Yup this is as strange as it sounds...
My input was an image such as:
data_type FLOAT32
dim1 128
dim2 128
dim3 45
pixdim1 1.7187500000
pixdim2 1.7187500000
pixdim3 4.5000000000
I ran 3dcalc -a input.hdr -expr 'step(a)' -prefix output.hdr -byte
and I got the following header:
data_type UINT8
dim1 128
dim2 128
dim3 45
pixdim1 1.7187500000
pixdim2 1.7187498808 <----
pixdim3 4.5000000000
As silly as it sounds, trying to use this as a mask gave me an error because the two images "have different dimensions"...
Is this something that can be fixed?
Thanks a lot!
Stathis Kanterakis