Problems with snr function

I want to find the snr for signal xn and noise ns but I keep getting an error.

clear;
n = [0:1023];
omega = 0.25*pi;
xn = sin(omega*n);
count = 1024;
ns = sqrt(0.2)*randn(1,count);
r = snr(xn,ns);
plot(r);
??? Undefined function or method 'snr' for input arguments of type 'double'.

Answer 1

[Image Analyst]

How about:

theRatio = xn ./ ns;
theSNR = mean(theRatio);

If you want it in dB, see http://en.wikipedia.org/wiki/Signal-to-noise_ratio

[Dick Rusell]

I'm not sure which formula you used to find snr in the first place.

[Image Analyst]

xn is your signal. ns is your noise. The SNR is theRatio xn/ns, but this gives the SNR element by element. So you have a bunch of SNR's - one for each element. So to get it down to just one SNR I took the mean of all the individual SNRs. If you want something different, then say what you want.

Answer 2

[Guillaume]

Sounds like you don't have the signal processing toolbox.

ver

will tell you which toolboxes you have installed.

[Image Analyst]

Or it's an antique version. I think snr() has not always been part of the Signal Processing Toolbox, but it's simple enough to calculate manually.

[Dick Rusell]

I'm using matlab 2011, how would I calculate it manually?

Answer 3

[Youssef Khmou]

Generally the formula is SNR=20log10(std(signal)/std(noise)) , in your case you have :

        snr=20*log10(std(xn)/std(ns)) % 3.8dB