Master Thread: Are measurements Everything or Nothing?

[CleanSound]

Well ignoring info that goes against our beliefs is one way to go about things. I prefer to trust but verify. No need to talk about the video anymore since nether of us can retest Danny in person.

I did addressed that video, therefore I did not ignore it. Despite that video makes me sick to my stomach that such charlatanry exist.

You said trust but verified, then why would you reference to this video if you haven't verified it yet? Legit and honest question. The authenticity of the test in that video is not verifiable because none of us were there and that is by design. If Danny Ritchie wants a real Blind AB, I am sure Amir or anyone else on this forum can arrange it with neutral parties verifying the legitimacy of the test.

Do you have any thoughts on the rest of my post, more specifically, “What if we aren’t measuring everything? Could there be something involved with the whole complex system of amplification and signal delivery that we’re missing?”

I'm not sure how long have you been in this HiFi hobby, but this ("what if we aren't measuring everything") has been going on for decades now.
1) First there has to be an scientifically observable difference before we can investigate what/how to measure something that we don't yet know about. Therefore first, provide me such evidence that such indeed difference exists. "Evidence, are you kidding me, half of the audiophile world are all saying they hear this x-factor." If that is your evidence that we should just go back to the stone ages now. No one has ever passed any blind AB test under verifiable legitimate environment. Have someone with the scientific and engineering knowledge, neutral party, with verifiable credentials host such a test and invite the general audiophile public to such blind AB event to prove such difference, then we can further investigate.

2) Dr. Toole has already done research and showed us what is that we "aren't measuring." He figured it out and published extensively about it. That is cognitive bias.

3) I go back, if we can't measure it, or have no clue on how to measure it, then how do you build it? How does an amp designer know that some people like "air," "warmth" and to add some, if s/he does not know much much air and warmth exist and how much to add?

Nope. I don’t hear a difference, but I believe there are others out there with better hearing than me that can hear a difference. I can’t call them liars just because I’m not as skilled, but I can ask questions to spur thought from others and point me to directions of research. It seems the more I learn, the less I really know…someone smarter than me said something similar back in the day.

The good old "golden ear" trick. I am sorry my friend, but you've been tricked by charlatans.

If you passed a hearing test and you trained your ears (using an application like Harman's How to Listen, only available on Windows), then your hearing is as good as any ones. Charlatans have made claims that they have a better hearing then me. When I was much younger, I didn't know better, I kind of believe them. But now I laugh at such claims, because I have passed a hearing test and spend a considerable amount of time on the Harman How to Listen app (I must have logged about 20 plus hours on it over the years). And I am able (like Amir and many others on ASR) to differentiate the difference between CD and hi res (it's not what you think and you can search for this topic on this forum). EDIT: So any charlatan, of the likes of Danny Ritchie or Jay Lee, dares to claim that somehow they have a golden ear or my hearing is insufficient, I am more than happy to take any challenges they can throw at me. You too, don't let any charlatan tell you they have a golden ear and you don't. That is complete utter BS.

 

[CapMan]

Black swan event ??

Black swan theory - Wikipedia

en.wikipedia.org

 

[Jimbob54]

Serious objectivists simply don't have the guts to stop lying to themselves (and others) and face facts about audio equipment.

Best change that toot sweet before the pitchforks get sharpened.

 

[CleanSound]

But just think how boring it would be to live in a world where there weren't fools who sell unicorns to others like them.

Very true. And very sad at the same time.

 

[kemmler3D]

Best change that toot sweet before the pitchforks get sharpened.

When I used to draw cartoons for the college paper, I would say "If you don't get hate mail, you haven't said anything at all." I still believe that.

 

[Sgt. Ear Ache]

When I used to draw cartoons for the college paper, I would say "If you don't get hate mail, you haven't said anything at all." I still believe that.

you meant to say "serious subjectivists" though correct?

 

[anome]

So,that was not a brain fart.

 

[kemmler3D]

Do you have any thoughts on the rest of my post, more specifically, “What if we aren’t measuring everything? Could there be something involved with the whole complex system of amplification and signal delivery that we’re missing?”

Microphones and recording equipment are more sensitive than the human ear.

Also, recording and mixing audio is no different on an engineering basis than the process of measuring audio gear.

As such, if it can't be measured, it can't be put into a recording.

So if there is something audible, but not measurable in principle, then by definition it's not in the recording - i.e. it's distortion, again by definition.

Is there distortion in DACs or Amps that we can't measure? I really don't think so, we can compare even very complex signals quite accurately with software like Deltawave, we can measure gear with a lot more sensitivity than the ear, etc.

Is there distortion that speakers produce that is not measurable in principle, but can be heard? How would that even be possible? The ear and microphones work on the same basic principle, a membrane being vibrated by the air. So if your ear can pick it up but a microphone can't, what are we even talking about? Sound waves that aren't sound waves?

Also... consider the fact that audio signals aren't even a big challenge for measurement. We have MHz and GHz signals being measured for computers and telecom equipment out there, what is distortion at 400hz to equipment like that?

Not every test measures every relevant variable - true. But if you hear something, you can in theory track down the cause via measurements. It's the people who dispute this who are fundamentally out of whack with reality.

 

[Killingbeans]

Do you have any thoughts on the rest of my post, more specifically, “What if we aren’t measuring everything? Could there be something involved with the whole complex system of amplification and signal delivery that we’re missing?”

The particle detectors in the Large Hadron Collider are probably one of the most sensitive examples of amplification and delivery of signals. I bet the people who built it would laugh their asses off when being told about something "missing" in audio signal processing.

I don't want to steer this thread into one of the big no-no discussions, but the 'what if you're wrong?' argument is, in my mind, reserved for those who are desperate for legitimizing magical thinking.

 

[kemmler3D]

you meant to say "serious subjectivists" though correct?

Yes, naturally, d'oh!

 

[kemmler3D]

Best change that toot sweet before the pitchforks get sharpened.

Yeah that was a typo.

 

[Jimbob54]

Yeah that was a typo.

Nice knowing you!

 

[Saponetto]

Fearing another poll knocking on ASR doors...

 

[Frank2]

I'm not a subjectivist nor an experienced electronics engineer. But I do have two serious questions related to this debate:

1. Measurements are often done with one frequency at a time (frequency sweeps and harmonic distortion) or sometimes two frequencies simultaneously (intermodulation distortion). A real music signal is a complex mix of frequencies and since audio electronics is not perfectly linear, these frequencies will influence each other. How well do the usual measurements represent the behavior of the electronics when processing real music?

2. Amplifier output measurements are often performed with a fixed resistive load. How well do these measurements represent the amplifier's behavior on a more complex speaker impedance?

I would really like to learn more about this from some of the experienced electronics engineers on thus forum.

 

[solderdude]

1. Measurements are often done with one frequency at a time (frequency sweeps and harmonic distortion) or sometimes two frequencies simultaneously (intermodulation distortion). A real music signal is a complex mix of frequencies and since audio electronics is not perfectly linear, these frequencies will influence each other. How well do the usual measurements represent the behavior of the electronics when processing real music?

Multitone and noise come pretty close in actual waveform.
Besides... there is nulling which can use music and actual loads but only in the electrical domain and have to take phase variances and time delays into account.

2. Amplifier output measurements are often performed with a fixed resistive load. How well do these measurements represent the amplifier's behavior on a more complex speaker impedance?

This is indeed the case. Certainly with speaker amps and headphone amps with a highish output resistance this can be quite audible but also is measurable.
There are also tests showing inductive and capacitive loads and complex loads exist.
This is often used to make some designs to sound different yet would not appear to do so in simple resistive load FR measurements.

So also measurable BUT not every reviewer/measurebator does this. This is what may be confusing for those that have little to no knowledge of measurements/protocols and electronics which is ... the vast majority of people on this planet .

 

[CleanSound]

I'm not a subjectivist nor an experienced electronics engineer. But I do have two serious questions related to this debate:

1. Measurements are often done with one frequency at a time (frequency sweeps and harmonic distortion) or sometimes two frequencies simultaneously (intermodulation distortion). A real music signal is a complex mix of frequencies and since audio electronics is not perfectly linear, these frequencies will influence each other. How well do the usual measurements represent the behavior of the electronics when processing real music?

2. Amplifier output measurements are often performed with a fixed resistive load. How well do these measurements represent the amplifier's behavior on a more complex speaker impedance?

I would really like to learn more about this from some of the experienced electronics engineers on thus forum.

These are VERY GOOD questions. In fact, these are probably the best, legit and most thoughtful questions so far. I am not trying to take credit, but I ponder on these questions myself. The other question I have that runs along the same line is, actual continuous music signal over a period, basically taking time domain into the mix, such as the raise and fall time of signal waves.

I don't have an AP, never have used one, so I don't what what an AP is capable of doing. So anyone here who is an analog engineer and knows the AP well, please comment.

HOWEVER, I go back to saying: If you can't measure it, how can you engineer it? If you can't measure what is wrong or what is right, then how do I know how wrong or how right it is and how much right I have to make it? This logic alone, is an universal logic to cover any what ifism or other very good questions.

 

[kemmler3D]

actual continuous music signal over a period, basically taking time domain into the mix, such as the raise and fall time of signal waves.

Frequency / phase response and compression tests tell you a lot about this.

 

[solderdude]

The other question I have that runs along the same line is, actual continuous music signal over a period, basically taking time domain into the mix, such as the raise and fall time of signal waves.

Note that this is an issue for transducers and not for electronics unless one is near the Nyquist of digital systems. Very predictable and measurable for electronics.
Rise and fall times is bandwidth limiting and thus also output level dependent. Also predictable, simulatable and measurable.

 

[Sokel]

I think the whole "differences" thing is a little insane.
When I listen to stuff I don't search to hear "differences" ,I want to hear if it's better.
Louder,cleaner,bigger (specially this)etc,big time.

I found this only with speakers,and I don't mean similar ones.
As it is it's only about size and power to me,everything else is minor to the point of non existing the next day.

 

[Jim Taylor]

First of all, there is nothing particularly shocking or impossible about getting 8 out of ten right in what is essentially a coin flip.

Correct. Theoretically, a monkey could get 10 out of 10 correct on the first try out of a million.

Jim