GEnie@willett.UUCP (ForthNet articles from GEnie) (12/22/89)
Date: 12-19-89 (10:34) Number: 49 (Echo) To: NICK JANOW Refer#: NONE From: CLIFFORD KING Read: NO Subj: DSP CONFERENCE Status: PUBLIC MESSAGE Nick: You've definitely picked a tough subject to start out with. Adaptive filtering techniques involve changing the filter coefficients in a predictable way and also in many cases the range of coefficient changes are also controlled so that the system doesn't become unstable. It would be much better to start out with basic filtering algorithms. A good reference is "Discrete-Time Signal Processing", Alan V. Oppenheim and Ronald W. Schafer, Prentice Hall, 1989. I have it and have been reading through it as a review. It uses current terminology and is up to date on what techniques are useful. Thanks for your feedback and interest in the conference. I will begin posting messages on how to actually generate coefficients for filters before too long. FFT's are very important to FIR filter coefficient generation. Cliff King NET/Mail : RCFB Golden, CO ... Vesta Technology & Denver FIG for Forth! ------------ ----- This message came from GEnie via willett through a semi-automated program. Report problems to: 'uunet!willett!dwp' or 'willett!dwp@gateway.sei.cmu.edu'
ForthNet@willett.UUCP (ForthNet articles from GEnie) (01/06/90)
Date: 12-28-89 (20:30) Number: 59
To: CLIFFORD KING Refer#: 55
From: NICK JANOW Read: NO
Subj: DSP CONFERENCE Status: PUBLIC MESSAGE
>You've definitely picked a tough subject to start out with.
Actually, I haven't chosen adaptive filters as a starting point; I have
studied regular filter theory before (both digital and analog) but what
I'm doing now is trying to get the practical aspects of it. Theory is
fine, but it's necessary to actually program some of these things and
run them with different levels of input noise, different word widths,
etc to get a true feel for it.
The reason I'm looking at adaptive filters now is that I'd just finished
my FFT program and was trying to decide what to do next. Then my local
bookstore had a good Adaptive Signal Processing book on their 30% off
shelf. For books that cost $100+, that 30% off is too tempting to pass
up. I also got "Digital Signal Processing, Engineering Aplications" by
Douglas F. Elliot; it seems a well written book. I really like that 30%
off shelf. :)
Adaptive filters theory seems to differ enough from regular filters that
being an expert in regular filter design isn't required. Besides, I've
already done some regular filter design and it's more interesting to
look at something new. Keeping my interest up is important.
I'll have to look for that new book by Oppenheim; I've used some of his
other books.
NET/Mail : British Columbia Forth Board - Burnaby BC - (604)434-5886
-----
This message came from GEnie via willett through a semi-automated program.
Report problems to: 'uunet!willett!dwp' or 'willett!dwp@gateway.sei.cmu.edu'
ForthNet@willett.UUCP (ForthNet articles from GEnie) (01/06/90)
Date: 12-30-89 (20:21) Number: 60 (Echo) To: NICK JANOW Refer#: NONE From: CLIFFORD KING Read: NO Subj: DSP CONFERENCE Status: PUBLIC MESSAGE >Theory is >fine, but it's necessary to actually program some of these things and >run them with different levels of input noise, different word widths, >etc to get a true feel for it. Nick, You've got that right. It requires actually working with filters to get a feel for what is happening. I am not familiar with the two books you mention. I will have to look them up when I have some free time. Cliff King NET/Mail : RCFB Golden, CO (303) 278-0364 VESTA & Denver FIG for Forth! ----- This message came from GEnie via willett through a semi-automated program. Report problems to: 'uunet!willett!dwp' or 'willett!dwp@gateway.sei.cmu.edu'
ForthNet@willett.UUCP (ForthNet articles from GEnie) (01/06/90)
Date: 12-30-89 (20:32) Number: 62 (Echo) To: NICK JANOW Refer#: NONE From: CLIFFORD KING Read: NO Subj: DSP CONFERENCE Status: PUBLIC MESSAGE Nick, I thought of something after my first reply. I am interested in the tools you use to generate coefficients. Have you written your own or used canned software? I would be interested in talking about it. Thanks Cliff King, RCFB NET/Mail : RCFB Golden, CO (303) 278-0364 VESTA & Denver FIG for Forth! ----- This message came from GEnie via willett through a semi-automated program. Report problems to: 'uunet!willett!dwp' or 'willett!dwp@gateway.sei.cmu.edu'
ForthNet@willett.UUCP (ForthNet articles from GEnie) (01/06/90)
Date: 01-02-90 (01:18) Number: 63
To: CLIFFORD KING Refer#: 62
From: NICK JANOW Read: NO
Subj: DSP CONFERENCE Status: PUBLIC MESSAGE
>I am interested in the tools you use to generate coefficients.
I assume you mean what do I use as a "signal generator". I wrote some
tools to build various waveforms, then added some extras, like FM and AM
modulation. It's simply for making an N-point array, since I hadn't
gotten to the point of running long sequences.
I use MacFORTH's floating point tools (80 bits) to generate the initial
values for a sine table, based on the sample length I'm using. Every
sine-based signal is then calculated using that table; quickly and
easily. The integer trig functions just didn't have the resolution. I
went crazy trying to find a bug in the fft program at one point. The
problem turned out to be lack of resolution in the original sine table
(using integer trig).
For output I have a simple "oscilloscope" or "FFT" window. It's not at
all polished, but it's worked well enough for my present needs.
I haven't used any canned software. That would defeat the whole purpose
of this exercise: education. It also forces me to write more 4th; also
good for educational purposes. Besides, the commercial programs for DSP
on the Mac seem rather expensive.
I'll provide more information if you like. The n-point fft program is
available on the BCFB as FFT.MFP (in the Mac section, File area 9).
It's only 8k, but it is MacFORTH Plus specific. Criticism welcome. :)
NET/Mail : British Columbia Forth Board - Burnaby BC - (604)434-5886
-----
This message came from GEnie via willett through a semi-automated program.
Report problems to: 'uunet!willett!dwp' or 'willett!dwp@gateway.sei.cmu.edu'