Thread Subject: steganography- High PSNR

hi all....
please i am getting very high PSNR values like above 50 between by original and stego image, and i know it is supposed to range between 20-40db. I have debugged my code and the embedding seems to be working just fine....i am at loss as to what could be causing it.....
Any ideas as to why i am getting high PSNR values and how i can get a value between the normal range..
thanks

Script to compute PSNR and EQM(Error Quadratic Mean):


PSNR.m:

y=imread('lena.tif');
[taille_i_l taille_i_c]=size(y);
figure(1)
colormap('gray')
imagesc(y);


Q=8;
z=double(dec2bin(double(y),Q))- 48;
Y = z(:);
[l_z c_z]= size(z);

fRe = 2*Y(1:2:length(Y))-1;
fIm = 2*Y(2:2:length(Y))-1; %
dk=(fRe+sqrt(-1)*fIm);
N_symb=length(dk);

%% Image reconstruction
X(1:2:2*N_symb) = real(dk);
X(2:2:2*N_symb) = imag(dk);
X_p=zeros(1,l_z*Q);
X_p(1:length(X))=X;



z2=reshape(X_p,length(X_p)/Q,Q);
mindex_1=find(z2==-1);
z2(mindex_1)=0;
num=2.^(Q-1:-1:0);
z_quantifie=z2*num';
image_dec=reshape(z_quantifie,taille_i_l,taille_i_c);
figure(3)
colormap('gray')
imagesc(image_dec)

%%PSNR
%EQM :
EQM = sum ((double(y(:))-image_dec(:)).^2)/(size(y,1)*size(y,2));

%PSNR: Peak Signal-Noise Ratio
PSNR = 10*log10(255*255/EQM)

My car is getting really bad gas mileage. It is supposed to get 35 mpg but gets 30 mpg. Can you tell me why?

If you answer No, and you think about WHY you cannot tell me why, you may have some insight into your own question as posted.

"Matt Fig" <spamanon@yahoo.com> wrote in message <h4suqh$40$1@fred.mathworks.com>...
> My car is getting really bad gas mileage. It is supposed to get 35 mpg but gets 30 mpg. Can you tell me why?

but - matt - of course(!): your counter has a FP issue...

us...

"us " <us@neurol.unizh.ch> wrote in message <h4svkr$lhs$1@fred.mathworks.com>...
> "Matt Fig" <spamanon@yahoo.com> wrote in message <h4suqh$40$1@fred.mathworks.com>...
> > My car is getting really bad gas mileage. It is supposed to get 35 mpg but gets 30 mpg. Can you tell me why?
>
> but - matt - of course(!): your counter has a FP issue...
>
> us...

Oh, I forgot!

eps(30)
ans =
 5