Meyer wavelet orthogonal!? Scalar product of two dilates isn't zero.

I'm plotting the Meyer wavelet function at different scales. I'd assume that 0 == P(1,:) * P(2,:)' since each two out of such a system ought to be orthogonal. It's not the case, however, not even close to zero.

Do you know where I'm mistaken?

L = 1024;
[psi,~] = meyer(-8,8, L, 'psi');

np = 3;
P = zeros(np,L);
for j=0:np-1
   P(j+1,:) = 2^(j/2)*[zeros(1,(L-L/2^j)/2) psi(1:2^j:L) zeros(1,(L-L/2^j)/2)];
end
xL = [300,L-300];
yL = [-3 3];
for i=1:np
   subplot(np,1,np-i+1);
   plot(P(i,:));
   xlim(xL);
   ylim(yL);
   line([L/2 L/2], yL, 'Color', [0 0 0]);
   line(xL, [0 0], 'Color', [0 0 0]);
   set(gca,'xtick',L/2); set(gca,'xticklabel','0');
end