[taymat]

Originally Posted by Sycraft

But here's the thing: Currently, the empiricists (of which I am one, have my copy of The Logic of Scientific Discovery and everything) rely almost exclusively on scopes, FFTs and so on. Objective measurements, in other words. The reason is because we know that people lie, and even when they don't their perception isn't absolute. Measurements are, my scope never lies to me (provided it's calibrated).

Ok, great, however what we empiricists do NOT know is if we have discovered everything that is important to human hearing. We know frequency response is, so we measure that. We know noise is, so we measure that, etc. But, what if there is a property we DON'T measure, beacuse we aren't aware of or believe it is unimportant, that is important to perception? Well that would mean, despite our claims that two things sound the same because they are identicle on the scope, we are wrong.

One possible example would be the nature of distorion. All I ever see measured is the amount, never the kind. I don't know enough about audio distortion to say for sure, but I would infer that distortion does not distort all harmonics of a note evenly. Thus the distortion itself will have a harmonic shape, just as the note does. The difference between a trumpet at A440 and a claranet at A440 isn't the pitch, the fundimental is 440Hz in both cases. Nor is the difference in the frequency of the harmonics, they are all integer multiples of 440Hz. The difference is in the relitive amplitudes of the harmonics, which are loud, which are soft. THAT is what makes sound to a human, and how we identify speech and so on.

So, maybe the shape of a device's distortion (if indeed distortion has a shape as I postulate) is something that effects our perception. We hear two things with the same THD, but one has a shape that sounds better than the other.

I'm not saying this is for sure that it exists or if it does that it matters, I am simply using it as an example of something that we do not attempt to measure and figure out what is good and bad.

Well, rather than just chase my tail around, the research I think would be interesting to do would be to take two things that are widely regarded to have good, but different sound that ought not, like the op amps, and do a proper, multi-level test on them. If the study shows that there probably IS a difference in sound, I'd then want to do more research to try and find out what causes that difference.

Or maybe (more likely I believe) I'll find that there IS no difference. In that case, we have some real, valid, empirical evidence to argue that. Claiming "it's all in your head" when you don't have evidence of that is kind of going out on a limb. It's probably all in our heads, but we need to do more testing to validate that hypothesis.

I've been mulling this one over and I might see if it's something that some funding and facilities could be gotten for. Need to hash out the design better first though. It's an interesting topic that is pretty much absent from the psychological journals.

You are completely right. There are many different types of harmonics and they all have slightly different effects on how we percieve the reproduced sound. For example it has long been know that odd harmonics are more pleasing to the ear than even harmonics. THD figures mean little, although it's arguable that we can still hear/feel the effects of in-audiable harmonics on sound in the audiable range, much more porgress would be made if the main types of harmonics were measured individually.