"Things that cannot be measured"

[RexrothPigeon]

a basketball must be a certain diameter, weight, and bounce when inflated to a certain pressure.

not true. there are some governing bodies with rules about this (and those rules vary lol) but 99% of the basketball played on earth is not subject to the rules of any of those governing bodies.

 

[Curvature]

So what? Establish that final connection for yourself. If after checking gear measurements and trying to correct behavior with Dirac et el... you clearly have to try something else, whatever it is.

Like I said, establish which measurements are truly relevant for YOU. If you like some harmonic distortion etc... the measurements will reveal it too. Shrug off the personal opinions and conclusions. Measurements can help you establish what YOU like, and who cares if anyone in the audio community doesn't approve of your not perfectly linear and distortion free personal preferences?

Do you leave it to measurement to establish the kind of woman you love? Your favorite pizza joint?

This topic is a long noodling of misapplying the practice of measuring for a ridiculous purpose.

PS: the "you" is generic, not directed at you, I agree with what you say.

Preference established under blind circumstances is not the same thing as liking something in others.

Liking something is subject to whims and stubbornness. Preference can be measured and and evaluated.

 

[Ladipoe@98]

I think the interesting part is figuring out what actually matters to the listener versus what we simply haven’t found a good way to measure yet. Measurements are incredibly useful, but they don’t necessarily tell the whole story.

 

[solderdude]

Measurements, if accurate enough, can only show the aspects that are measured.
Poor accuracy, incorrect measurements, poor resolution, not measuring relevant aspects are also measurements.
They may appear useful but may not be.

For audio we need to divide measurements in at least 4 different realms before saying something about measurements and their usefulness/accuracy.

• Electric domain
• Mechanical domain
• Acoustic domain
• Perception

The above list is in order of difficulty and accuracy for measurements.
Some are really easy to do with high accuracy where other measurements are extremely complex and involve human perception.

 

[rgpit]

It all can be measured. Just have to know what to measure and how?

 

[stuartkohl]

Measurements, if accurate enough, can only show the aspects that are measured.
Poor accuracy, incorrect measurements, poor resolution, not measuring relevant aspects are also measurements.
They may appear useful but may not be.

For audio we need to divide measurements in at least 4 different realms before saying something about measurements and their usefulness/accuracy.

• Electric domain
• Mechanical domain
• Acoustic domain
• Perception

The above list is in order of difficulty and accuracy for measurements.
Some are really easy to do with high accuracy where other measurements are extremely complex and involve human perception.

You could even break perception into biological perception and mental perception.

Biological being due to our hardware. Mental being learned or other emotional responses.

 

[AJM1981]

It all can be measured. Just have to know what to measure and how?

True. Also the "why" question. Hi-fi is not made for labs or studios, its made for homes. It is made to compliment furniture and personality. A brand like KEF s known for its cooler kind of sound. But also it seems optically designed to match cool / clinical interiors. Speaker manufacturers like Wharfedale seem to compliment warmer interiors and their products are also sonically often described as such. Consumers associate it with warmth, a cool sounding speaker would be off-key.

Companies don't have a so called "house sound" label for no reason with specific amp models and their own take on that. All a result of certain design philosophies, damping factor, form factor etc. Now, when combining an amp and a room and a loudspeaker. These multiple parameters per device (and the room) combined provide kind of a unique outcome. No one should be interested in what I or a reviewer thinks a set sounds under these conditions. But well..

It is still all measurable, but the question is, is it interesting enough to do so when parameters combined kind of provide their unique parameters. Personally I think such sets of data would fall into a void of tldr by most and a desire to choose for what we think is best.

Therefore I like the backstories that are in my opinion more interesting than data. The why behind a manufacturers choice in a certain era. What was the competion doing? why something stood out for its time? Why something failed. Something like the Vfets used by Yamaha and Sony was kind of the middle ground split between vacuum tubes and solid state. Literally best of both eras, but also a complicated solution to challenges perhaps solved in a better way in later times.

 

[Timcognito]

It is very simple. It is not a preference thing, it is an accuracy thing. What is accuracy? Simple the sound wave out matches the the signal in that was captured by the microphone. If one has DSP and an accurate speaker that speaker can be tuned to your preference and also eliminate many of the absorbed, reflected and augmented parts of the frequency range to restore the desired sound profile. So many things in life we accept the measurements, the pressure in our tires, the weather report, oven temperature, your fuel gauge or when your doctor tells you your EKG shows an arrhythmia, do you say are you sure, I don't trust those measurements. Yet somehow despite many years of development and scientific research correlating sound output of transducer to the input signal all of sudden the science is flawed. Everything that is audible is captured on a calibrated microphone, can be displayed graphically and baked into those measurements, is all that you are hearing. Reading and understanding those measurements takes some learning and some things are shades of gray and debatable but it's all there. Accurate and neutral may not ones preference but it acts a control that the transducer is replicating the sound captured by the microphone. No speaker is perfect and distortion exists but the biggest flaws in an accurate system come from how the content was recorded and mixed.

 

[pablolie]

It is very simple. It is not a preference thing, it is an accuracy thing. What is accuracy? Simple the sound wave out matches the the signal in that was captured by the microphone. If one has DSP and an accurate speaker that speaker can be tuned to your preference and also eliminate many of the absorbed, reflected and augmented parts of the frequency range to restore the desired sound profile. So many things in life we accept the measurements, the pressure in our tires, the weather report, oven temperature, your fuel gauge or when your doctor tells you your EKG shows an arrhythmia, do you say are you sure, I don't trust those measurements. Yet somehow despite many years of development and scientific research correlating sound output of transducer to the input signal all of sudden the science is flawed. Everything that is audible is captured on a calibrated microphone, can be displayed graphically and baked into those measurements, is all that you are hearing. Reading and understanding those measurements takes some learning and some things are shades of gray and debatable but it's all there. Accurate and neutral may not ones preference but it acts a control that the transducer is replicating the sound captured by the microphone. No speaker is perfect and distortion exists but the biggest flaws in an accurate system come from how the content was recorded and mixed.

Heart arrithmia is a good example. Would anyone go on a crusade against the limited measurement capabilities of the device... or simply do something about the things that result seems to require? I'd hope the latter.
Performance measurements are targeted to reveal the capabilities of a device in audio. Pleasure measurements are irrelevant. Learn to check for the performance measurements you actually enjoy, and that'll help you pick devices tailored to your requirements from a sound profile perspective. No one wants to force anyone to always pick the most linear, highest SINAD etc box in the world. I know I am not geared that way.
And the tradeoff between performance -tailored to your preferences if you learn to read measurements- and looks is yours to decide.

 

[kemmler3D]

I'm sort of guessing, since everyone I know is pretty rational, and for that matter they're engineers, scientists and lawyers. But to a non-scientific person all this stuff is pretty mysterious and well, quasi-mystical. So, easy prey for the Shunyatas and AudioQuests of the world, who are long on flowery pseudo-technical explanations largely devoid of meaning. Worse yet, if they're in certain forums, they are essentially immersed in a cult. And those forums have a lot of members.

Agree. Thinking back to when I was a kid, it was pretty brain-breaking that so many sounds, each with depth, breadth, and motion, could come out of a single speaker cone at the same time, driven by a single signal. It's still kind of crazy to think that we perceive so much at once, just from one squiggly line.

If you don't know which way is up when it comes to reproducing audio it's really easy to be impressed by things you don't understand. The "simple" process of turning digital audio into hi-fi sound feels unimaginably complex. Like how does the DAC know what to do when the trumpets and drums are going at the same time? It sounds totally dumb to us, but if you actually don't know anything about audio that's how it is.

I'm pretty sure that before I knew better, I believed every silly marketing name for a bass reflex design (like power bass X or whatever) was genuinely new technology. Imagine if I'd had $40k to spend on gear back then.

 

[fpitas]

Agree. Thinking back to when I was a kid, it was pretty brain-breaking that so many sounds, each with depth, breadth, and motion, could come out of a single speaker cone at the same time, driven by a single signal. It's still kind of crazy to think that we perceive so much at once, just from one squiggly line.

If you don't know which way is up when it comes to reproducing audio it's really easy to be impressed by things you don't understand. The simple process of turning digital audio into hi-fi sound feels unimaginably complex. Like how does the DAC know what to do when the trumpets and drums are going at the same time? It sounds totally dumb to us, but if you actually don't know anything about audio that's how it is.

I'm pretty sure that before I knew better, I believed every silly marketing name for a bass reflex design (like power bass X or whatever) was genuinely new technology. Imagine if I'd had $40k to spend on gear back then.

Some very smart guys have been developing audio for over a hundred years now. There's no wonder it's daunting to a non-technical sort.

 

[Cote Dazur]

What I find daunting is the constant vociferous barrage of inexact information (either by ignorance or malice) that those people are exposed to. The constant noise is drowning any access to understanding what matters and what does not matter.
It takes years of Herculean efforts and a strong determination to make any sense of what make audio great.

 

[pablolie]

Some very smart guys have been developing audio for over a hundred years now. There's no wonder it's daunting to a non-technical sort.

The engineering basics are pretty straightforward to understand. The implementation thereof and the advancements in material science and tuning and optimizing are beyond my comprehension (and I have an advanced engineering degree, so why would I lie?). My specialization is elsewhere, and any engineer that claims to be a stellar universal multiple use case expert is either mediocre or a liar.
But it's important to understand the correlation between measurement capabilities that didn't exist when everything was done by ear, and the here and now. Obviously a divide is growing between designers that do things the artisan way, and those who consult their Kippel.

 

[solderdude]

What I find daunting is the constant vociferous barrage of inexact information (either by ignorance or malice) that those people are exposed to. The constant noise is drowning any access to understanding what matters and what does not matter.
It takes years of Herculean efforts and a strong determination to make any sense of what make audio great.

That's the main thing.

From the huge amount of 'information' out there the biggest part of it is misinformation and only some small part is factually correct information.
It is nigh-on impossible for the public that does not have a lot of (real) knowledge of electronics, transducers, recording, (psycho)acoustics to filter out what 'information' is correct and what is misinformation.
One really cannot fault 'the interested readers' for this.

AI is not helping here at all either unless you prompt it correctly otherwise it just generates even more nonsense disguised as 'facts'.

 

[wilcaz]

AI is not helping here at all either unless you prompt it correctly otherwise it just generates even more nonsense disguised as 'facts'.

Indeed. When I tried using Gemini for amp recommendations it told me I should get a class AB amp for its "rhythmic drive and warm authority" over class D amps lol

 

[Ropeburn]

Indeed. When I tried using Gemini for amp recommendations it told me I should get a class AB amp for its "rhythmic drive and warm authority" over class D amps lol

That's how you describe women, not amplifiers

 

[wilcaz]

That's how you describe women, not amplifiers

I personally prefer class D to class A/B. DD even better.

 

[Ropeburn]

I personally prefer class D to class A/B. DD even better.

Kinda reverse huh. Suddenly class D is large, heavy and impractical and can lead to strain...

 

[fpitas]

From the huge amount of 'information' out there the biggest part of it is misinformation and only some small part is factually correct information

Unfortunately, that's how the internet as a whole is turning out. AI seems to dredge up misinformation and make matter worse.

 

[Ropeburn]

I personally prefer class D to class A/B. DD even better.

Oh and, there's a reason class AB is so common. It's practical enough and still performs perfectly fine...