Step 1: Sampling My Voice using MATLAB

[y, fs] = wavread('C:\Users\Habad\Dropbox\SPRING 2012 CLASSES\EEGR491\myvoice.wav');

%w = [min(fs) max(fs)]

%sound(y,fs);

%plot(y);

%stem(y);

t=0:1/fs:(length(y)-1)/fs; %get sampling frequency

figure(1);

plot(t,y); %graph it – try zooming while its up…not much visible until you do

FIG 1: AUDIO SIGNAL PLOT WITHOUT NOISE

%clear

[y,fs] = wavread('C:\Users\Habad\Dropbox\SPRING 2012 CLASSES\EEGR491\myvoice_nonoise.wav');

%w = [min(fs) max(fs)]

%sound(y,fs);

%plot(y);

%stem(y);

t=0:1/fs:(length(y)-1)/fs; %get sampling frequency

figure(2);

plot(t,y); %graph it – try zooming while its up…not much visible until you do

FIG 2: AUDIO SIGNAL PLOT WITH NOISE

COMMENTS

So far, I have been able to input my voice as a .wav audio signal. I then sampled it. For the quantization which is step 2, I started writing a function but I decided to write a C code with similar algorithm that would perform the same operation. The program would take in a string of discrete values and then assigned each one of the discrete values to a level. This was taken as a step to enable me understand exactly how the quantization would take place.

C CODE (Incomplete)

int main()

{

    int quant[8] = {0,1,2,3}; //levels

   

    int y, yq, i;

   

    y = [2,3,4,5,6,8,1,2,4,5,2,4,5,6,0,3,4,-1,4,5,1,2,10,9]; //input signal

   

    for (i = 0; i <0; i++)

    {

        //B = 4, number of levels

        //so break max and min into 4 different levels/ranges

        //how many numbers are between max and min, then divide that number by number of levels i.e. 4

        //min=-1, max = 10, so number of numbers  = 12, so there would be 4 elements per range

       

       

         yq = quant[i]; //quantized

    }

   

    system("pause");

    return 0;

}