Image Processing (Deep Sky)

In the following I would like to introduce my preferential approach from taking a picture of a deep sky object to a ready processed "pretty picture". This concerns a sequence of steps which are appropriate for the most images. Particulary the kind and order of the used processing methods depends, nevertheless, strongly of the quality of the raw picture.

Photo

My equipment preferential nowadays looks as follows:
I shoot photos with a modified Canon Digital Rebel XT (with Baader filter) through a Pentax 75 SDHF on a Vixen GP-DX mount. I manually guide with a 16mm crosshair eyepiece on a GSO 680 Newton with the Boxdoerfer MTS-3 SDI drive.
To the compensation of the lightened sky in the outskirts of Krefeld I use an IDAS LP2 filter.
Normally I focus with the software "DSLR Focus".
For each object I take at least ten photos, better more, save them in raw format and try not to use more than two different exposure times (e.g. 240s und 480s). For these exposure times I take several dark frames in raw format.

Rawconversion and calibration

For this I use the software Deep Sky Stacker:

1.) Options

First I adjust the "Raw/Fits DDP Settings...". Normally I choose "1" for the coefficients at "Colors Adjustment". At "White Balance" I deselect both alterbatives and for the "Bayer Matrix Transformation" I choose "Bilinear Interpolation".

2.) Loading images and registration
I load the lights and darks and choose the button "Check all". After that I registrate the images (Register checked pictures...).  In the tab "Advanced" I choose the "Star detection treshold", that approx. 100 stars are recognized.
The result of the registration is presented with descending quality. I examin especially the columns "Number of stars", "Score" and "FWHM". Images which have a quality of less than 80% than the reference image at "Number of stars" or "Score" or have a bad "FWHM", I exclude from further treatment.

3.) Stacking of the images
For stacking I choose the button "Stack checked pictures..." and after that the button "Stacking parameters". For the covering I select "Standard Mode".
In the tab "Light" I select "Average" and "RGB Channels Background Calibration" for "Stacking Mode". With the RGB Channels Background Calibration all 3 channels are levelled to the same level. Therefore there is no need to select a special White Balance.
In the tab "Dark" I select "Average" and "Hot Pixels detection and removal" as "Stacking Mode".
In teh tab "Alignment" I select "Automatically" as method.
In the tab "Cosmetic" I select "Detect and clean remaining Cold Pixels" with the parameters "Filter Size = 1px", "Detection Treshold = 50%" and "Replace pixel value with the median.

4.) Saving of the result
I save the result additionally to the automatically saved image "autosave" as 16Bit TIFF with the parameters "Compression = none" and "Do not apply adjustments to the saved image".

Final Processing

1. Darken the background and bringing out the object
I open the image in Photoshop. At first the image looks very dark. Now I use the curves for bringing out the visual content. In a first step I raise the contrast very highly (Curve left), later more conservative (Curve right).

These steps I repeat several times with correcting the background using the levels after each curve operation:

                             before correcting the levels                                               after correcting the levels

The result looks for instance as follows:

                    M42 before                                                                                                                 M42 afterwards

2. Removing Gradients
For removing gradients I use the Freeware Pixinsight LE. With the tool "Dynamic Background Extraction" you can extract the background from the image.