ARJULE is nothing we'd consider an FIR filter in tbe classic sense. It's some kind of application of autocorrelation, giving you the similarity of the signal to itself over time.

Not sure whether this is what you wanted to get.

There is a DC offset in your raw EMG signal, which is equivalent to a signal with zero frequency. I think your designed FIR filter (may be high pass or band pass filter), filtering the low-frequency signals which includes DC offset also.

So, try to avoid high pass or band pass filter if you want to retain the DC offset in the signal to be filtered.

I suggest, use 'smooth' command in MATLAB, i.e.,

Filtered_EMG_signal = smooth(Raw_EMG_signal);

if, it does not filter well then you can multiple smooth commands together, e.g.,

Filtered_EMG_signal = smooth(smooth(smooth(Raw_EMG_signal)));

Another option is using Savitzky-Golay filter in MATLAB, for this, you can use this command- 'sgolayfilt', e.g.

Filtered_EMG_signal = sgolayfilt(Raw_EMG_signal, 1, 21);

here, order is 1 and frame length = 21.

Both the filters are kind of moving average filter.

Thank you for your answers Govind N. Sahu and U. Dreher . I think I start to understand more about the kind of filter I'm using and the noise it has filtered.

FIR filters are one of two primary types of digital filters used in Digital Signal Processing (DSP) applications, the other type being IIR.