Hi,
In the calculation of the 2D Gaussian kernel there appears to be a factor 0.5 missing. The standard
definition of a Gaussian is with exp(-(x^2+y^2)/(2*sigma^2)), see e.g. http://mathworld.wolfram.com/NormalDistribution.html
So c2raytools uses a non-standard definition of sigma.
Please also remember that the width of the Gaussian is not uniquely defined. The FWHM is 2_sigma_sqrt(2*log2). Perhaps more options could be added to the function so that you can
specify either a sigma or a FWHM?
We are also finding that the variance of a white noise field goes down with s^2 where s=2*FWHM.
Hi,
In the calculation of the 2D Gaussian kernel there appears to be a factor 0.5 missing. The standard
definition of a Gaussian is with exp(-(x^2+y^2)/(2*sigma^2)), see e.g. http://mathworld.wolfram.com/NormalDistribution.html
So c2raytools uses a non-standard definition of sigma.
Please also remember that the width of the Gaussian is not uniquely defined. The FWHM is 2_sigma_sqrt(2*log2). Perhaps more options could be added to the function so that you can
specify either a sigma or a FWHM?
We are also finding that the variance of a white noise field goes down with s^2 where s=2*FWHM.