Steve Freides <[email protected]> wrote:

   Jeff, fascinating stuff, not what I would have expected, but also not
   a result consistent with the science of sound as I know it.  How can
   the overtones drop in pitch and not in amplitude when you record
   yourself from the next room?  It makes sense to me that our
   psycho-acoustic _perception_ could be that the pitch is lower because
   of the change in amplitude of the overtones - what is the explanation
   for the change in _frequency_ of those overtones?

The explanation is that the tools being used are not being used
appropriately.  I don't have much time to respond right now -- need to
walk the dogs and get to a rehearsal -- but I hope to respond more fully
to this issue and also Cabbage's lengthy explanation, with which I
generally agree that overtones (of a steady-state instrument) will
generally be accurate overtones, except that Cabbage missed (teaser!) a
few side issues.

Briefly, Jeff did not document the length of the fft he used, not the
kind of window.  It really makes a difference.  With the default fft
frame length of 512 samples, the 353 fundamental is represented by about
only four cycles.  Fft windowing could easily account for the skewed
frequency identified for the fundamental.  It would have less effect for
the higher partials.  So, the fundamental may be analyzed incorrectly by
audacity, not the upper partials.
_______________________________________________
post: [email protected]
unsubscribe or set options at 
https://pegasus.memphis.edu/cgi-bin/mailman/options/horn/archive%40jab.org

Reply via email to