mixing audios according to array value

hi i have the following code which mixes two audios and plots them :

Fs = 44100;

[y,Fs] = audioread("sound and 1000 new.wav");

[x,Fs] = audioread("160 hz.wav");

mix = y+x;

sound(mix,Fs)

Nsamps = length(mix);

t = (1/Fs)*(1:Nsamps)

y_fft = abs(fft(mix));

y_fft = y_fft(1:Nsamps/2);

f = Fs*(0:Nsamps/2-1)/Nsamps;

figure

plot(f, y_fft)

xlim([0 5000])

xlabel('Frequency (Hz)')

ylabel('Amplitude')

the values of y and x are x = 192000 and y = 192000

now i create a simple sine tone by the following code:

Fs = 44100;

dt = 1/Fs;

StopTime = 3

t = (0:dt:StopTime-dt);

Fc = 1000;

A = 1;

z = A*sin(2*pi*Fc*t);

sound(z,Fs)

the value of z is

as we can see Z is different than X and Y although they are all 3 seconds so i cant mix them, how can i generate a simple sine tone with tha value of Z similar to X and Y so im able to mix them?

Answer 1

[Mathieu NOE]

hello

the simplest trick here is to first generate z with the same dimension (length) as x and y

use the initial t vector already defined for x / y in the first section of your code

then simply force the values of z for t > StopTime to zero

Fs = 44100; % nb : already defined in first section, no need to duplicate this line 
   dt = 1/Fs;                  
   StopTime = 3             

StopTime = 3

%    t = (0:dt:StopTime-dt);     not used here , use t = (1/Fs)*(1:Nsamps)
%    already defined in first section

Nsamps = 192000

t = 1×192000

1.0e+00 * 0.0000 0.0000 0.0001 0.0001 0.0001 0.0001 0.0002 0.0002 0.0002 0.0002 0.0002 0.0003 0.0003 0.0003 0.0003 0.0004 0.0004 0.0004 0.0004 0.0005 0.0005 0.0005 0.0005 0.0005 0.0006 0.0006 0.0006 0.0006 0.0007 0.0007

   Fc = 1000;
   A = 1;
   z = A*sin(2*pi*Fc*t);
   z(t>StopTime) = 0;
 

ans = 1×2

1 192000

   
   sound(z,Fs)

[Mohamed Turkmani]

they are still incompatible , i wrote

[y,Fs] = audioread("sound and 1000 new.wav");

Fs = 44100;

dt = 1/Fs;

StopTime = 3

Nsamps = length(y);

t = (1/Fs)*(1:Nsamps)

Fc = 1000;

A = 1;

z = A*sin(2*pi*Fc*t);

z(t>StopTime) = 0;

mix = y+z;

sound(mix,Fs)

[Mathieu NOE]

hello

can you check size of y and z ?

I suspect they have same number of samples but different orientaion (one is a row vector and the otheroneis a column vector maybe) . In that case, force both of them to be column vectors :

mix = y(:)+z(:);

[Mohamed Turkmani]

ok i did it i converted the array by

[y,Fs] = audioread("sound and 1000 new.wav");

Fs = 44100;

dt = 1/Fs;

StopTime = 4;

Nsamps = length(y);

t = (1/Fs)*(1:Nsamps);

Fc = 1000;

A = 0.1;

x = A*sin(2*pi*Fc*t);

x = x';

mix = x+y;

sound (mix,Fs)

thanks for your help Matheiu i appreciate it

[Mathieu NOE]

ok - got it , simply changed a few things to make sure it works for any number of channels

I added so normalisation of the output otherwise you may clip (wav export assumes your total signal does not exceed +/- 1 range !)

%% my test files
% [y,Fs] = audioread("test_voice_mono.wav"); % mono
[y,Fs] = audioread("test_voice.wav"); % stereo

% Fs = 44100; % why ? - do not overwritte Fs obtained from line above

dt = 1/Fs;                  
StopTime = 3;             
Nsamps = length(y);
  t = (1/Fs)*(1:Nsamps)';    % !! NB : transpose t to make it col oriented like y
   Fc = 1000;
   A = 1;
   z = A*sin(2*pi*Fc*t); % is now also col oriented like y
   z(t>StopTime) = 0;
   mix = y+z;
   
   % important : normalize output amplitude to avoid clipping 
   mix = mix./(max(abs(mix)));
   
   sound(mix,Fs) 

[Mathieu NOE]

hello again

yeap - finally we converged more or less to the same solution !

still don't forget to normalize mix output to stay within +/- 1 range !

all the best