How can i generate coefficients (Num, Deno) for allpass filter?

Question

Yvan cadoux 2022-1-19

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1631995-how-can-i-generate-coefficients-num-deno-for-allpass-filter

评论： Mathieu NOE 2022-2-2

I want to add to different signals wich come from filters ( 1 High pass filter , and the other from low pass filter). Both signals have phase shift. Before adding both signals, i need to correct the phase to get same phase on each signal. For this i decided to use allpass filters.

On simulink there is an object : Allpass . To make it runs for your specific application, you need to set coefficients : Which matlab function do i have to use to generate : Num_coef and Deno_coef for the allpass filter from my datas :backward phase, forward phase, cutoff frequency, filter order. An example of code will be very usefull for me. Thanks a lot.

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Mathieu NOE 2022-1-20

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1631995-how-can-i-generate-coefficients-num-deno-for-allpass-filter#answer_878110

在 MATLAB Online 中打开

hello

try this :

function [b,a]=allpass(n,Fst,Fed,mag,Fs)
%Creates an allpass filter
%
%Syntax:
%		>>[b,a]=allpass(n,Fst,Fed,mag,Fs) or
%		>>allpass(n,Fst,Fed,mag,Fs)
%
%where
%		n = number of pole/zero pairs
%		Fst = Start frequency of pole/zero pairs in Hz
%		Fed = End frequency of pole/zero pairs in Hz
%		mag = Magnitude of pole (should be less than 1)
%		Fs = Sample frequency in Hz
%		b = coefficients of the allpass filter numerator
%		a = coefficients of the allpass filter denominator
%
%****  Inputs  ***********************************************************
if nargin ~= 5
   fprintf('\nNot enough input arguments for ALLPASS\n');
   return
else
end
if Fed<Fst
     fprintf('\nEnd frequency must be greater than start frequency for ALLPASS\n');
     return
else
end
f(1)=Fst;
if n~=1
     fdiv=(Fed-Fst)/(n-1);
     for k=2:n
          f(k)=(k-1)*fdiv+Fst;
     end
else
end
r=mag;									%magnitude
%****  Computation of poles and zeros  ************************************
for k=1:length(f)
     p(k)=r.*exp(i*2*pi*(f(k)/Fs));
end
a=1;										%initializing den
b=1;										%initializing num
for k=1:length(f)
     if imag(p(k))< 1*10^(-10)				%if p(k) is purely real
          p(k)=real(p(k));					%create only one pole/zero
          a=conv(a,[1 -p(k)]);
          b=conv(b,[-p(k) 1]);
     else
     a=conv(conv([1 -p(k)],[1 -conj(p(k))]),a);
     b=conv(conv([-p(k) 1],[-conj(p(k)) 1]),b);
	end
end
%****  Display Z Plane and Freq. Response  ********************************
if nargout ~=2			
     zplane(b,a);
     title('Z Plane of Allpass Filter');
     figure
     freqz(b,a,256,Fs);
     title('Frequency Response of Allpass Filter');
else
end