Page 1 of 1

ISIS Calibration with polynomial

Posted: Tue Jul 03, 2018 9:16 pm
by Andrew Smith
I reduce a spectrum in ISIS 5.9.3 with the "Spectral Calibration" not ticked.

I load a polynomial in Profile -> Dispersion and then Calibrate. The spectra is calibrated but the CUNIT1 and CTYPE1 is not updated in the header and can't be edited in ISIS.

I may well not be doing this right.

Regards Andrew

Re: ISIS Calibration with polynomial

Posted: Wed Jul 04, 2018 9:36 pm
by Robin Leadbeater
Hi Andrew,

I think ISIS only writes to the header when you run the full reduction pipeline so in this case you would need to rerun the calibration with "predefined dispersion equation" ticked in the "general" tab, remembering to set the x coordinates to zero in the "calibration" tab if you want it to use the A0 constant in the dispersion equation you entered in the "profile", "dispersion" tab

Cheers
Robin

Re: ISIS Calibration with polynomial

Posted: Wed Jul 04, 2018 9:54 pm
by Robin Leadbeater
Actually trying this, I see calibrating via the profile, dispersion tab wipes the whole header except for the updated wavelength calibration parameters. Very odd

Robin

Re: ISIS Calibration with polynomial

Posted: Thu Jul 05, 2018 8:58 pm
by Christian Buil
There is a bug in ISIS if we save in FITS format a spectral profile initially in .DAT format, after a spectral calibration. The header CUNIT1 and CTYPE1
fields are not updated. The problem comes from the SAVE function (Profile tab). I think that's the trouble.

I have updated a version that fixes this bug. Downloaded the current ISIS version from my WEB page (version 5.9.3). Look at the bottom of the Settings tab: if you see 5.9.3c, you are using the new version (http://www.astrosurf.com/buil/isis-software.html)

The correct procedure is: from a .dat file (or from an uncalibrated internally generated spectrum), do the calibration, and from the Profile tab save in the FITS format.

Note: Version 5.9.3c also slightly improves the spectral calibration quality of the eShel spectrograph data.

Christian B

Re: ISIS Calibration with polynomial

Posted: Fri Jul 06, 2018 7:57 am
by Andrew Smith
Thanks Christian, it now works if I select the "Predefined dispersion equation" option.
Regards Andrew