[linux-audio-dev] IR FFT smoothing

Denis Sbragion d.sbragion at infotecna.it
Thu Jul 6 12:56:59 UTC 2006

Hello Andrew,

On Thu, July 6, 2006 14:03, Andrew Gaydenko wrote:
> x[n] is real (abs() of the FFT result). It is interesting, I still have not
> found any
> article about fractional octave smoothing. All googling results lead to cited
> AES article.
> So, applied smoothing is as an "invention" for me :-)

I suspect it is an invention for almost everyone... :) There's little
documentation about the common smoothing procedures used in audio. Most of the
available papers have been written after the Mourjopoulos ones. A better
analysis of the effects of smoothing has been started because an accurate
smoothing procedure of the complex response is needed for digital room
correction purposes. Before this it was pretty much just trickery used to make
the results easier to read to our eyes.

> Is your paper about smoothing free?

"The use of continuous phase for interpolation, smoothing and forming mean
values of complex frequency response curves"

It is available here:



	Denis Sbragion
	Tel: +39 0362 805396, Fax: +39 0362 805404
	URL: http://www.infotecna.it

More information about the Linux-audio-dev mailing list