3 Jul
2004
3 Jul
'04
8:30 a.m.
Hi,
does anyone know a library (preferably C/C++, or anything that
interfaces to it) which implements the Prony algorithm (a.k.a. least
squares fitting of a sampled signal to a sum of damped sinusoids)?
TIA
Albert
--
Dr. Albert Gr"af
Dept. of Music-Informatics, University of Mainz, Germany
Email: Dr.Graef(a)t-online.de, ag(a)muwiinfa.geschichte.uni-mainz.de
WWW: http://www.musikwissenschaft.uni-mainz.de/~ag
3 Jul
3 Jul
2:57 p.m.
Es geschah am Samstag, 3. Juli 2004 14:26 als Albert Graef schrieb:
Hi,
does anyone know a library (preferably C/C++, or anything that
interfaces to it) which implements the Prony algorithm (a.k.a. least
squares fitting of a sampled signal to a sum of damped sinusoids)?
I never heard about the "Prony" algorithm, but if it does what you describe
then it's easy to implement, isn't it? Have you tried it? Any problems? Or
are you concerned about performance?
CU
Christian
8:18 p.m.
On Sat, Jul 03, 2004 at 08:47:38PM +0200, Christian Schoenebeck wrote:
Es geschah am Samstag, 3. Juli 2004 14:26 als Albert
Graef schrieb:
It's *not at all* easy. Exponential fitting is one of the more challenging
algorithms in the field of numerical software.
--
FA
Hi,
does anyone know a library (preferably C/C++, or anything that
interfaces to it) which implements the Prony algorithm (a.k.a. least
squares fitting of a sampled signal to a sum of damped sinusoids)?
I never heard about the "Prony" algorithm, but if it does what you describe
then it's easy to implement, isn't it? Have you tried it? Any problems? Or
are you concerned about performance?
4 Jul
4 Jul
5:47 p.m.
Christian Schoenebeck wrote:
I never heard about the "Prony" algorithm,
but if it does what you describe
then it's easy to implement, isn't it? Have you tried it? Any problems? Or
are you concerned about performance?
No, I haven't implemented it yet. That's why I'm looking around for
available solutions before trying to do it myself. As Fons already
pointed out, the algorithm isn't trivial at all if you want to do it
right and in an efficient manner. There are also issues with numerical
instability when the input is noisy.
In case anyone's interested in this stuff, here's a closer description
of the problem with some references:
http://www.statsci.org/other/prony.html
I'd really appreciate further pointers about descriptions and efficient
implementations of that method. I think I read somewhere that netlib has
a corresponding module, but I wasn't able to locate it.
Alternatively, is there any other spectrum estimation method which gives
a reasonably good approximation of the main partials of a
(pseudo-)periodic signal (FFT is too bad in the lower frequencies)? Does
anyone have any experience (or even code) with using wavelets for that
purpose?
Cheers,
Albert
--
Dr. Albert Gr"af
Dept. of Music-Informatics, University of Mainz, Germany
Email: Dr.Graef(a)t-online.de, ag(a)muwiinfa.geschichte.uni-mainz.de
WWW: http://www.musikwissenschaft.uni-mainz.de/~ag
6:04 p.m.
On Sun, 04 Jul 2004 23:43:44 +0200
Dr.Graef(a)t-online.de (Albert Graef) wrote:
Alternatively, is there any other spectrum estimation
method which gives
a reasonably good approximation of the main partials of a
(pseudo-)periodic signal (FFT is too bad in the lower frequencies)?
This floated by a month or so back:
http://ab-initio.mit.edu/harminv/
but I haven't had a chance to look at it yet.
Does
anyone have any experience (or even code) with using wavelets for that
purpose?
I have messed with wavelets (dyadic and the discrete approximation of
continuous wavelets) and I always found that they never lived up to
their reputation.
Erik
--
+-----------------------------------------------------------+
Erik de Castro Lopo nospam(a)mega-nerd.com (Yes it's valid)
+-----------------------------------------------------------+
"Hey, I've re-dorkulated." -- Prof. Frink (The Simpsons)
6:44 p.m.
Albert Graef sagte:
Alternatively, is there any other spectrum estimation
method which gives
a reasonably good approximation of the main partials of a
(pseudo-)periodic signal (FFT is too bad in the lower
frequencies)? Does
You might want to have a closer look at linear prediction (LPC). It is
very good at displaying only the main peaks and does not suffer from a
coarse resolution at lower frequencies, because you basically get a
continuous filter back)
You can play around with comparing LPC and FFT results at different
settings by using my software :-) http://sonasound.sf.net/
Have fun*
Niklas
5 Jul
5 Jul
4:06 p.m.
Niklas Werner wrote:
You can play around with comparing LPC and FFT results
at different
settings by using my software :-) http://sonasound.sf.net/
Thanks, I'll try that, too.
--
Dr. Albert Gr"af
Dept. of Music-Informatics, University of Mainz, Germany
Email: Dr.Graef(a)t-online.de, ag(a)muwiinfa.geschichte.uni-mainz.de
WWW: http://www.musikwissenschaft.uni-mainz.de/~ag
4:08 p.m.
Erik de Castro Lopo wrote:
This floated by a month or so back:
http://ab-initio.mit.edu/harminv/
Thanks, that looks like a viable alternative, I'll have a look at that.
I have messed with wavelets (dyadic and the discrete
approximation of
continuous wavelets) and I always found that they never lived up to
their reputation.
What where the specific problems you encountered? I've never tried
wavelets for the purpose of spectrum estimation, so it would be very
kind if you could share your experiences.
Albert
--
Dr. Albert Gr"af
Dept. of Music-Informatics, University of Mainz, Germany
Email: Dr.Graef(a)t-online.de, ag(a)muwiinfa.geschichte.uni-mainz.de
WWW: http://www.musikwissenschaft.uni-mainz.de/~ag
5:20 p.m.
On Mon, 05 Jul 2004 22:01:29 +0200
Dr.Graef(a)t-online.de (Albert Graef) wrote:
What where the specific problems you encountered?
I've never tried
wavelets for the purpose of spectrum estimation, so it would be very
kind if you could share your experiences.
The results I obtained with wavelets were never as good as the results
I was able to obtain with FFT based methods.
Erik
--
+-----------------------------------------------------------+
Erik de Castro Lopo nospam(a)mega-nerd.com (Yes it's valid)
+-----------------------------------------------------------+
"Debugging is twice as hard as writing the code in the first place.
Therefore, if you write the code as cleverly as possible, you are,
by definition, not smart enough to debug it."
--Brian Kernighan
9 Jul
9 Jul
6:32 a.m.
Es geschah am Sonntag, 4. Juli 2004 23:43 als Albert Graef schrieb:
Christian Schoenebeck wrote:
Sorry for my naive reply :) , I thought it was just a simple linear least
squares problem.
CU
Christian
I never heard about the "Prony"
algorithm, but if it does what you
describe then it's easy to implement, isn't it? Have you tried it? Any
problems? Or are you concerned about performance?
No, I haven't implemented it yet. That's why I'm looking around for
available solutions before trying to do it myself. As Fons already
pointed out, the algorithm isn't trivial at all if you want to do it
right and in an efficient manner. There are also issues with numerical
instability when the input is noisy.
7264
days inactive
7270
days old
linux-audio-dev@lists.linuxaudio.org
9 comments
5 participants
participants (5)
-
Christian Schoenebeck -
Dr.Graef＠t-online.de -
Erik de Castro Lopo -
Fons Adriaensen -
Niklas Werner