Help with Alpy Calibration

Design, construction, tuning of spectroscopes
Information and discussion about softwares (telescope remote, autoguiding, acquisition, spectral processing ...)
Paul Luckas
Posts: 101
Joined: Tue Jun 24, 2014 5:08 pm
Location: Perth, Western Australia

Help with Alpy Calibration

Post by Paul Luckas »

I'm experiencing some strange things working through Christian's Alpy calibration methods guide a part of commissioning a new Alpy.

Using the Balmer lines method on a nice A0V standard I get a very good result - an RMS of 0.42, and a set of Balmer lines within an angstrom or two (eye-balling) on my resulting profile. So far so good.

Then I try the pre-defined Alpy calibration module mode. Now I'm getting an RMS of 12, and resulting Balmer lines that are upwards of 15Å from where they should be. My wavelength fit deviations are all over the place, and I'm wondering what's going on here. Of note, my spectral image has quite severe fishtailing at the short end, and the corresponding neon image shows significant 'softening' at the blue end. I'm wondering if this is the cause - as during testing with the Calibration Wizard, I get an error message informing me that my "calibration error is too high" (UV on or off).

I have managed to manually compute a quite acceptable dispersion from my neon using 5 lines from 4200 to 7635 giving me an RMS of 0.052 using a 3rd order polynom, but my resulting spectra are not as 'aligned' with an A0V reference as I'd like - to within a couple of pixels, but not quite right. In fact, my aforementioned 'Balmer lines' version appears more accurate when comparing with a database A0V, hence the alarm bells (that and a reported R of 850 @ 5852Å which seems a bit too good to be true).

I'm using an 80mm fluorite doublet - so I'm not sure whether the fishtail is due to optics or my focus (though I did try pretty hard to narrow down the focus to within the Alpy's seemingly 'quarter turn' tolerance while observing the solar spectrum).

Advice or thoughts on any of the above would be most appreciated. All in all, I'm loving the Aply experience (coming from a LhiresIII). And, as always, much appreciation for the excellent information on Christian's web site.

Cheers,

Paul
Francois Teyssier
Posts: 1520
Joined: Fri Sep 23, 2011 1:01 pm
Location: Rouen
Contact:

Re: Help with Alpy Calibration

Post by Francois Teyssier »

Hi Paul,
Have you check the radial velocity of your reference star ?
Not sure this is the right answer to your issue, but sometimes this is the explanation
Cheers,
François
Benjamin Mauclaire
Posts: 258
Joined: Thu Sep 29, 2011 10:14 am

Re: Help with Alpy Calibration

Post by Benjamin Mauclaire »

Hello Paul,

As calibration lamp spectrum has the fishtail and other geometric defromations as yours stellar spectra, calibration law computed with lamp contains information adapted to the spectra taken before or after the lamp spectrum.

Today I answered Etienne (http://www.spectro-aras.com/forum/viewt ... 5688#p5688) about Alpy calibration and his Vega spectrum's shape (Planck continuum).
I did a quick processing of his Vega spectra (A0V type) with 6 Ar-Ne lines coming from his lamp spectrum well spreaded over the whole spectrum (see link above) giving a third order calibration law (upper order is quite instable and not so accurate at edges).
As you see, superposed Pickels spectrum (green) on Vega's spectrum (blue) shows quite good calibration at blue end.
superposision_vega_a0v-zoom_bleu.png
superposision_vega_a0v-zoom_bleu.png (4.8 KiB) Viewed 11916 times
superposision_vega_a0v-zoom2_bleu.png
superposision_vega_a0v-zoom2_bleu.png (2.61 KiB) Viewed 11916 times
Anyway, using more lines would give a more accurate calibration law.
In calibration step, Spcaudace shows Relco lamp annoted spectrum from O. Garde for guiding.

Clear skies,

Benji
Last edited by Benjamin Mauclaire on Wed Sep 09, 2015 7:20 am, edited 1 time in total.
Spcaudace spectroscopy software: saving you hundred hours of frustration.
Paul Luckas
Posts: 101
Joined: Tue Jun 24, 2014 5:08 pm
Location: Perth, Western Australia

Re: Help with Alpy Calibration

Post by Paul Luckas »

Thanks guys - much appreciate your comments. This may be more of a question for Christian, as I'm using the method outlined here http://www.astrosurf.com/buil/isis/guid ... ration.htm.

It's not a show stopper, but I'm curious as to why I am unable to get a good result using the predefined Alpy calibration mode in ISIS. Why such a large RMS ?
ISIS Predefined mode using calibration module (1 line)
ISIS Predefined mode using calibration module (1 line)
The Balmer line calibration method gives me a very respectable RMS = 0.4.
ISIS Predefined calibration mode using Balmer line
ISIS Predefined calibration mode using Balmer line
Finally, here's a result using predefined dispersion from upwards of half a dozen lines from 4200Å to 7635Å (and IR corrected this time).
Screenshot 2015-09-09 12.41.09.png
Cheers,

Paul.
Robin Leadbeater
Posts: 1926
Joined: Mon Sep 26, 2011 4:41 pm
Contact:

Re: Help with Alpy Calibration

Post by Robin Leadbeater »

Hi Paul,

Are you using the automatic line finding function in ISIS? ie you enter the pixel size, click on the position of the neon 5852 line and ISIS does the rest? If so I suspect your problem is that ISIS has misidentified the lines it uses. I also had this problem. I needed to fine tune the pixel size, rerunning until the fit errors suddenly dropped to a very low value as the lines were identified correctly. (The pixel size is used to roughly calculate where to expect the lines and presumably there is some small variation in dispersion between the instrument that Christian used to set up this feature in ISIS and the one I am using)

Cheers
Robin

EDIT - the Ne reference line is at 5852 not 5400A
Last edited by Robin Leadbeater on Thu Sep 10, 2015 12:47 am, edited 2 times in total.
LHIRES III #29 ATIK314 ALPY 600/200 ATIK428 Star Analyser 100/200 C11 EQ6
http://www.threehillsobservatory.co.uk
Paul Luckas
Posts: 101
Joined: Tue Jun 24, 2014 5:08 pm
Location: Perth, Western Australia

Re: Help with Alpy Calibration

Post by Paul Luckas »

Hi Robin,

Yes, that's exactly what I'm doing. It's not a show stopper, but I still found it curious - particularly as the "Balmer line" method yielded surprisingly good results. I fiddled a bit with pixel size, but it didn't seem to make that much of a difference, and so suspected something else was going on. I'll take another look.

Cheers,

Paul
Vincent Bouttard

Re: Help with Alpy Calibration

Post by Vincent Bouttard »

Hello Paul,

What pixel size do you enter in the General panel ? You have to be very careful with this value because it is the basis for Isis to successfully find the lines in the automatic calibration process. Christian suggests a formula to reach the correct pixel size : p = 3123 / dx where dx is the distance (in pixels) between Ha and Hb lines in a star calibration picture and p is the size of the pixel in µ.

I use the mixed method : Balmer lines calibration for the blue part of the spectrum and Neon calibration for the redder part.
With the calibration module, you could use the neon calibration method only if the picture of the neon is clearly surexposed because the blue part of the neon spectrum is too faint otherwise.

I hope that it could help you.

Vincent
Paul Luckas
Posts: 101
Joined: Tue Jun 24, 2014 5:08 pm
Location: Perth, Western Australia

Re: Help with Alpy Calibration

Post by Paul Luckas »

Thanks Vincent - great advice, thanks for confirming. I will take another look at pixel size.

Paul
SteveCuthbert
Posts: 45
Joined: Fri Aug 09, 2013 6:25 am

Re: Help with Alpy Calibration

Post by SteveCuthbert »

Hi Paul
Probably a daft question and someone will no doubt correct me but are you using an unbinned 1x1 or 2x2 binned image as I gather you have to double the pixel size if using a 2x2 one?.
Steve
Robin Leadbeater
Posts: 1926
Joined: Mon Sep 26, 2011 4:41 pm
Contact:

Re: Help with Alpy Calibration

Post by Robin Leadbeater »

I just checked a few of the ISIS txt files for some of my observations calibrated using the ALPY calibration module Ne and Ar lines only. I use an ATK314L camera with 6.45um pixels. To get a good fit with the correct lines identified I used pixel values of 6.40 unbinned and 12.80 binned which agrees well with the formula Vincent found. My RMS fit is very good, typically 0.2A unbinned and 0.4A binned. The absolute wavelength calibration is also good with <0.5A error at H beta for example. I have not tried the technique of mixing the Ne/Ar lines and blue end Balmer lines though. I am not sure there is much advantage in this though provided the Ar lines at the blue end are used

Cheers
Robin
LHIRES III #29 ATIK314 ALPY 600/200 ATIK428 Star Analyser 100/200 C11 EQ6
http://www.threehillsobservatory.co.uk
Locked