[music-dsp] a-weighting filter coefficients

Al Clark aclark at danvillesignal.com
Mon Apr 15 09:16:26 EDT 2002


At 12:01 PM 4/15/2002 +0100, Adam Green wrote:
>hello there,
>
>would anyone know if there's a bona fide resource for calculating biquad
>coefficients in order to implement a standard a-weighted filter response
>curve? or how is a-weighting officially defined in terms of frequency
>response?
>
>at present i'm guessing the response visually from graphs i've found on der
>web, arriving at a crude approximation of combining a high pass with -3db
>cutoff @ 500 hz and a low-pass with -3db cutoff @ 9000 hz.  i'm currently
>using some assembler routines to generate coefficients for 12db/octave
>slope, which is too steep (graphs would indicate 6db/octave to be
>preferable), but this is another matter and probably a job for something
>like matlabs' filter design toolbox i guess.
>any tips greatly appreciated...
>
>aG
>

The A weight curve is defined in the standards as a set of real poles and 
zeros defined in the s plane.
You can get the definition by searching Google. Martin Saxon has the 
equations at http://www.ptpart.co.uk/

The filter is difficult to implement as a digital filter since filters 
defined in the s plane do not translate directly into the z plane.

Al Clark
Danville Signal Processing, Inc.


>dupswapdrop -- the music-dsp mailing list and website: subscription info,
>FAQ, source code archive, list archive, book reviews, dsp links
>http://shoko.calarts.edu/musicdsp/



dupswapdrop -- the music-dsp mailing list and website: subscription info,
FAQ, source code archive, list archive, book reviews, dsp links
http://shoko.calarts.edu/musicdsp/




More information about the music-dsp mailing list