'cause there is no one ear! I can play tracks for you where I can hear compression artifacts and you might not. By measuring, we know what exists as a difference absolutely.
I don't think that I ever argued that we all hear the same. My point is that ultimately what is important if I (Meaning the human race) can't hear it, it does not matter. We don't design products for the hard of hearing, old guys like me, nor that that thirteen year old girl with hearing like a bat. So measurements are important to the designers but not to me(end user). All that is important is what I can hear,
Further, if an instrument shows there is no difference, you cannot use your ear and says there
(is?) isn't. If I play two tracks and I difference them and it shows zero, you cannot tell me your ear says otherwise. Unfortunately it is entirely possible for your ears to mistakenly say there is a difference.
I'm not going to fall into the null test argument. If I can paraphrase; if I'm hearing something inconsistent with the measurement then you must be measuring the wrong thing. If I hear something that's not there. it's not my ear that is mistaken. It's my brains perception that is wrong.
So the notion that the ear is better than an instrument is false on multiple grounds. So is the assertion that if we measure it, it is audible.
Unfortunately for you science guys human perception is the most important thing. The reason for that is you are trying to create the illusion inside my head. Measurements are only useful to create a machine to try and create that illusion. That is why the ear is best. You can take a machine and make a gazillion colors. My eyes have the last word. Yeah you can trick my eyes. But that's an aberration. Any true test has to assume the ear is working properly. If the ear is malfunctioning (being tricked) the test is invalid
Where this leaves us is that both extremes of this discussion are wrong. You cannot stick 100% to one side or the other without being subject to challenge.
Most absolutes are wrong
I don't see saying the "ear is best under the totality of circumstances" as an extreme. I see that as a moderate statement leaving open the notion that machines do many things better when not acting under the totality of the circumstances. The ear does not have to be perfect because whatever the ear does is right. (It's kinda like a woman.) It is what as it is. The machines have to be perfect. We humans are subjective by nature. Whatever the ear hears is what we have to deal with. That's right the ear wags the machines. The machines don't wag the ears.
It absolutely is not academic. Measurements can be ultra fast and ultra reliable. Such is not the case with listening tests. I can do 1000 measurements overnight to figure an optimum design for something. You can't possibly conduct the same number of listening tests even if you decided to not sleep overnight
.
Machines are notoriously bad at measuring real world performance. They do excel at repetitive tasks and computations. most measurements require human interpretation. Additionally applying those measurements to real world situation require some "artistic ability. Trying cooking from a recipe. What I can do is take my time and see if the effect wears off because of listening fatigue.Again our machines have to make hear music. The measurements assist the machine not the ear.
Everything in audio needs to start with a measurement. Listening tests is confidence builder. Anyone who is designing equipment by ear alone is not someone I will give a penny to. Thankfully, hardly anyone does that, even though their marketing material scuffs at the idea of measurements having value.