I did a bit of work with the Hilbert transform (instantaneous frequency IF)=
a decade ago. The IF worked well when tracking a single frequency generator=
, but could become confused where signals had mixed or multiple generators.=
However the newer techniques - which you have alerted me to - look interes=
ting and may overcome the limitations BUT the reference I have found uses i=
ndependent components analysis (ICA) so you would need an array of mics lar=
ger than the number of independent signal generators
Hopefully I am not misleading here as my signal processing days are a few y=
ears ago. :)
--- In Bernard Swart <> wrot=
e:
>
> Hi Lew
>
> Why would two syrinxes make the analysis more difficult?Thanks
>
> Bernard
>
> On Tue, May 5, 2009 at 7:56 PM, Lew Proudfoot
> <>wrote:
>
> >
> >
> > Any feedback on this? Sounds interesting, although twosyrinxes might m=
ake
> > the analysis more difficult!
> > Lew Proudfoot
> > The Wind in My Face
> > Vancleave, MS
> >
> > ________________________________
> > From: Bernard Swart < <bsbirding%40gmail.com>>
> > To: <Naturerecordists%40yahoogroups.co=
m>
> > Sent: Friday, April 17, 2009 8:58:55 AM
> > Subject: [Nature Recordists] Re: Applying Hilbert spectrum theory
> >
> > Hi All
> >
> > Anybody applying Hilbert spectrum & EMD theory to analyse bird calls?
> >
> > I'm especially interested in obtaining a working algorith in C++ or Pyt=
hon.
> >
> > Many thanks
> >
> > REgards
> >
> > Bernard Swart
> >
> >
> >
> >
> >
>
>
>
|