Compare 16 vs. 32 Bit

[Morten Lerager]

Hi'
My subs are 16bit fits
The stacked fit file are 32 bit

How do i easily compare ADU's on this to image.
I would also like to compare standard deviation.

Thank's Morten Lerager

Edit: I will try to pixel edit to value 70000. It could be a work around to
get the file in 32bit.

Edit Edit: Ohh, a better work around would be to stack only the One sub.
....If this will do the trick, to convert it to 32Bit

Edit Edit Edit: Ok now. Stacking a single sub wont change it to 32bit.
Pixel edit to a V70000 will change to 32bit without scaling. So AA automatic change
på 32bit. only if needed.

The "error" was my bad. I had a single sub with a background standard deviation reading 37.3
After preprocessing 20 x subs the same standard deviation was 143 ....No so funny.
The if I change Sigma + Sum to Average the reading is 6.5

[fabdev]

Hi, to average (or sigma+average) a sequence of 16 bit images, consider the option "Keep result in 16 bits", since the flat field correction can indeed exceed the 16 bit range if a saturated star (65535) is present.
To compare quickly the dynamic of several images, open the Statistics Window for one image, then set it to "Unlocked" mode, then switch to other images.
To compare only a subregion, select a rectangle or a mask over an image, and click "Real time" in the context menu of the statistics. Then move the region, for example with CTRL + mouse drag.

[Morten Lerager]

Thank you for all your great tips & trick.
And to keep it 16 bit seems to me, the way to go from here.

[einari]

Will continue this old thread because my 'issue' is related.

I stacked 7 2 min exposures and the result is that 4 of them are 32 bit and 3 are 16 bits.
Why is that, even if I have set "Keep result in 16 bit." ?

Something to do with fact that nights are light and ADU readings high.

[fabdev]

Hi, the option indeed work on the "result" image of preprocessing.
The intermediate images will not be modified (if you mean that you're saving them with the option "Save each image").
Fabio.

[einari]

Ok.
Can I make sure that each of the "Save each image" are 32 (or 16) bit images?

[fabdev]

If you really need to obtain a sequence of 16 bit images on disk from a preprocessing, you could then use Apply Macro with the command "Clamp maximum 65535".
However, if this is just for saving disk space, consider the option to compress FITS.
If you don't care about the range scaling, you could also solve adding an "Extra" filter in preprocessing: Multiply Constant 0.9

[einari]

It's not for saving space. Was trying to do photometry and it would not tolerate mix of 16 and 32 bit files.
I know there's a problem if I had to use something like Multply constant in processing but that sounds like a viable solution.
Thanks.

[fabdev]

Another solution could be enabling the option "Open all images in floating point" in the Options. All images will be opened and saved as floating point FITS (if your software supports it).