Add the phase of a signal to magnitude of another signal

I have two voice recordings A and B in the time domain. I changed into the frequency domain using the fft command.Then I got the magnitude and phase of the transform for both of them. My task is to take the phase of the second recording to the magnitude of the first. However, magnitude_A is not equal to phase_B in length. I append some zeros. The following command gives error saying that the two vectors must agree in dimensions [Although both now have the same length]. >> mixed = Amag .* exp(j*Bph); Any suggestions to solve this?

[David Goodmanson]

Hi AbdAlla, once you have checked that both are vectors and length(Amag) == length(Bph), one possibility is that j is not a scalar, due to something going on earlier in the code.

[AbdAlla Hefny]

Thanks David for reply. But the 'j' is not the issue. I tried two other signals, and it worked. I do not know what is the problem with the first signals.

[AbdAlla Hefny]

K, I found out what was the problem, one signal seems to be a matrix not a vector although the vector lengths are the same!

[David Goodmanson]

yes, in retrospect comparing size(Amag) and size(Bph) would have been a better suggestion.