filtfilt provides excessive transient
15 views (last 30 days)
I observed that 'filtfilt' suffers from an undesired behavior when I provide IIR bandpass filters having steep transition bands. Specifically, the output signal exhibits excessive transient response; nevertheless, this behaviour does not emerge if I use a 'home-made' version of zerophase filtering based on 'filter'. I guess that it is not a numerical issue involving the filter coefficients or structure, nor the 'filter' function.
% design bandpass filter having transition bandwidth of 200 Hz (Fs = 8000)
bp = designfilt('bandpassiir', 'StopbandFrequency1', 50, 'PassbandFrequency1', 250,...
'PassbandFrequency2', 3600, 'StopbandFrequency2', 3700, 'StopbandAttenuation1', 30,...
'PassbandRipple', 0.1, 'StopbandAttenuation2', 30, 'SampleRate', 8000, 'DesignMethod', 'cheby2');
% check stability
% apply filtfilt to a random (white) long input signal; output signal shows an undesirable transient
x = randn(2^20,1);
y = filtfilt(bp,x);
% apply 'home-made' filtfilt to the same input; output signal shows a more accptable transient
y2 = flipud(filter(bp,flipud(filter(bp,x))));
% compare effects
As a rule of thumb, the effect grows as the transition bands get narrower, while it tends to vanish as they get broader.
Where is the problem? Have I missed some recommendations or hints in the function's help?
Anh Tran on 4 Jan 2018
The transients observed are due to a combination of using a marginally stable filter coupled with the initial condition matching performed by "filtfilt" to minimize start-up and ending transients, as described in the documentation. This is the reason behind the difference in behavior between using "filter" and using "filtfilt". It is an inherent limitation of the "filtfilt" function.
Regarding the stability of the filter, you may view the filter's stability using "fvtool", (fvtool(bp) in your case). In the Filter Visualization Tool window, select "Analysis > Pole/Zero Plot", and you will see a discrete system pole/zero map If you notice, your zeroes are very close to the edge of the unit circle, indicating the marginally stable behavior of this filter. When this type of filter is combined with the initial condition matching, the numerical conditioning yields the transient results that you observed.