Reply by jhearon July 26, 20102010-07-26
I'm new to and struggling with a c code implementation for an available LP
filter, http://www.musicdsp.org/files/filters004.txt. The filter function,
float iir_filter(float input,FILTER *iir) takes a float input and a
reference to the FILTER structure.  I believe I have the filter code
working, and in my rudimentary implementation I can also generate white
noise, window it, and analyze it using an FFTW3 dft r2c ld plan, and graph
all using gnuplot.  But I don't see any change in the spectrum when I feed
samples to iir_filter and adjust the Q, or Filter cutoff.  I do see the
coefficients change for different SR and num of samples.  I'm wondering if
I'm not using the right FFTW plan or sending the right values to the
filter?
Wondering if someone perhaps knows this bit of code and has done an
implementation?