argument should be a vector

figure(1)
subplot (5,1,1)
suspect1=audioread('Bashar.wav'); %importing audio
audiolength1=length(suspect1) %length of audio
[x1,fs1]=audioread('Bashar.wav'); %vector, fs
xFFT1=fft(x1);
f1=-fs1/2:fs1/length(x1):fs1/2-fs1/length(x1);
plot (f1, fftshift(abs(xFFT1)))
noiseAmplitude=0.09;
Noisy_thiefAudio= x2 + noiseAmplitude*rand(audiolength2,1)-noiseAmplitude/2;
noisylength=length(Noisy_thiefAudio)
subplot(5,1,5);
plot(x2,'r-');

WHENEVER I TRY TO CORRELATE, THE ERROR IS "Error using matlab.internal.math.parseXcorrOptions (line 71)

%%%Second argument must be a vector.

%%%Error in xcorr (line 72)

%%% matlab.internal.math.parseXcorrOptions(varargin{:});

I know my Noisy_thiefAudio should be a vector but im not finding a way to convert it. Help please

[Razan Al Hakim]

This is how im correlating

figure(2)
Newaudiolength1=audiolength2-audiolength1;
padaudio1=cat(1,suspect1(:,1),zeros(Newaudiolength1,1));
[Corr1,lag1] = xcorr(f1,Noisy_thiefAudio,'none');
subplot (5,3,1);
plot(lag1,Corr1)

[Jan]

What dimensions does Noisy_thiefAudio have? Maybe the imported sound file is recoreded in stereo?

[Walter Roberson]

Noisy_thiefAudio= x2 + noiseAmplitude*rand(audiolength2,1)-noiseAmplitude/2;

x2 is not defined. audiolength2 is not defined.

Answer 1

[Walter Roberson]

I suspect that your x2 is the output of an audioread(), and that your audio file has two channels (stereo). Your code is expecting that your .wav files only have one channel (mono)

[Razan Al Hakim]

Walter, this the whole code.

figure(1)

subplot (5,1,1)

suspect1=audioread('Bashar.wav'); %importing audio

audiolength1=length(suspect1) %length of audio

[x1,fs1]=audioread('Bashar.wav'); %vector, fs

xFFT1=fft(x1);

f1=-fs1/2:fs1/length(x1):fs1/2-fs1/length(x1);

plot (f1, fftshift(abs(xFFT1)))

title('Voice of Bashar: Suspect #1 using FFT')

ylabel('Amplitude')

xlabel('Hz')

subplot (5,1,2)

suspect2=audioread('Razanne.wav'); %importing audio

audiolength2=length(suspect2) %length of audio

[x2,fs2]=audioread('Razanne.wav'); %vector, fs

xFFT2=fft(x2);

f2=-fs2/2:fs2/length(x2):fs2/2-fs2/length(x2);

plot(f2, fftshift( abs(xFFT2)))

title('Voice of Razanne: Suspect #2 using FFT')

ylabel('Amplitude')

xlabel('Hz')

subplot (5,1,3)

suspect3=audioread('Aya.wav'); %importing audio

audiolength3=length(suspect2) %length of audio

[x3,fs3]=audioread('Aya.wav'); %vector, fs

xFFT3=fft(x3);

f3=-fs3/2:fs3/length(x3):fs3/2-fs3/length(x3);

plot(f3, fftshift( abs(xFFT3)))

title('Voice of Aya: Suspect #3 using FFT')

ylabel('Amplitude')

xlabel('Hz')

subplot (5,1,4)

suspect4=audioread('Malak.wav'); %importing audio

audiolength4=length(suspect4) %length of audio

[x4,fs4]=audioread('Malak.wav'); %vector, fs

xFFT4=fft(x4);

f4=-fs4/2:fs4/length(x4):fs4/2-fs4/length(x4);

plot(f4, fftshift( abs(xFFT4)))

title('Voice of Malak: Suspect #4 using FFT')

ylabel('Amplitude')

xlabel('Hz')

%Now, I am going to generate the noisy sound signal of the actual thief which is suspect 2 :p

noiseAmplitude=0.09;

Noisy_thiefAudio= x2 + noiseAmplitude*rand(audiolength2,1)-noiseAmplitude/2;

noisylength=length(Noisy_thiefAudio)

subplot(5,1,5);

plot(x2,'r-');

title('The noisy signal of the thief')

ylabel('Amplitude')

xlabel('Hz')

soundsc(Noisy_thiefAudio,fs2);

%Noisy_thiefAudio, audiolength2,and audiolength3 have the same length, so

%zero padding is done using one of them .

figure(2)

Newaudiolength1=audiolength2-audiolength1;

padaudio1=cat(1,suspect1(:,1),zeros(Newaudiolength1,1));

Newaudiolength4=audiolength2-audiolength4;

padaudio4=cat(1,suspect4(:,1),zeros(Newaudiolength4,1));

[Corr1,lag1] = xcorr(f1,x2,'none');

subplot (5,3,1);

plot(lag1,Corr1)

ylabel('Amplitude');

title('Cross-correlation between noisy sound and padaudio1')

[Corr2,lag2] = xcorr(f2,Noisy_thiefAudio,'none');

subplot (5,3,2);

plot(lag2,Corr2)

ylabel('Amplitude');

title('Cross-correlation between noisy sound and padaudio1')

[Corr3,lag3] = xcorr(f3,Noisy_thiefAudio,'none');

subplot (5,3,3);

plot(lag3,Corr3)

ylabel('Amplitude');

title('Cross-correlation between noisy sound and padaudio1')

[Razan Al Hakim]

I messed up in the correlation part. My goal is to correlate the noisy signal with each of the previous 4 signals. Can you help with this?