Quantcast

[SciPy-User] inverse filter design, remez?

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|  
Report Content as Inappropriate

[SciPy-User] inverse filter design, remez?

R Schumacher
We have sets of ADC data collected with different front-end filters at acquisition; given an empirically determined roll off in dB/octave or dB/Hz, I'd like to perform equalization (in this case the low frequencies) and I know from sweeps that roll off begins under ~200Hz and -3dB is ~15Hz.

One general example is Bank's algorithm:
http://nullege.com/codes/show/src@p@o@porc-HEAD@...
where people correct sound for their room acoustics.
There is also the Nelson-Kirkeby inverse filter methodology.

Our specific case is flattening the response due to a single-pole RC high-pass filter so I think remez() might be just fine; can anyone here supply links or examples in scipy to this application of remez()?
I'm guessing
scipy.signal.remez(
2,
range(.625, 512, .625),
desired= <inverse of measured gain, range(.625, 256, .625)>,
weight=<unsure of this...>,
Hz=1024,
type=<unsure, what to use for high-pass?>,
maxiter=25,
grid_density=16
)

- Ray
_______________________________________________
SciPy-User mailing list
[hidden email]
https://mail.scipy.org/mailman/listinfo/scipy-user
Loading...