do digital filters also affect the phse of a signal?

[weissi]

Excuse me for asking so stupid, maybe the answer is writen in my books..... but I don´t want to spend so much time for finding the answer ;-)
the question belongs to the integrated filters in cooleditpro.... (has this anything to do with FIR or IIR - filters.....)
thanks, wssi

[Brian Guralnick]

Your topic doesn't exactly match your thread contents. I'll answer the question in the topic title:

Cool edit pro did their FFT transforms (filters/noise reduction) correctly. They do not create phase errors.

However, their quick eq option may not. In fact, I found it to be slower than their FFT stuff with 8kpoint FFT conversions.

I'm also not sure about their sample rate conversion, I havn't analysed it's performance.

Now, it seems you need to read up on FFT/FIR/IIR, here is a good link:
http://www.dspguru.com/info/faqs/index.htm

--------
Extra:
If you use win2k/winxp, look for a copy of Cooledit96. A bug within it makes the real-time FFT analysis & vu meter work really smooth and good. Check out playback with a 1K plot. In win95/win98, it would run at the same speed as the newer cooledits, this is sad and annoying since I need to go to CE96 when I analyse audio to check for cheap thing like compression artifacts, 15Khz leakage usually recieved by microphones or cheap cabling picking up the yoke EM signal from monitors, or 50/60Hz leakage. The newer cooledits really suck here.

Does anyone know how to get the new cooledit's spectrum analysis to run like cooledit96's?
---------

[Michael Grant]

Yes, they can. Digital filters can be separated into two categories: "linear phase" and "nonlinear phase".

As their name implies, linear phase filters have a phase response that is a straight line on a linear/linear phase vs. frequency graph. These filters have a constant group delay across all frequencies. (The group delay is basically the slope of the phase response.)

However, it is entirely possible to construct a nonlinear phase digital filter, so named because the phase graph is not a straight line. In this case the group delay is not constant across all frequencies.

FIR filters can be both linear phase and nonlinear phase. Linear phase FIR filters have a symmetrical impulse response. IIR filters are, except in certain degenerate cases, nonlinear phase.

[Brian Guralnick]

Originally posted by Michael Grant
Yes, they can. ...

Funny, I was just about to mention the same thing, ty...

[weissi]

ok, thank you for this, now it´s a little bit clearer for me, but at this time I think I´ve drunk to much to think about it deeper... but it´s a definitive answer to that what I´am looking for...[well, if that sound confusing to you -> its the easter weekend.... so forgive me....;-)]
cheers!