🚀 go-pugleaf

Subject: 1-D Time Domain Filters
Newsgroups: comp.soft-sys.khoros
Keywords: 1-D signal processing, filters

   I am using 1-D signal processing routines in Khoros and have run into a 
problem (or a lack of understanding) of dfiltbp (the Frequency Domain filter
design routines) and dfilter ( the 1-D time domain filter routines).

   Specifically, I have tried to create a bandpass filter using dfiltbp, 
with parameters:
                (Butterworth)
                Sampling Freq.      100 Hz.
                Lower Reject. Freq.   5
                Lower Cutoff  Freq.  10
                Upper Cutoff  Freq.  20
                Upper Reject. Freq.  25

This creates an ascii file for the z-domain transfer function for this filter.


   I then tried to use dfilter to apply the filter designed by dfiltbp 
described above.  According to the Help pages for dfilter, dfilter filters 
signals according to a transfer function which is input as an ascii file and
can be a z-domain function. (I assume of the kind generated by dfiltbp.)

   I tested this on a signals generated by dgsin and dgunif (sin wave and
uniform noise - the sin wave was generated with 100 Hz. sampling,
freq = 15 Hz so it should have been passed by the filter).  The results are 
absurd numbers (lots of really large numbers and NaN's), so I assume that
I am not using these routines correctly.

   Does anyone out there know where I've gone wrong?  Thanks in advance
for your help.


                                        Allan Moser



  +-----------------------------+--------------------------------------------+
  | Allan R. Moser              |    E.I. duPont de Nemours & Company (Inc.) |
  | moser@eplrx7.es.dupont.com  |    Experimental Station Bldg. E357         |
  | (302) 695-7918 (office)     |    P.O. Box 80357                          |
  | (302) 695-7395 (lab)        |    Wilmington, Delaware 19880-0357         |
  +-----------------------------+--------------------------------------------+

-- 

  +-----------------------------+--------------------------------------------+
  | Allan R. Moser              |    E.I. duPont de Nemours & Company (Inc.) |
  | moser@eplrx7.es.dupont.com  |    Experimental Station Bldg. E357         |