[sdiy] RE: Digital noise source

[Sowa Roman]

I made a little simulation this morning. Looks like
31/32 tap FIR made of pseudo random shift register
would have flat response up to about 800Hz at 32kHz
sample/clock rate. That is when I wanted roughly
3dB/oct slope. It looks more like pink response
with stage count reaching 100.

If anyone is interested, I wrote simple program
that caltulates FIR values to aproximate mouse
drawn filter response. Unfortunately it's in
Polish...
Anyway it here:
http://membres.lycos.fr/romcio/firfilter.zip
(click 'ignore' at first dialog box)

Roman

__-----Original Message-----
__From: Tim Ressel [mailto:madhun2001 at yahoo.com]
__Sent: Tuesday, May 14, 2002 1:30 AM
__To: René" Schmitz; Sowa Roman; Synth-Diy
__Subject: Digital noise source
__
__
__FYI:
__
__I found this lecture thingie
__
__http://www.ee.ucla.edu/ee236b/filters2.pdf
__
__At the bottom it mentions a 1/f filter. They used 50
__taps and got a pretty good response. Didn't mention
__filter wieghts however. Lots of math that I don't get.
__
__
__--TR
__
____________________________________________________
__Do You Yahoo!?
__LAUNCH - Your Yahoo! Music Experience
__http://launch.yahoo.com
__
__
__---------------------------------------------------------
__This message has been scanned for viruses by MIMESweeper.
__
__Wiadomosc zostala sprawdzona pod katem wirusow przez
__oprogramowanie MIMESweeper.
__