Why does the approximations at different levels have the same size for wavelet swt2 function?

In the documentation for SWT2, it says:

 [A,H,V,D] = swt2(X,N,'wname')
 For 1 ? i ? N, the output matrix A(:,:,i) contains the coefficients of approximation of level i.

When I run the example given for this example, I found the sizes at different, i.e. different i, all the matrices of A(:,:,i) have the same size.

I think it's downsampled for approximation at different levels.

Where's my fault?

Answer 1

[L.]

swt2 doesn't do downsampling, this is the difference between dwt and swt