How to interpret strange gain values in BasicCCDParameters?

[magnusl]

Hi!

I am playing with the BasicCCDParameters script, following instructions found here: https://www.lightvortexastronomy.com/measuring-your-camera-sensor-parameters-automatically-with-pixinsight.html

Screenshot of the report is attached.

The data here is from a Canon 1000D. I use 2 identical flats, 2 bias frames, and 2 darks with exposures differing with a factor of 10. ANd I get what I think is strange gain values, and generally, strange differences between the colums. What am I missing here?
- I would not expect the gain value to be more than 2. In fact, it is roughly the same despite ISO value on the images. I normally use ISO 400, here I tried 100 to see if gain would be lower. But not.
- I would expect the values to be quite similar in the different columns - that is what I have seen when others try. Why not here?

The files I used are available here:

https://www.dropbox.com/sh/g955gg2qht6e02h/AACE6cp3sx33i7ZceOIav5xZa?dl=0

Any ideas? I'm quite confused.

Magnus

[sharkmelley]

A couple of things I notice:

   1) You have a mixture of ISO 100 and 400.  That won't help.
   2) Your flats show significant vignetting.  You need to select a region of interest that is as flat as possible.

Mark

[magnusl]

Hi!

Thanks for your reply!

So, I do the same with all frames at ISO400, and I define a small preview in the center, which is the flattest part of the image. And again I get what I consider to be strange values, see attached screenshot. Subs are in the same folder as before.

Magnus

[sharkmelley]

Your flat frames are saturated in the green channel.  That's your main problem.  The green channel is clipping at a value of 3651 which you can clearly see if you look at the histogram in the HistogramTransformation tool.

I didn't notice it earlier because I missed the fact that the camera is only 12bit. 

Mark

[magnusl]

Hi!

Ok, thanks! Using another pair of flats with lower exposure makes a difference - now the values are rather consistent across the channels!!

But still - my gain vaiule is 2.7-2.8. I'm surprised by that  - is that really to be expected from a Canon 1000D?

Magnus

[sharkmelley]

Why the surprise?  The gain seems to be a perfectly reasonable figure to me.

It's also broadly in line with the results at PhotonsToPhotos:
http://www.photonstophotos.net/Charts/Gain.htm#Canon%20EOS%201000D_12

Mark