[SciPy-user] filtering without phase shift

Ryan Krauss ryanlists@gmail....
Mon Jul 23 22:27:47 CDT 2007


I have a vector of data from an impact test.  The data has some
"noise" in it that is really ringing in the mass used in the test.  I
have spent a fair amount of effort getting rid of the ringing or
trying to design the test so that the ringing is at as high a
frequency as possible.  I now must resort to filtering the data.

I can get fairly decent results with a 10 kHz filter (see attached).
But I can get really nice results with a 1 kHz filter and then
shifting the vector back in time - somewhat arbitrarily.  This feels
slightly unscientific.  I googled for zero phase filters, but don't
really understand what I was reading.  Is there a way to get magtitude
attenuation without phase lag that is easy to implement and accepted
as good practice among signal processing experts?

Thanks,

Ryan
-------------- next part --------------
A non-text attachment was scrubbed...
Name: filtering.png
Type: image/png
Size: 64927 bytes
Desc: not available
Url : http://projects.scipy.org/pipermail/scipy-user/attachments/20070723/bd5e0e20/attachment-0001.png 


More information about the SciPy-user mailing list