Consider the following MATLAB code: load handel; b2 = fir1(200,[0.48 0.85]); n2 = filtfilt(b2,1,randn(size(y))); yn2 = y+n2; sound(yn2,Fs) The resulting signal yn2 is a corrupted version of the original Handel signal y. Select the causal filter design and implementation best eliminates the noise.

b3 = fir1(400,[2000 3500]/(Fs/2),'stop'); yf2 = filter(b3,1,yn2);

b3 = fir1(200,[1000 3500]/(Fs/2),'stop'); yf2 = filter(b3,1,yn2);

b3 = fir1(200,[2000 3500]/(Fs/2),'stop'); yf2 = filtfilt(b3,1,yn2);

b3 = fir1(300,[1500 4000]/(Fs/2),'stop'); yf2 = filter(b3,1,yn2); sound(yf2,Fs)