Measurements all aim at laboratory standards, royally ignoring the requirements of psychoacoustics. That is, we do not measure sound quality as perceived by the human brain. We measure reliability.I'm going to contest this. The measurements aren't the problem- we can easily measure what we need to (at least insofar as electronics are concerned) to know that the electronics will have both good distortion figures **and** also sound musical.
But for the most part the industry does not do that. As a result the spec sheet becomes the Emperor's New Clothes.
Often super low THD is associated with a 'neutral' amplifier that sounds bad. When this is the case there's been a bit of slight-of-hand.
If you know of an example of this, look at the **frequency** at which the distortion measurements were made. You'll see that its only 100Hz. This is important- at 100 Hz most solid state amps made in the last 60 years have enough feedback to allow for excellent reproduction. This is why solid state amps have had 'good bass' for so long. But those same amps otherwise sound bad because they are bright and harsh. This is also why tubes are still around and it comes down to the same thing:
Those amps have increased distortion at 1KHz or 10KHz!
IOW the distortion production rises with frequency. The ear interprets the higher ordered harmonics as brightness and harshness; if the 7th is present in enough amplitude from a 1KHz fundamental the amp is going to sound bright because the ear assigns brightness and harshness to the 7th harmonic (as well as other higher orders).
The slight of hand is that to get around this problem the industry measures at 100Hz and calls it good.
So if you want to hear an amp that has both low distortion and sounds like music, look for one in which the distortion at 100Hz, 1KHz and 10KHz is exactly or nearly the same. Manufacturers don't usually show that in their measurements but they should. One reason that they have not is an engineering problem called Gain Bandwidth Product. GBP is to feedback in electronics what gas is to a car. When you run out of gas the car stops. When your GBP is insufficient, the feedback goes down. Because most of the amplifiers, tube or solid state, made in the last 90 years have had insufficient GBP, they have also had insufficient feedback at the frequencies *where it counts to the human ear*. This is why feedback has gotten a bad rap in high end audio, and why a good number of manufacturers have made zero feedback amplifiers. A zero feedback amp gets around this problem of variable feedback by having none- in this way, as long as its got the bandwidth (IOW well past 20KHz), its distortion will be exactly the same at 100Hz as it is at 10KHz.
So if the spectra of the distortion is correct (IOW the 2nd and/or 3rd is prominent enough to mask the higher orders from the ear), a zero feedback amp can be quite musical, even though its distortion is 'high' (THD numbers don't tell the story because to the ear the 2nd and 3rd are nearly inaudible while the ear is keenly sensitive to the higher orders). But this is also true if the amp has feedback, if it has enough GBP to support the feedback properly up to 20KHz. Quite simply it was probably not until sometime in the late 1980s before the semiconductors needed to pull this off even existed. It was some time later before the will to design an amplifier with this vital character of distortion was designed.
So its no surprise that people still think that excellent specs and really 3d musical sound are mutually exclusive because the examples that disprove this are a minority in the high end audio world. We certainly have the ability to measure what's important- although the will to do so is an entirely different matter!