Mark E. Skeels wrote:
>     * I don't know much about DSP. Could you explain to me what you
>       mean when you say "a true impulse is of course impossible"?

This is straight from linear systems and signal processing theory, which I
have neither time, inclination, nor space to try to explain here.

Look up impulse or "gamma function", and you'll see it's a function defined
to have zero duration but a area of 1.  That's rather hard to do in
practise.  Usually you make due with short enough for the highest frequency
of interest and adjust the amplitude or compensate for it in the math as
appropriate.  Another trick is to use a step function, which is generally a
little easier to realize well enough.  Since a step is the integral of a
impulse, the system response you get is the integral of the impulse
response.  Math can convert that to the impulse response, but since it's
taking the derivative is susceptible to noise.

>       Maybe another way to
>       approach it is to sweep through the relevant part of the audio
>       band and note the amplitude at various levels?

Right, as I said in my original post on this thread.

>     * Let's say one wanted to apply some kind of 'impulse" to the
>       room....how could you go about that? Thump on the wall? Run some
>       kind of signal through a speaker or some other kind of
>       transducer? What would be the character of that "impulse" ?

A thump on the wall will be nowhere near good enough for the purpose, and
will be in the wrong location too.  You can probably put a short pulse into
a sufficiently high frequency transducer to get a good enough impulse for
the purpose.  However, there are lots of things that can cause practise to
deviate from theory with this method.  If you have to ask, it's probably no=
t
the right approach.


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.
--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.