I captured Vega with the ALPI600 + ASI533MM the results looked good and Demetra was used to perform data reduction.
Demetra’s automatic calibration was successful and the finished sprectra looked good. I submitted it to the AAVSO and it was rejected.
Here’s the message from the AAVSO…
Inaccurate wavelength calibration Comments:: There appears to be a significant shift of 7-8 Angstroms blueward on the blue end around 3839. This is on the cusp of what is acceptable, but could you please check this.
Cropping everything below 4000Å does NOT appeal to me.
Question 1) Since ALPI’s calibration module lamp doesn’t produce any usable spectral lines below 4158.59 Å does this cause Demetras calibration to fail on the Blue end?
Question 2) Can a different lamp be employed to provide spectral lines below 4000Å?
Question 3) Will ISIS do a better job of calibration on the blue end?
I am willing to used Demetra’s manual calibration if that will suffice, but I don’t know how to go about it.
Hi Steve,
Apologies for not directly answering your questions, however I thought should mention this.
Sony Semicon does not provide QE data about most of their sensors below 400nm (4000A).
In part, because they don’t warranty performance in the deep blue/UV end of the spectrum for sensors like the IMX533.
In my day job at Diffraction Limited, we worked with some folks at the University of Toronto to evaluate several of the Sony Semicon sensors, and found they still have significant response down to 350nm or lower, depending on the particular chip. This is outside the specifications from Sony.
In practice, the camera chamber window will block the UV, depending on what material it is made of. At DL, we use a different window for UV applications to pass stuff below 4000A.
Regarding Q2 - yes, you could use a different lamp. The problem is that you don’t want to be around that lamp as the UV can damage skin and eyes. Some Halogen bulbs do emit in the UV - this is why you’ll see inexpensive IKEA desk lamps have a thin piece of glass over the bulb.
Clear skies! Colin
For reference I have only some experience with a SA100 + grism on an ES127 triplet.
What telescope are you using, f ratio, etc. ?
I would ask Shelyak to comment.
I would guess from my small experience that perhaps there is field-curvature or something like that going on. Even using the grism to center the spectrum on the main optical axis on my system I have different focus between blue and red. One night I will have to experiment taking two spectra with focus optimzed for the blue-end and red-end and combining in RSpec.
Still wondering what scope you are using, focal length, etc.?
4th order fit in ISIS… seems like more than should be needed. As perhaps you know higher order fits and then extrapolating outside the fit interval may have significant variation!
I wonder what a 2nd order (parabolic) fit would look like for your calibration? At least perhaps the extrapolation outside the fit interval would be better behaved perhaps.
