Low-pass filter at 200Hz with a 2khz sampling rate?

Hi there,

I have ECoG data and need to only look at frequencies from 1-200Hz. I would like to use a band-pass filter to achieve this but I am having a hard time understanding how to calculate the coefficients and the number of taps. I would like to use a least-squares linear phase FIR filter design. I sampled a 2khz.

[Greydon Gilmore]

This text file has multiple channels, I would only need one channel to be filtered, in this instance I would like channel #7. In the textfile it is the 10th column.

Answer 1

[Star Strider]

Your design seems unnecessarily complicated to me.

I would use something like this:

Fs = 44100;                                                     % Sampling frequency
fcuts = [10 20 20E+3 21E+3];                                    % Frequency Vector (Hz)
mags =   [0 1 0];                                               % Magnitude (Defines Passbands & Stopbands)
devs = [0.05  0.01  0.05];                                      % Allowable Deviations
[n,Wn,beta,ftype] = kaiserord(fcuts,mags,devs,Fs);
n = n + rem(n,2);
hh = fir1(n,Wn,ftype,kaiser(n+1,beta),'scale');

figure(1)
freqz(hh, 1, 2^14, Fs)

Use the correct sampling frequency (as ‘Fs’) and bandstop and bandpass frequencies (in ‘fcuts’) for your signal.

See the documentation for kaiserord for details. The kaiserord function provides the normalised frequencies for the fir1 function.

[Greydon Gilmore]

When I input this I get an n = 8930, that seems fairly high?

[John BG]

To Greydon;

1.

try Low pass filter, not BPF

2.

could you supply a sample of the sampled signal in a file attached to your question?

To Star Strider;

may be you would like to consider changing Fs to 2kHz, oversampling without having access to the original signal may distort rather than increase accuracy.

[Star Strider]

@Greydon Gilmore —

I didn’t see ‘Example File.txt’ earlier.

See if this does what you want:

fidi = fopen('Example File.txt','rt');
D = textscan(fidi, ['%*s%s' repmat('%f',1,11) '%*s'], 'CollectOutput',1);
t = datenum(D{1}, 'HH:MM:SS.FFF');
t = (t-t(1))*24*60*60;                                          % Time Vector (sec)
Ts = mean(diff(t));                                             % Sampling Time (sec)
Fs = 1/Ts;                                                      % Sampling Frequency (Hz)
Fn = Fs/2;                                                      % Nyquist Frequency (Hz)

DesiredChannel = 8;                                             % This Should Be Channel #7
s = D{2}(:,DesiredChannel);

fcuts = [0.2  1.5  195  205];                                   % Frequency Vector (Hz)
mags =   [0 1 0];                                               % Magnitude (Defines Passbands & Stopbands)
devs = [0.05  0.01  0.05];                                      % Allowable Deviations
[n,Wn,beta,ftype] = kaiserord(fcuts,mags,devs,Fs);
n = n + rem(n,2);
hh = fir1(n,Wn,ftype,kaiser(n+1,beta),'scale');

figure(1)
freqz(hh, 1, 2^14, Fs)

sfilt = filtfilt(hh, 1, s);                                     % Filter Signal

figure(2)
subplot(2,1,1)
plot(t, s)
grid
title('Raw Signal')

subplot(2,1,2)
plot(t, sfilt)
grid
title('Filtered Signal')

The filter (passband depicted in figure(1)) appears to do what you want. I don’t see much difference in the filtered signal, other than the elimination of the d-c component, but the filter appears to work correctly.