ISIS-mode fichier (file mode) for wavelength calibration

[Robin Leadbeater]

I am now changing over to ISIS from using VSpec for LHIRES spectrum reduction and I have some questions about this mode of wavelength calibration:

1) I am trying to use "file mode" for wavelength calibration but am seeing some unexpected results. I am doing a linear calibration using 2 lines. I have included in the file an estimate of the dispersion so that program can identify the second line. The data reduction runs successfully and the calibration looks good but the dispersion given in the fits header always uses the same estimated value I entered in the file, not an actual calculated dispersion. Is this correct?

2) I plan to use the same "file mode" for non linear calibration on many lines. Some lines are weak and close together so there is a risk the program will not identify them correctly using just an estimated linear dispersion. I need to test this. How can I check that the lines have been correctly identified and the quality of the fit is good. Is there a list of residual errors generated somewhere? Can I view the calibrated lamp spectrum somewhere?

Thanks
Robin

[Christian Buil]

For the first point, the final dispersion is hard coded by ISIS.

For the second point, for the moment only the RMS error of calibration is returned, but the possibility to check individual line error can be added for a future version.
Also, from the setup tab, unselect option "Erase automaticaly intermediate files". Re-run the processing and display @calib.dat file. It is
the calibrated spectrum of your reference spectral lamp.

Christian

[Robin Leadbeater]

Also, from the setup tab, unselect option "Erase automaticaly intermediate files". Re-run the processing and display @calib.dat file. It is
the calibrated spectrum of your reference spectral lamp.

Hello Christian,

I had already found the file @calib.dat but it is just the lamp profile in pixels, not calibrated in wavelength. Have I missed something?

Thanks
Robin

[Robin Leadbeater]

Also, from the setup tab, unselect option "Erase automaticaly intermediate files". Re-run the processing and display @calib.dat file. It is
the calibrated spectrum of your reference spectral lamp.

Hello Christian,

I had already found the file @calib.dat but it is just the lamp profile in pixels, not calibrated in wavelength. Have I missed something?

Thanks
Robin

Also what is @correl.dat ? and what is the difference between @plot.dat and the final result? ?

Thanks
Robin

[Christian Buil]

Exact Robin, @calib.dat correspond to a raw extracted profile.

Two solutions for check calibration:

1 - semi-manual calibration - after loading @calib file, from Profile tab, click
Calibration button, then, Calibration. (with actual polynomial parameters).

2 - make a full processing of calibration image - see attached screen copy

Pour cette dernière technique, il y a pas mal d'astuce.

Par exemple, je ne traite qu'une seule image, dont je peux indiquer
le nom complet avec l'indes "... neon-1", mais je doit mettre "0" dans le nombre
d'images.

J'aurai pu ici aussi faire "... neon-" et mettre "1" image.

Comme on traite une image spéciale, comme celle d'un objet étendu, je sélectonne
les option "Valeur Y..." et "Fond de ciel retiré".

Enfin,pour le nom de l'objet, je donne le nom que je veux (ic' TEST).

Aller directement dans l'onglet "Go" et lancer le traitement.

Il est très intéressent de bien analyser comment fonctionne c'est exemple,
on y apprend pas mal de chose sur ISIS.

Christian

[Robin Leadbeater]

Thanks Christian,

I will try those suggestions

Cheers
Robin