on my own... just where?)

 

Both are ways to describe a physical phenomena, and both are quite valid. Transforming into the frequency domain is almost always provides simpler manipulation, at least mathematically of course. Just do a search on "time domain" and I am sure you will find more than enough material to satisfy your curiosity.

 

 

You have mentioned "SCIENCE" before. What do you mean with the quotes and all caps?

Thanks.

 

There is normal science, which is probably the most accurate invention the human race has for describing reality. There there is "SCIENCE!", which is equivalent to a religious cult where anything and everything is blessed because it is "SCIENCE!" One can usually recognize the difference between someone who practices "science" (they are always questioning everything...) and one who practices "SCIENCE!" The latter usually doesn't want anything questioned because "science has already settled that issue" - or something similar. Also easy to recognize because of their strident and unrelenting tone.

 

Obviously, I am rather contemptuous of "SCIENCE!" while I quite respect and agree with the science and the use of the scientific method.

[Paul R]

Fourier analysis allows the conversion of a signal in the time domain (infinite) into a signal in the frequency domain (infinite). This mathematical conversion is always appropriate but not typically practical. When limiting the frequency domain, the so-called Discrete Fourier Transform (DCT), allows digital analysis. Analyzing signals by analyzing individual frequencies in an individual fashion is only appropriate for a subset of systems -- I have used the term: linear to describe such systems. Other systems might have factors which are multiplicative between coefficients of individual frequencies e.g. a[57khz]^2*b[5khz] etc. which thus are nonlinear.

 

Yes, but the important part I recognize in your comment is that we use DCTs as a practical thing, not because it is an exact description of the system. It is an approximation. That is not recognized or fully understood (or fully thought out?) by some folks. It's the same old heehaw we hear all the time.

So really the incorrect assumption is that all systems can be linearly analyzed via Fourier transform. Sort of like assuming that the Earth is the center of the solar system's gravitational field -- makes for weird orbits. If you assume that Mars orbits in an ellipse around the Earth you will be wrong.

 

=+1

 

Unfortunately Nicotine is very addictive. Many people abuse substances in an attempt to self-medicate. I wonder what would happen to the rate of lung cancer if edibles became legal -- yet another example of bad science on the part of our government promoting policies that are more political than scientific!

 

 

 

In many things: Caveat emptor

 

Yep. Though smoking is just slow suicide, and every kid in every 1st world country knows that. Health care professionals that smoke also should be aware of resources to help them quit without experiencing dramatic withdrawal effects.

 

-Paul

[Jud]

Is my above definition suitable?

 

 

Generally:

-- Nonlinear system - Wikipedia, the free encyclopedia

 

The salient point is that even if our ability to hear frequencie s > 13khz is significantly diminished and even if the signal does not contain much information in this bandwidth (low coefficient), in a non-linear system there still may be a vastly outsized contribution to some aspect of auditory perception. Moreover such systems cannot be properly analyzed by analyzing the response to individual frequencies -- you will entirely miss any coefficients which are the multiplicative result of multiple simultaneous frequency components.

 

Its really a huge difference.

 

Now of course we do everything to cause our electronics to be linear -- that's a perfect amp.

 

The digital filters that you say you hear the effects of - are these linear systems?

Dither, quantisation noise, DAC noise modulation, DAC noise shaping, ........

What about microphones & speakers?

Finally, is Auditory perception linear?

 

jabbr, in this context your response may not be as specific as I might have liked. There are certain inputs for which common types of analysis provide linear output, and others for which they don't - see the textbook chapter I referenced, for example.

 

mmerrill99, auditory perception clearly isn't linear. The current analysis tools available to model it are in my own insufficiently informed opinion very, very good, but not entirely adequate to completely substitute for reality; or to put it simply, the next time I am honestly confused between an audio system and a live performance will be the first.

 

Aside from whatever unknown (at least to me) shortcomings mathematical analysis of audio and our perceptions may have, there are some known ones that pique my interest. For the manyth time I'll link to the Resonnessence site's very nice explanation (in my inexpert opinion, anyway) of some of the problems with digital filters: Digital Filters | Resonessence

[mmerrill99]

The salient point is that even if our ability to hear frequencie s > 13khz is significantly diminished and even if the signal does not contain much information in this bandwidth (low coefficient), in a non-linear system there still may be a vastly outsized contribution to some aspect of auditory perception. Moreover such systems cannot be properly analyzed by analyzing the response to individual frequencies -- you will entirely miss any coefficients which are the multiplicative result of multiple simultaneous frequency components.

 

Its really a huge difference.

 

Now of course we do everything to cause our electronics to be linear -- that's a perfect amp.

Or, another example in auditory perception - the missing fundamental where we perceive a fundamental frequency when only it's overtones are played so some smaller speakers can give the impression of reproducing a lower frequency than they are physically capable of playing.

[mmerrill99]

mmerrill99, auditory perception clearly isn't linear. The current analysis tools available to model it are in my own insufficiently informed opinion very, very good, but not entirely adequate to completely substitute for reality; or to put it simply, the next time I am honestly confused between an audio system and a live performance will be the first.

Sorry, I don't think you are talking about the analysis tools of auditory perception here, I think you are really talking about the analysis tools that are use to model audio electronics.

 

The fact that we aren't fooled into thinking that audio reproduction sounds like a live performance, I would suggest, is because of the non-linearities in the audio reproduction chain & our lack of a clear model that accounts for auditory perception so that we can focus on the perceptually important non-linearities. It may also be the case that we have reached the limits of how realistic 2 channels can appear (although I think there is still more to play for) but only a better understanding of what is perceptually important will guide us towards better multi-channel systems.

 

The mistaken focus in audio electronics still, too often involves specmanship - a blunderbust approach to reducing measured metrics that is still not achieving perceptual transparency.

[Jud]

Sorry, I don't think you are talking about the analysis tools of auditory perception here, I think you are really talking about the analysis tools that are use to model audio electronics.

 

Well, no. I'm talking about things like designing digital filters that will hopefully sound as realistic as possible (also taking into account that for different people, different aspects of the sound may make this facsimile of reality better or worse).

 

The fact that we aren't fooled into thinking that audio reproduction sounds like a live performance, I would suggest, is because of the non-linearities in the audio reproduction chain & our lack of a clear model that accounts for auditory perception so that we can focus on the perceptually important non-linearities. The mistaken focus in audio electronics still involves specmanship - a blunderbust approach to reducing measured metrics that is still not achieving perceptual transparency.

 

I think "non-linearities" is way too general. There are quite specific known issues that, to return to the same example, digital filter designers must deal with, such as the balance between time domain and frequency domain distortions. Our "lack of a clear model" appears to be at least partially because different people react differently to these distortions, some favoring filters that, relatively speaking, minimize frequency domain distortions, others favoring filters that minimize time domain distortions.

 

I'm not sure when you talk about "audio electronics" whether you are including DACs or not; if you are, then the filters used in a DAC and/or associated software are of course an important factor in the resulting sound.