Hi Stanislaw. There are several ways you can remove 50 Hz (I guess it's line noise) from a signal. The easiest is to create a notch filter, which basically involves taking the FFT of the signal, zeroing out the Fourier coefficients at/around 50 Hz (best to use gentle slopes to minimize introducing artifacts resulting from sharp edges), then taking the inverse FFT. More sophisticated techniques involve source-space separation methods to identify a "50 hz subspace" and project that out of the data.
I have lots of teaching resources on my website (mikexcohen.com) about signal processing and time series analysis. Maybe you would find those useful.
Hi Stanislaw. There are several ways you can remove 50 Hz (I guess it's line noise) from a signal. The easiest is to create a notch filter, which basically involves taking the FFT of the signal, zeroing out the Fourier coefficients at/around 50 Hz (best to use gentle slopes to minimize introducing artifacts resulting from sharp edges), then taking the inverse FFT. More sophisticated techniques involve source-space separation methods to identify a "50 hz subspace" and project that out of the data.
I have lots of teaching resources on my website (mikexcohen.com) about signal processing and time series analysis. Maybe you would find those useful.
the answer by Michael is perfect but I have for you a proposition to use well known tools for do the job. If you are not 'signal processing freak' this would be the best solution for you. For example MATLAB with Signal processing toolbox will fulfill your expectation. Try to import your signal and run sptool. Use for example lowpass Butterworth filter with stopband option and Fs=25 Hz, Fpass=25 Hz, Fstop=32 Hz. Use zero phase algorithm for applying the filtration. Results? Look at the picture.