Master Thread: Are measurements Everything or Nothing?

[uwotm8]

That is what happens when someone gives credibility to personal opinions as if they were valuable for others. They're not. Personal opinions are varied, they are changeable, they are unfounded and unsupportable without data to back them up, and they are not reproducible. (Follow the hyperlink.)
The only meaningful information about "things" (cell phones, audio speakers, airplanes or hydraulic seals) is to test them and measure them

Reproduction seem to not working as good as intended if some fraction of subjectiveness is involved, once we're talking about things that are more difficult to measure than fps, screen brightness or data transfer speed. It is enough to look at the camera tests of the smartphones you mentioned. And "normal" digital cameras as well. Once a certain acceptable level of noise and dynamic range, as well as color accuracy - I would compare this in importance with the frequency response and distortion - completely different things become a factor in choosing, such as behavior in difficult lighting conditions, subtle technical shortcomings and especially the subjective perception of the result and what kind of "frequency response distortions" the cameras introduce. Oddly enough, Sony or Nikon, which are very accurate in test color chart shots, are still ready to surprise in difficult or uneven conditions with strange colors and questionable skin tones, the pleasant rendering of which is still the reason for choosing Canon for a lot of people. Moreover, not everything can be corrected in the post-processing. In the case of smartphones, the repeatability of the good chart result in a field is where some models struggle nowadays: products of young brands can be equipped with excellent sensors, but the software part makes them show strange things in non-ideal situations. That's why expertise and field tests are very important here. Which is what the DxOmark does. Despite how high their sensor rating mark is, it's way better to check the photos and see what you'll be dealing with and if you want to. And also to figure out if there's really much progress or you should as always divide the rave around it by 10.

A different approach would be declaring that precision on test charts under controlled light is all that matters, nobody cares about how some ugly faces look in a rainy day, just shut up pleb, this is science

Oh, you mentioned airplanes. I'm not a pilot and never will be yet just from a common knowledge - aren't aviation rules and principles written by blood? Aren't those perfect mathematical plane models tested by people and gets cancelled after billions and billions of investments when it appears that science didn't take to account of some factors or some 3rd part manufacturer failed making bolts to specification? It continues to happen nowadays. And how actually do you measure real, say, maneurability in numbers? It doesn't matter however that aviation should be cancelled, exactly the opposite - year by year science gets improved and concepts get overlooked.

It was not so far ago when the whole idea of a plane or nucler reaction was a heresy

 

[Audionaut]

To recap .... the opinions of the gear are not the same, person to person. That makes them useless for transmitting information accurately from one person to another. The only way to present information that is equally accurate from one person to another (and for all people concerned) is to use a science-based format that rejects personal opinion.

This is only true for some areas of science.
In others, it is certainly possible to use iteration, hermeneutics, deduction and induction to approximate human-oriented sciences.

When it comes to how someone reacts psychologically to a drug, how they judge certain situations, how they judge a taste, how they judge the sound of an automobile door slamming, etc., you can only make progress with humanistically oriented, qualitative data analysis.

If the data collection group is large enough to achieve significance, the data is scientifically valid.

 

[Newman]

Edit: I've just deleted two posts. It seems I have allowed myself to be manipulated, in my gullibility, by a clever little troll. I apologize to all our members for wasting their time.

I hope you reported it. Manipulative baiting posts are in bad faith and the mods usually view that very dimly.

Especially when, as is usual in your case, your wasted effort is science-based, educational, and very calm.

 

[Megaken]

I just know that I've heard a few systems that measure excellent that sound ok, and a few systems that measure so-so and sound excellent

 

[dshreter]

I just know that I've heard a few systems that measure excellent that sound ok, and a few systems that measure so-so and sound excellent

Maybe you weren’t measuring the right things

 

[AdamG]

Manipulative baiting posts are in bad faith and the mods usually view that very dimly.

We have seen enough…

 

[MAB]

This is only true for some areas of science.
In others, it is certainly possible to use iteration, hermeneutics, deduction and induction to approximate human-oriented sciences.

When it comes to how someone reacts psychologically to a drug, how they judge certain situations, how they judge a taste, how they judge the sound of an automobile door slamming, etc., you can only make progress with humanistically oriented, qualitative data analysis.

If the data collection group is large enough to achieve significance, the data is scientifically valid.

Perhaps you need to cite this, with some context. Specifically examples of drug side effects or primary effects, and those specific situations that preclude quantitative evaluation and comparison. And lastly, how those situations and studies relate to audio reproduction. Thanks.

 

[ahofer]

Maybe you weren’t measuring the right things

He’s ‘taking the piss’, no?

 

[dshreter]

He’s ‘taking the piss’, no?

Not sure, but I suspect not.

 

[Megaken]

Maybe you weren’t measuring the right things

I'm not talking about my measurements I'm talking about Amirs etc

 

[solderdude]

If the data collection group is large enough to achieve significance, the data is scientifically valid.

Do you mean if a large (and vocal) group of people hear differences in op-amps or say interlinks, power cables or even audiophile vs 'run of the mill' fuses.

1. people can hear things that aren't real but appears convincing because perception is a tricky thing
2. people can hear things like differences in op-amps, interlink or power cables and fuses thus they actually sound different because there are that many reports so it must be true and not everything can be shown in measurements.

An analogy for this is if all people see a static picture 'move' or see 'curved lines' or 'depth' in a 2 dimensional picture (like your avatar) when in fact lines are straight or a picture is 2-dimensional then one may conclude that.

1. people can see things that aren't real because perception can be 'flawed'
2. people see static pictures 'move' or 'depth' and thus the picture may actually move or have 3D properties

or.... measurements are everything or nothing

1. measurements are indicative for technical performance and thus have a certain value
2. Many people report clear discrepancies between heard sound quality and measurements and music enjoyment is what you hear so measurements say little to nothing

What is clear is that when a collection group is large enough to appear to have a significance you can bend that 'data' to suit a narrative.

 

[Beave]

I just know that I've heard a few systems that measure excellent that sound ok, and a few systems that measure so-so and sound excellent

Wanna share those specific examples you have encountered?

1. The room and setup are always factors that can make even good gear sound not so good.

2. Sighted listening isn't a very reliable indicator of good sound - except to that listener at that particular time and at that particular location.

So maybe it was the setup. Maybe the room. Maybe the music being played. Maybe the lighting in the room. Maybe your mood.

 

[Megaken]

Hegel stuff and Ps audio equipment don't measure well, but ask anyone who owns them - they love how they sound

 

[Beave]

Hegel stuff and Ps audio equipment don't measure well, but ask anyone who owns them - they love how they sound

That's just about meaningless. One, although they may not measure as well as some other electronics, they measure well enough to be audibly transparent, or very near so. Two, it's not at all surprising that people who bought a product like it. They probably didn't do any controlled testing before buying the product, and even if they did, refer to my previous point.

 

[Mart68]

Hegel stuff and Ps audio equipment don't measure well, but ask anyone who owns them - they love how they sound

You're not taking audibility thresholds into account. Both are good enough to sound fine.

Almost everyone buys new equipment and gushes over it - it's called the honeymoon period. The Hegel or PS Audio amp or DAC is likely sounding no different from whatever they had before. People don't do proper listening comparisons. That's the root of it.

 

[Jimster480]

I just know that I've heard a few systems that measure excellent that sound ok, and a few systems that measure so-so and sound excellent

That is because music itself doesn't always need all the measurements. People forget that lots of albums have extremely low dynamic ranges. Lots of music wasn't recorded well. Part of the reason why some people hate well measuring setups is because those setups can reveal how badly recorded their favorite music is.

Also when it comes to measurements; not EVERYTHING is measured always. It is near impossible to measure every scenario with a track and see the reproduction of that music in every single waveform across the entire system. This is why there is so much subjectivity in the world of audio and continues to be. Since people can purchase something that measures well but not like how it sounds.
I find that the measurements of SINAD and Linearity and Multitone provide a good guide on the actual available clarity in music with the right speakers or headphones.... but lets not forget that speakers themselves have their own measurements, as do headphones. Again both of those tests have variability and testing conditions, and not every condition can be tested.
If everything was so cut and dry; there wouldn't be so many audio companies.

 

[dshreter]

The most important part of a system is the room. So it’s also expected that there are many lousy rooms with well-measuring components that still don’t sound very good.

The same is true of the converse. Most electronics that are considered hifi will sound good in a good room.

 

[MattHooper]

I just stumbled across this old article: a visit to Bill Ying’s place, of Shun Mook.

Most of us old timers recognize Shun Mook as one of the pinnacles of nonsense tweakery. Their notorious mpingo discs and all that.

For those who dare to read or skim it, you can see the rabbit hole somebody can go down on this. Once you think that resonances affect everything, you can reach the point of madness. The steps he has taken get fairly jaw-dropping.

I mean it gets to the point where he “ tested” all sorts of newsprint for its Sonic virtues and he ended up with “Civil War-era newspaper attached to the face of his fireplace.”

The Audio Beat - East Meets West Coast; Bill Ying of Shun Mook Audio

read audio reviews, ratings, and user opinions on audio equipment before you buy an audio system

www.theaudiobeat.com

Now the main reason I even bring this up is this, and I have mentioned this before:

Despite this level of craziness, these people still manage to create one of the most beautiful sounding loudspeakers I’ve ever heard! I don’t even know if they used measurements at all… I remember, they claimed they listened to the resonant quality of the wood used for the cabinet, and that they were mpingo discs within the cabinet tuning it and all that. I have a hard time imaging that they didn’t use measurements at all, but they emphasized a tuned by ear approach.

This is one reason why I don’t write some audio gear off only on the basis that it is made by or praised by someone who believes in some nonsense. Because even people who believe in an extreme level of nonsense like these guys managed to produce something really compelling,

I don’t think for a moment that the quality of the sound from their speaker justified the nuttier implausible aspects that went into the design. Rather, even amidst some level of delusion, it looks like some people can actually still have a good pair of ears and create something interesting.

So I totally understand most ASR members would look at a crazy set up like the above and run the other way, especially when it comes to loudspeakers, I’d say “ OK let me hear it, let’s see what you can actually do.”
And I’ve had a number of pleasant surprises down the years.

 

[Mart68]

The loudspeakers look pretty conventional to me - big ported three-way - they usually sound good unless the design is badly botched.

 

[Purité Audio]

Amazing how far uncorroborated listening can take one, you start by adding add one suspended violin….
Keith