A more practical question is that I wonder if the sampling component in the switched filter acts as a excitation of the resonating filter, and does so in a multiplicative sense, thereby increasing the resonace peak, and introducing the (audible) noise in the signal.
As a more general question, is a perfectly avaraging filter before a sampling circuit sufficient to eliminate all sampling (aliasing) noise? That would practically be fairly achievable by taking a reseteable integrator preceded by a siple high cut filter, the latter to comentsate for the small transfer and reset time gap in the averaging integrating sample and hold (another way of defining a resettable integrator).
A general thought is how Maxwell Equations (which in itself have an appeal for their composition of very intuitive mathematical concepts, maybe even beating the wave equantion in that sense) can yield stable particles. There are probably loads of answers to these type of questions, but a flow and field equation that holds a structure stable for relatively speaking very long times is not very easy to picture. Does that require quantized variables? Can one say that the processes represented by the ME take place at such high speeds (frequencies) that all interactions with them (that would normally break them up I would say) average out to zero interactions, and therefore only very high frequency radiation can seriously affect them? What bearing does relativity have on all this (probably known by people in th. phys. field, I'm just thinking out load).
I remember that (I think it was Mill-Young type of thinking: to give EM variables an intrinsik complex component) quark theory somehow contained the thought of rotating particle conglomerates (with relatively high mass) is that linked with the above?
A though I'm a bit more familiar with has to do with event based models. A event driven simulator draws heavily from the equal time concept. Events either occur at equal time, and are then directly related, or occur at different time instances, meaning that they have causal relationships. representing a time axix in such a system can be generalized by taking a ratinal numbers of sufficiently large denominator as the basis for thetime axix, yielding distinguishable events. This could even be taken to infine numbers of events by including non-rational numbers. One of my questions is, wether this is a fundamenental concept, since discretization is a factor of consideration even in continueous field or particle representations.
Combining these thought with my recent encounter and
subsequent thoughts abuot the remarkable observations concerning
computer random numbers and their appearent relation with
shared consciousness (see 'Dean Martin'comments on page ***),
I start to suspect strongly that a discrete, observeable or
observed event or state has something to do with above thoughts, and
is more than coincidental.
In other words that putting a continuously stated problem into
a discretely observable relation is a fundamental concept not just
in terms of observability (and related uncertainty, see also the
split photon experiment), but also in long term sense, and
possibly related to mathematical ordering and uniqueness.
More later.