[sdiy] Harmonic bandwidth - nonlinearities

[JH.]

>> However, within the domain of any individual's hearing, according to
>> theory, we should
>> not be able to hear the difference between two signals with the same
>> harmonic
>> components when harmonic phase relationships are constant, but different
>> between two
>> signals.
>
>Scott, have you read the article that Ian posted a link to yet
>(http://www.silcom.com/~aludwig/Phase_audibility.htm)? Apparently it's
>been known for quite awhile that the older theory is not 100% correct, and
>the article makes an argument for it being even less correct than in the
>papers that are cited.

I think it's quite convincing that nonlinearities will cause intermodulation 
between partials, and differenetly so, when th epartials have different 
phase in two otherwise indetical samples. And *everything* in nature is 
nonlinear to some degree - the human ear included.
(Just think of it, we have a mechanical compressor in our middle ear, with 
muscle tension on the little bones forming a "VCA".)
The big question is, how subtle these effects are, but for me there is no 
doubt that there _are_ such effects to _some_ degree.
So the conclusion that - paraphrasing what I read out of this article - that 
phase mostly doesn't matter, but on rare occasion it does, is quite 
convincing, IMO.

JH.