Design, construction, tuning of spectroscopes
Information and discussion about softwares (telescope remote, autoguiding, acquisition, spectral processing ...)
I am trying to help someone who is having a problem with the automatic wavelength calibration of the ALPY using ISIS. The spectra have small tilt and slant errors (<1.5deg) but when ISIS makes the geometric corrections, the X position of the lines moves 15 pixels. This means the entered X position of the reference line is not correct and the lines are not found when the wavelength calibration is run. Has anyone else seen this problem? I expected ISIS to compensate for the change position of the lines caused by the geometric correction.
J'essaie d'aider avec un problème de l'étalonnage automatique de la longueur d'onde de l'ALPY à l'aide d'ISIS. Les spectres présentent de petites erreurs géométriques(<1,5 ° ), mais lorsque ISIS effectue les corrections géométriques, la position X des lignes se déplace de 15 pixels. La position X entrée pour la ligne de référence n'est pas correcte. Quelqu'un d'autre a-t-il vu ce problème? Je pensais qu'ISIS compenserait le changement de position des lignes par la correction géométrique.
I have made some more checks. This is an error in the way ISIS corrects the smile in the ALPY spectrum. The smile correction should not shift the spectrum at the Y position of the star spectrum. ( Note the slant correction in ISIS used with the LHIRES works correctly. There is no shift at the Y position.)
A work around is to check the geometric corrected image, note the X offset in the spectrum, apply this to the entered X location of the reference line and run the calibration a second time.
Hi Robin, a tilt of 1.5 ° is not a small error of orientation, it is a BIG error ! I strongly recommend to reduce the physical tilt to nearly 0.1°-0.2°, this is a basic precaution for a correct use of the spectrograph.
ISIS is not free from defects and improvements can be made (the order of operations can be discussed), but I have not anticipated such a severe case of geometric distortions. The effort is first at the level of the spectrograph setting.
I agree in part. The user admits he did not align the spectrum horizontally but the slant caused by the rotated grism is not his fault. It came from Shelyak like this. His is not the only one with this problem either unfortunately I have explained to him how to adjust the ALPY grism.
Even with the spectrum rotated so the tilt is 0, the slant in the lines is 1.7 deg and the position of the lines shifts after the smile correction by 5 pixels which is enough to stop the ISIS automatic calibration working.
I have just had this problem myself. The ALPY automatic calibration would not run. When I looked at the geometric corrected lamp image, the position of the reference line had moved 8 pixels. The camera had rotated slightly so the star image had a 0.8 deg tilt. This, combined with the smile correction, was enough to stop the automatic wavelength calibration running. It ran perfectly when I entered the reference line position from the corrected image.
screen shots attached
Robin
Had to use reference line position from the corrected image for the calibration to run correctly
ISIS_original_lamp_annot.png (303.25 KiB) Viewed 3826 times
The geometric corrected lamp image
ISIS_lamp_corrected_annot.png (454.68 KiB) Viewed 3826 times
The original star image
ISIS_star_annot.png (254.67 KiB) Viewed 3826 times
The corrected star image
ISIS_corrected_star.png (195.96 KiB) Viewed 3826 times
I have explained to him how to adjust the ALPY grism.