to generate a noise signal

Read any audio file given and play it from Matlab. Find the

number of samples N in the audio file. Generate a noise signal of length N in which

each sample is from a Gaussian distribution with mean 0 and variance 1. Add the

noise signal in the audio and listen it from Matlab.

Pass the noisy signal generated abouve through a MA filter with M1 = M2 = 2 and

generate the output. Listen the output signal and comment on your observation.

code giving randn error.

[y,Fs] = audioread('hello.mp3');

N = length(y);% Number of sample in y;

variance = 1;

mean=0;

W = sqrt(variance).*randn(1,size(y,1:N))+mean; %Gaussian white noise W

y =y + W; %Add the noise

M1=2;

M2=2;

r = ones(1,M1+M2+1);

impulse_response = (1/M1+M2+1).*r;

for k=-M1:M2

y(k)=.5*y(k) + .5*y(k-1); % taking successive average

end

w = -pi:pi/100:pi;

l=length(y);

k=length(impulse_response);

H=[y,zeros(l+1)];

X=[impulse_response,zeros(k+1)];

for i=1:l+k-1

Y(i)=0;

for j=1:k

if(i-j+1>0)

Y(i)=Y(i)+H(i-j+1)*X(j);

end

end

end

sound(Y,Fs);

plot(w,Y)

[Jeffrey Clark]

@maulik bhagat, I think your randn(1,size(y,1:N)) should be randn(size(y)) ?