spectro-aras.com

Flat produced by ISIS, normal view, it is quite ordinary I think:
Having enlarged the dynamics (see bars for interval, linear), you can see the strange column pairs:
To make visible all, I did a box median [11 x 1] and deducted the flat itself:
Has anybody seen this effect?

On real spectrum, flat corrects it somewhat at higher ADUs, but at lower (<1000 ADU) it deviates, as seen when ISIS processed spectra caused extra 'Z' shapes there.
My workaround was to average together those columns, then it becomes usable.

Peter,


I will be getting a 414ex soon so I can experiment with it. This seems a very good replacement for the 314L+ but need to be tested of course.

Now, I noticed your spectrum is not very aligned with the CCD chip. In general, it is a good practice to align the spectrum (you can use a dust making an horizontal line as a good mark) with your CCD.


Note also the option in ISIS to sum up column when doing the master flat - this may help you in creating a higher signal/noise flat:

Cordialement,
Olivier

Thank you Olivier for the help, I will adjust to the dust line before next observation and will do a retry.
This time the vertical gain correction did not help, maybe I will give a try it on the next session.
The SNR should be already good, taken 21 flats (8 sec) + 15 bias + 21 flat-darks with exposure time match at -15C (and my diagrams above is showing the stacked flat).
Wish I had waited for your evaluation.

Cheers,
Peter

Hi Peter,

I have not seen any column defects like this in the 3 ATIK cameras I have owned. As a comparison, I can see one very faint column defect in the bias frame of my ATK314L+ but it is only detectable in a big stack of exposures at less than 2 counts above the background. It does not appear in long exposure thermals (dark-bias) and is completely invisible in the flat. You can just see it in the attached screenshot. The profile is of the binned full height of the CCD

The ripples are usual ones seen in low resolution spectroscopic flats, I think. It will be interesting to compare the intensity with other Sony CCD ICX285, ICX694

Robin

Updated firmware (3.42 -> 3.44, got from support telling this is a known issue been fixed) solved the problem completely!
(Right before I checked the lines are still there, so it was definitely the upgrade.)

Uploaded now the proof of the fix (median filter of flat - flat). I tried to attach as .fit to compare with others, but can't upload .fit as attachment.
Anybody interested having 314L+ and Alpy modules, we can exchange an ISIS processed flat by mail to compare by ADU as a very rough estimate (and calc. to electrons by factory gain, for example).

Interesting that I started the entire use with seeking for a firmware update, but couldn't find any. Got it via support only.