source: SH_SHM/trunk/doc/recfilters.txt @ 373

Recursive filters from poles and zeros files using MATLAB:

- type the FFT filter file (e.g. GRSN_S+G_WWSSN_SP.FLF)
! file GRSN_S+G_WWSSN_SP.FLF
!      =====================
!
! version 1, 25-Nov-94
!
! simulation filter for
! seismometer  h1 (0.670000), T1 (1.250000)
! galvanometer h2 (0.840000), T2 (0.750000)
! recorded at instrument
! seismometer  h0 (0.707000), T0 (120.900002)
! K. Stammler, 25-Nov-94
1357913578
1
7.018386e+01
2
(-3.674286e-02,-3.675396e-02)
(-3.674286e-02,3.675396e-02)
4
(-3.367788e+00,-3.731514e+00)
(-3.367788e+00,3.731514e+00)
(-7.037168e+00,-4.545562e+00)
(-7.037168e+00,4.545562e+00)

- call MATLAB and define column vectors with complex zeros and poles:
>> Z = [
-0.0367 - 0.0368i
-0.0367 + 0.0368i
]
>> P = [
-3.3678 - 3.7315i
-3.3678 + 3.7315i
-7.0372 - 4.5456i
-7.0372 + 4.5456i
]

- define gain K and sample frequency Fs
K = 7.018386e+01
Fs = 20

- compute transfer function using zp2tf:
[num,den] = zp2tf( Z, P, K )

- compute recursive coefficients using bilinear
[NUMd,DENd] = bilinear( num, den, Fs )

- create FLR-file:
1. line '1357913578'
2. line '3'
3. line sample distance in s
4. line gain factor
5. line number of NUMd elements
from 6. line: list NUMd elements
N. line: number of DENd elements
from N+1. line: list DENd elements