Thread Subject: gaussian fit : center and scale x data normalization

Hello,
I wonder how it is possible to use the 'center and scale x data' function during a gaussian fit with the cftoolbox.
Without checking the 'center and scale x data' box, I obtain 'correct' a1, b1, c1 coef results, with a fit roughly matching the data set. I write 'correct' to express the fact that these coefs are in the right order of magnitude, but the fitting curve could obviously be closer to the data.
To improve things, I checkthe 'center and scale x data' box : then the curve matches far better the data. But how can one go back to the mean and std values? How to deal with the values given by Matlab ?

Thanks for your help,
Cedric