Page 1 of 1

slant correction in ISIS and influence on calibration

PostPosted: Tue Jun 12, 2018 8:28 am
by Erik Bryssinck
Hello all,

I'm working with a L200 spectrograph with 1800 l/mm grating with a TRIUS SX814 CCD-camera on a C11 telescope and use ISIS 5.9.2 software for datareduction. For an observing session i take first a calibrationspectrum, then some spectra images of the same star and at the end again a calibrationspectrum.

I follow the tutorial of ISIS to calibrate the Ne-spectrum obtained by the internal Ne-calibrationlamp of the L200. so far so good.
I select also the 'slant' on the same calibrationline as selected as referencepoint in the spectrum of the Ne-lamp for ISIS, in my case the 6506.528 Ang. line.
After that i do the complete calibration on the spectrum of the Be-star.
Then i check my calibrated spectrum of the star on the H2O-lines as provided in ISIS.

What I notice now is that I can not let the H2O lines coincide with my spectrum over the entire spectrum. In one way or another, there is a shift that is not linear. (see screenshot of VSpec - check the error on the left side is positiv, on the right end negativ). The first thing that comes to mind is that the calibration is wrong, so next step:
I took care of the complete calibration procedure and also carried out this calibration manually in ISIS.

Upon closer inspection of my calibration spectrum, I notice that the slant is different for each Ne-line (see screenshot in appendix). I can obtain a further slant here when choosing a Ne-line that nevertheless has an influence on the final result.

I thought it was a problem with the L200, but when I download the ISIS turorial file and see the calibration spectrum there, I also notice a similar variation in the slant for each calibration line.
I was thinking that this could be a possible cause of the problem of the variation in the H2O lines of the Be-star calibrated spectrum.

I do not know if there is a correction for this problem in ISIS, in any case I did not find it immediately. can a function be provided in ISIS that corrects the variation in slant over the entire calibration image (for example by measuring the slant for each calibration line or at 3 points)? Or maybe i'm wrong and this slant variation has no influence on the calibration ?

Or possibly I do not apply the calibration procedure correctly or i'm overlooked something in ISIS or another problem?

Has someone the same problem and how can this be resolved or deal with this ?

thanking in advance,
Erik
L200_slant.JPG
slant difference in Ne-calibration spektrum


L200_Vspec.JPG
VSPEC - H2Olines in calibrated spectrum

Re: slant correction in ISIS and influence on calibration

PostPosted: Wed Jun 13, 2018 2:11 pm
by Michel Verlinden
Hello Erik,
Did you select the option "in wavelength recenter" ? Individual 2D spectra are not aligned. Then you have to translate you final spectrum with H20 telluric lines
recentrage.png
recentrage.png (9.53 KiB) Viewed 684 times

Cheers,
Michel

Re: slant correction in ISIS and influence on calibration

PostPosted: Thu Jun 14, 2018 3:38 pm
by Robin Leadbeater
Hi Erik,

The change in slant along the spectrum does mean the dispersion is slightly different depending on the Y position where you measure it but this should not affect the calibration as the dispersion is calculated at the Y position of the spectrum. I do not believe it a problem of offset between the individual exposures that Michel has mentioned either as I think that would give a constant error, not a varying error along the spectrum. Which lines are you using to calculate the dispersion? Extrapolating the dispersion equation outside the limits of these lines cannot be relied on. Could this be the source of the problem ?

Cheers
Robin

Re: slant correction in ISIS and influence on calibration

PostPosted: Thu Jun 14, 2018 5:36 pm
by Erik Bryssinck
Hello Miciel & Robin,

Thanks for your reply !

The ‘wavelength registration’ is checked.
The ne-lines i used for calibration are: 6382.992, 6402.246, 6506.528, 6532.882, 6598.953 and i used the method ‘predefined dispersion equation’ – it is more work, i know but i see what is going on and if all is going well.

But now i moved the grating a little so, i use different ne-lines and the H-alph is now more in center of the CCD-chip and also in the complete spectrum: the ne-lines i use now are: 6506.528, 6532.882, 6598.953, 6678.276, 6717.043 and as method ‘file mode’ with a txt-file i made like mentioned in the tutorial.
I now notice on the spectra, i took last night, that the varying error along the spectrum is much less present, but has not yet completely disappeared. But when using 'crop' I do get much better results. The choice to twist the grating a little, so that I work in another area, and thus also use other ne-lines of the calibration lamp, looks a better choice. What the root cause is now, I'm not sure yet.

best regards
Erik