"Things that cannot be measured"

[Blumlein 88]

Having spent some time on this site I've noticed a trend with regards to these types of posts.

I can't really say if there is or isn't a difference with these devices but what is clear is that there are a number of individuals here who assume that the current measurement paradigms/protocols related to electronics capture EVERYTHING relevant to the performance of these devices. This seems to me to be somewhat of a dogmatic (if not outright arrogant) position on the matter. There was a time when the "objectivists" denied the relevance of jitter measurements when it came to digital devices.

I notice that they're now in the habit of altering thresholds, now demanding 0.1 dB of level matching, pretty soon they will insist on adding a decimal place in order to ensure that conducting the experiment is all but impossible.

What I see as most disturbing however is when they claim that they did the listening tests themselves and "heard no difference". I guess they're not acquainted with the concept of "observer bias". Not only do they not hear a difference, their entire worldview is based on them not hearing a difference. Can you imagine if they had to admit hearing a difference. You go on about how the measurements tell you everything so why are you pretending to listen for differences??

My suspicion is that static measurements do not tell the whole story but I am open to being wrong, something more folks on this site should also be open to.

It has been .2 db or .1 db for a long, long time. That is not new. Not even in this century.

Jitter was known about and accounted for in digital systems prior to their use in audio. They were a consideration in design from the very beginning of audio. So objectivists never did and do not deny its relevance. They may deny it is a problem with the devices we have because it is not a problem with the devices we have. The designs took care of it from the beginning. Chances are extremely good you have never heard a system with audible levels of jitter.

As for listening for themselves and hearing no difference, old complaint that does not mesh with facts. Some things are audible under controlled listening conditions. Sometimes when people say they have heard no difference they don't mean they listened willy nilly and made their claim. They mean they tried carefully with good level matching etc and didn't hear a difference. Their world view is not based on not hearing a difference. It is based on reality as much as possible. So next you wheel out the old and very tired shibboleth about static measurements. Plenty of measurements can be and are done with music vs music. Of course the right static measurements do tell you plenty. The supposition music will be different does not usually pan out.

 

[j_j]

Having spent some time on this site I've noticed a trend with regards to these types of posts.

I can't really say if there is or isn't a difference with these devices but what is clear is that there are a number of individuals here who assume that the current measurement paradigms/protocols related to electronics capture EVERYTHING relevant to the performance of these devices. This seems to me to be somewhat of a dogmatic (if not outright arrogant) position on the matter. There was a time when the "objectivists" denied the relevance of jitter measurements when it came to digital devices.

I notice that they're now in the habit of altering thresholds, now demanding 0.1 dB of level matching, pretty soon they will insist on adding a decimal place in order to ensure that conducting the experiment is all but impossible.

With all due respect (which is to say none whatsoever) this entire post is a heap of baloney.

1) There is NO evidence to the contrary in regard to the idea that modern measurements can miss anything relevant. Measurements can be hard to interpret, but the information is there.
2) Please don't talk crap like "objectivists denied ...", yeah. SOME of the objectivists may have, I don't know them personally, but those of us who came up through Telco engineering (Bell Labs Research in my case) were annoyingly well aware of things like jitter, were measuring jitter, and pointing out that the SPECTRUM of jitter was just as important as the level of jitter, FROM THE FIRST DAYS OF DIGITAL.
2a) Do you remember the old TDM lines, the old analog lines that "sung", and that would occasionally allow some leakage from an adjacent channel of voice? Did you realize that was a jitter problem of magnitude very very close to what's required to assure "below threshold of hearing" in 48kHz digital audio? Probably you don't, but what you'd fail to realize is that instead of one 20khz bandwidth channel, you were sending 14 4kHz channels? Do a little multiplication there and learn how that rather has a dastardly effect on crosstalk.
3) "altering thresholds" is sheer, tommyrot. Back in the first analog days, people like Fletcher, Zwicker, etc, ALL POINTED OUT THE NEED FOR LEVEL MATCHING TO UNDER .1 or .2 dB. A lack of level matching will ALWAYS create effects you can detect in a double-blind test, unless the test is fubar. If you make a good ABX test with two versions of the same signal, one of which is .2dB lower (or higher) in level, you'd BETTER see a positive result. If you don't, fix your test.

In short, your post is all of false, false-hearted, and completely inaccurate. It is also deeply offensive and pure propaganda.

Oh, and I should add, you DO REALIZE, do you not, that any test worth referring to has both positive and negative controls. Do you know what they are, do you know why? One of the things a positive control shows is that the test DOES have a specified level of accuracy and function. In short, you put in a test condition that SHOULD be audible, and if it's not, you've got a problem.

 

[j_j]

A reasonable answer to someone asking if an amp would make things more lively is that the perception of liveliness is often correlated to the high frequency response of the resulting sound. And this amp is flat through the audible frequency band. So it may be more lively than amps that are rolled off. Its damping factor might better control woofers making the bass they produce less bloated and thus sound more taunt and "quick". Etc. Etc.

First, things like "liveliness" and the like are effects that are much more energetic than the measured accuracy of very many amplifiers. If an amplifier has actually different sensation at that level, either it, or it's pair in the comparison, is broken, full stop. This will show up in an in-situ measurement like a blinding searchlight. Now, make sure your comparisons are level matched, because that kind of "feel" is often attributed to something that is a pure level difference.

As to the "damping factor" thing, almost all modern speakers are built with the idea that the amplifier is very close to a pure voltage source. As such, there really should be no effects of that sort AT ALL. If not, again, something is just plain broken.

 

[shion_ca]

Why were my posts moved to a thread called things that cannot be measured... when they were about things that could be measured and I gave exact examples of how?

 

[ahofer]

Why were my posts moved to a thread called things that cannot be measured... when they were about things that could be measured and I gave exact examples of how?

They don’t belong in the Buckeye amplifier thread, other than the simple answer to your initial question, and the conversation was rehashing old debates that the moderators quarantine in this and one other thread.

We don’t have a “audio review terminology is useful/useless” thread, although I know a couple of members who have spilled a trillion pixels on the subject.

 

[shion_ca]

They don’t belong in the Buckeye amplifier thread, other than the simple answer to your initial question, and the conversation was rehashing old debates that the moderators quarantine in this and one other thread.

We don’t have a “audio review terminology is useful/useless” thread, although I know a couple of members who have spilled a trillion pixels on the subject.

It wasn't my question for crying out loud. I own the amps. Your facts when you respond about my position, and even my having posited the initial question are completely wrong. And when I point this out... My posts are moved.

 

[SIY]

I’m not sure what you’re referring to when you talk about positive and negative controls

Res ipsa loquitur.

You need to know whom you're talking to here. I'd take those criticisms very seriously considering the source.

 

[ahofer]

I’m not sure what you’re referring to when you talk about positive and negative controls but it doesn’t address my main criticism related to the blatant confirmation bias done by those that are dogmatically convinced that any difference will appear in the measurements.

Do you know who you are arguing with? It is true you don’t know what negative and positive controls are. If you did, you might not have written the very next sentence.

 

[Beave]

Do you know who you are arguing with? It is true you don’t know what negative and positive controls are. If you did, you might not have written the very next sentence.

Beat me to it.

 

[JustJones]

I’m not sure what you’re referring to when you talk about positive and negative controls but it doesn’t address my main criticism related to the blatant confirmation bias done by those that are dogmatically convinced that any difference will appear in the measurements.

If you wanted to set up a proper test then you would need to include a device which measures in a way that should sound different along with the devices that shouldn’t

Google can be helpful

Positive Controls​

Positive controls are used in experiments to show what a positive result looks like. They ensure that the testing procedure is capable of producing results when the expected outcome is present. They involve using a material or condition known to produce a positive result.

Positive controls confirm that the experimental setup can detect positive results and that all reagents and instruments are functioning correctly and as intended.

Negative Controls​

Negative controls, on the other hand, are used to ensure that no change is observed when a change is not expected. They help confirm that any positive result in the experiment is truly due to the test condition and not due to external factors.

 

[Jim Taylor]

I don’t really care who I’m arguing with.

The issue is that those claiming to hear no differences are describing simple level matched listening tests, not employing controls of any kind to ensure their biases are addressed.

Have you viewed this?

Jim

 

[SIY]

Cool, now show me how many here are employing these controls when claiming they hear no difference.

I think you need to do some studying on experimental design.

Is this an audio forum or a cult?

If I'm blathering about string theory and Barton Zweibach harshly tells me that I don't know what I'm talking about, my first reaction would be to thank him and ask for some ideas on how to bring myself to a level where I could have a discussion with him. I wouldn't react by saying, "You physicists are just a cult!" But that's just me.

 

[shion_ca]

First, things like "liveliness" and the like are effects that are much more energetic than the measured accuracy of very many amplifiers. If an amplifier has actually different sensation at that level, either it, or it's pair in the comparison, is broken, full stop. This will show up in an in-situ measurement like a blinding searchlight. Now, make sure your comparisons are level matched, because that kind of "feel" is often attributed to something that is a pure level difference.

As to the "damping factor" thing, almost all modern speakers are built with the idea that the amplifier is very close to a pure voltage source. As such, there really should be no effects of that sort AT ALL. If not, again, something is just plain broken.

Sorry to take your time, I got moved to some form about things that cannot be measured when my posts were on the exact opposite topic. How you go about measuring perceived changes in sound

 

[ahofer]

those that are dogmatically convinced that any difference will appear in the measurements.

There is most certainly a scientific consensus that we know the ingredients of sound, and we have instruments that can measure those ingredients with much, much, greater sensitivity than the human ear. Therefore we can measure what can be distinguished by the human ear. There is plenty of proper experiimental evidence to back this consensus up and, most importantly, none that really calls it into question (or at least none that has been brought to this forum or the attention of the experts here). This is not ‘dogma’, and it is not discredited by the test design of some who claimed to hear no difference in comparisons of audio equipment.

And nobody has been abusive to you, you are reading too much into the text - but that happens to all of us.

I don’t really care who I’m arguing with.

I don’t really believe you.

EDIT: It appears I've said this before:

The only things worth arguing about with regard to measurements are a)whether we are doing it correctly or b)if we have the proper/full standard suite of measurements. The idea that there is some unmeasurable quantity that can be detected by ear but not instruments has no evidentiary basis in either audiology or audio comparison testing. The former is a ripe subject for discussion, although a lot more has been established in the scientific literature than most ”audiophiles” concede. The latter is just fantasizing, but endlessly attractive to the hobby, for some reason.

 

[SIY]

So me pointing out that the control measures to address listener bias is rarely ever mentioned or employed when doing level matched listening tests suggests to you that I should study experimental design??

Your absolute ignorance of experimental design is why you should study experimental design.

 

[JustJones]

So me pointing out that the control measures to address listener bias is rarely ever mentioned or employed when doing level matched listening tests suggests to you that I should study experimental design??

Not sure what site you're talking about but on this site proper controls are mentioned with regularity. We're human and all have biases that's a given

 

[j_j]

Well I see I’ve struck a nerve given the long winded attack aimed at my post.

This is typical of the kind of arrogance and sheer bullying frequently displayed in this forum.

I spoke facts. You speak libel. It's not arrogance nor bullying to simply point out the facts. You openly, in plain text, accused "objectivists" of cheating, and completely misrepresented each and every thing you brought up.

The fact that your drivel was rejected as a matter of fact is not bullying, it is educational.

 

[Blumlein 88]

 

[j_j]

I’m not sure what you’re referring to when you talk about positive and negative controls but it doesn’t address my main criticism related to the blatant confirmation bias done by those that are dogmatically convinced that any difference will appear in the measurements.

Wow, you certainly do NOT know anything about subjective testing. None the less, you make professional accusations. I suspect you don't belong in a scientific forum.

 

[DonH56]

Well I see I’ve struck a nerve given the long winded attack aimed at my post.

This is typical of the kind of arrogance and sheer bullying frequently displayed in this forum.

I’m not sure what you’re referring to when you talk about positive and negative controls but it doesn’t address my main criticism related to the blatant confirmation bias done by those that are dogmatically convinced that any difference will appear in the measurements.

If you wanted to set up a proper test then you would need to include a device which measures in a way that should sound different along with the devices that shouldn’t. This would address any observer (listener) bias and validate the trial methodology. No serious experimental protocol would ever allow a simple listening comparison when the biases of the observer are so blatant.

The fact that many of the hard core objectivists even mention their listening experiences as confirmation of their “measurements tell all” theory is laughable.

JJ's post went step-by-step through your post in an attempt to counter misinformation and explain what and why it was incorrect.

JJ has set up more than a few proper tests. He's also a Fellow of the AES and the IEEE, titles that mean something to technical folk, and his expertise garners respect from those who know him.

https://ieeexplore.ieee.org/author/37302284200

AES Member Profile » James Johnston

This is the member profile of an AES member.

www.aes.org