Originally Posted by Stephan Mire
Just a question that has always intrigued me for the, shall I call them Measurers?
Purely hypothetical: The specs tell you some piece of kit should sound crap and it does not; or, all your measurements, graphs and numbers tell you that something should sound fantastic and it does not... Which do you believe?
Or is this just that: a purely hypothetical scenario that does not compute in your world?
First of all, specifications are simply claims made by the manufacturer. They can claim anything (though some claims might get them some legal trouble). Typically, reputable manufacturers make claims that are less than what the equipment is likely to do. For example, I have an old Pioneer SX-1250 receiver that is rated at 160 watts per channel, 20-20kHz, @ 8 ohms with no more than 0.1% THD. After it was more than a decade old, without anything having been done to it, it was measured in a McIntosh Clinic (they used to travel about to audio shops and test people's gear, in an effort to explain to them why a McIntosh amp would be better). Because it was easier to do, it was measured putting out 200 watts into 8 ohms, 20-20kHz, with no more than 0.055% THD. So, being more than a decade old, it exceeded its power ratings with just barely more than half of the rated distortion. That is the sort of thing that occurs with reputable companies, that the piece of equipment typically will perform better than the specifications.
Now, in the case of something measuring well, if one forgot one important measurement, then it measuring great in all of the ways that one measured it, that will not tell one that it will sound good, because one forgot one important measurement that might turn out bad.
If I encountered a case where I thought that everything important was measured, and the results suggested it would sound great, but it sounded like crap, I would want to look into whether there was an error in one of the measurements, or consider the possibility that there was something important that was neglected and not measured. What specific thing I would look for first would depend on how it sounded bad (e.g., deficient in bass, distorted, etc.). It could also be that the gear is now malfunctioning, but was working fine when it was measured. And so it would need to be remeasured.
Now, do you have any examples of this, where there is something that has measured well in all important ways, but sounds bad, and sounds bad when compared with things that sound good, in level matched double blind tests?
My guess of the sort of thing that might come up in real life would be something like looking at the frequency response of a speaker, that is rated 33Hz–17kHz +/-4dB, and someone imagines that that must not be very good. That, however, would be a mistake, because there is little music outside that range, and human hearing is not overly sensitive to those frequencies outside of that range (not to mention the fact that many people cannot hear that high anyway), and the variation from any speaker in your room
is likely to be greater than +/-4dB anyway. So, I can easily imagine a case where someone imagines that the numbers are bad, when they are not that bad at all.
Of course, that would be a "bad" rating for an amplifier, but its badness in such a case is because it is easy to get much better performance from an amplifier than that, though it might sound okay in practice. But I would not buy such an amplifier, as it might not sound as good as another, and it is easy and cheap to get something that performs better than that.
(As an aside, that frequency response specification is of a speaker that retails for about $12000/pair. It is not my favorite speaker, but I would not describe it as crap either.)