Why does ISIS add 1 to the pixel X position ?

[Robin Leadbeater]

For some reason it appears ISIS adds 1 to the X pixel positions of the lines used for wavelength calibration before calculating the fit coefficients. See here for a test example
https://britastro.org/node/8842#comment-2644
This means the coefficients calculated by ISIS are only useable in ISIS and give unexpected results if used elsewhere, for example with VSpec, and vice versa.

Does anyone know why ISIS does this ?

Robin

[David Boyd]

It seems to be even more complicated. ISIS uses (CCD pixel + 1) X positions of the calibration lines when calculating the polynominal fit coefficients listed in the Go and Dispersion windows, but a different set of polynomial coefficients calculated using the original CCD pixel X positions when calculating the dlambda fit residuals listed in the Go window.

David

[Christian Buil]

I know this (old) display problem.

Most importantly, ISIS compute correctly the calibration internally (I hope !).

But, in a future version I will try to correct this disturbing problem (I admit) (it is related to the taking into account of the first pixel coordinate in the images (the coordinate of the first pixel is +1, not 0).

Christian