adding noise to a wav file

if I add noise to the script...

load handel.mat;

hfile= 'handel.wav';

wavwrite(y, Fs, hfile);

nsamples=Fs;

[y, Fs , nbits, readinfo]= wavread(hfile,nsamples);

   sound(y(round(2.85/4*end):end),Fs);

How would I do so, I know I need some kind of vector to do it but I dont know how to approach it

Answer 1

[Walter Roberson]

y = y + randn(size(y));

[AngelsaAtWar]

This worked, but it is coming out too loud How would I decrease the amount of white noise used.. and Thanks!

[Walter Roberson]

Multiply randn() by a fraction, such as 1/10 or 1/100.

[AngelsaAtWar]

It works, Thank you very much for the help. I really apprecite it

Answer 2

[Youssef Khmou]

hi,

You have many ways :

1) method one :

 Mean=0.00;
 Var=0.005;
 z=imnoise(y,'Gaussian',Mean,Var);

You can control the amount of noise to add, with N(mean,variance ), and also you can change the type of noise you want ' Poisson','speckle'.....

2)method two :

 SNR=15;
 z=awgn(y,SNR,'measured');

With awgn stands for Additive white Gaussian noise, u work here in terms of Signal to Noise Ratio with SNR=20 log10( signal_power/ noise_power ).

3) method three :

Like Walter answer :

 STD=sqrt(0.005); %  STANDARD DEVIATION
 z=y+STD*randn(size(y))

 z=y+randn(size(y)) ; % Means adding Zero mean noise and VAR=1 .

[Youssef Khmou]

example, i prefer working with SNR :

load handel.mat
Y=y(round(2/3*end:end));
SNR=11; % dB
Z=awgn(Y,SNR,'measured');
figure, plot(Z), hold, plot(Y,'r'), legend('AWGN snr=11dB',' Clean Signal')

Answer 3

[tayyab awan]

How to remove this noise in order to retrieve the original signal ??

[Walter Roberson]

There have been a number of Questions dealing with noise removal; you should do a search.