Converting Pre-Processing Procedures from CCDStack2 to PI

Carl Kwadrat

Hello,

I am currently using PixInsight (PI) for the post-processing
of my calibrated iTelescope.net FITS data.  I
continue to use CCDStack to do the pre-processing to produce master luminance,
red, green, blue, and Ha files to import into PI for post-processing.  I choose to use the calibrated form of the data
that is automatically produced by iTelescope immediately after the images are
captured using their bias, dark, and flat frame archives for that specific
telescope. 

I download the calibrated images to my computer and perform
the following  basic steps with CCDStack:
(1) Examine the Data, (2) Apply Hot/Cold Pixel Filter, (3) Debloom Data (if
necessary), (4) Register Images, (5) Normalize the Images, (6) Apply
Statistical Data Rejection and (7) Combine Images using the Mean.  These are exactly the same steps provided by
Adam in his CCDStack Tutorial except for the first step of calibrating the data
which is already provided for me by iTelescope’s as the pre-processing first
step.  I believe that iTelescope
calibrates the data using MaximDL or another popular astroimaging software
package using the current calibration frame archives of biases, darks, and
flats for the particular telescope that I am imaging with.

Now I am in the process of trying to convert my CCDStack pre-processing procedures to PI with the iTelescope’s  pre-calibrated
data as the starting point and bypassing CCDStack altogether.  I have studied the BPP Script section of PI Fundamentals
to orient myself in PI’s approach to the pre-processing of the image data.  BPP always starts with calibrating the data,
a step that I do not need to perform. 

So far, I have concluded that I need to use the following PI
processes of StarAlignment and ImageIntegration for CCDStacks Registration and
Combine Images processes, respectively,.as part of my sequence.  Perhaps CCDStack’s Image Normalization can be
achieved in PI’s BackgroundNormalization in post-processing.  This leaves me with me with Hot Pixel
Rejection, Deblooming, and Data Rejection in the CCDStack processing paradigm
to identify in the PI realm.  I have yet
to identify the PI processes that will perform these three tasks.  This approach looks like a lot of manual work
to perform if this is indeed the direction needed to pre-process my calibrated
data in PI.

Perhaps I am overlooking a simpler way to make this
pre-processing transition?  Maybe I can
use the Light frame portion of BPP without inducing the calibration process?  I am at an impasse in determining what to do
next and the best way to go about it in PI for my situation.

Any strategies that anyone can provide me with to achieve my
goal of performing the pre-processing steps that I used in CCDStack in PI with
my pre-calibrated data would be greatly appreciated.

Thanks,

Carl

BlockHead

Carl,

You can do everything that you need to in PI and relatively simply.

1. Use Cosmetic Correction as the hot pixel filter. (Automatic is fine). You can make an image container for your data and apply the CC process to it. (I explain both the usage of CC and image containers early on in my videos.)

2. Use StarAlignment to register your images.

3. Then use ImageIntegration to combine your data. 

That is pretty much it. The normalization step that is in CCDStack is part of the ImageIntegration process and there is nothing to select or anything like that. Rejection is in ImageIntegration as well.  I would highly recommend watching my Image Integration primer which gives enough background to lead into the Image Integration explanation.. another set of videos. PixInsight's ImageIntegration is much more sophisticated than CCDStack's normalize->reject method (which is very good). There are much options in PI...and some very good improvements as well. 

With regards to deblooming... that is another story. No software program really has a good solution for this. There are some ways of handling this with PI using Defect Map that would be equivalent to the CCDStack method. 

-the Blockhead