Post Processing - worked example
Please note that the following is a rather 'down and dirty' method and I explain a much better method in my book - 'Making Every Photon Count'.
There is nothing like a pictorial worked example to explain a set of procedures so, hopefully, this small article will be of assistance to newcomers to the 'art' of post processing. Although I have used PhotoShop throughout this tutorial, the same (or very similar) functionality is available with many other graphics programs like PhotoShop Essentials, The Gimp (free of charge) and PaintShop Pro as obvious examples.
The Andromeda Galaxy is a popular first object to image because it is relatively easy to find (being our closest galactic neighbour, it is quite large and quite bright). This does not make it an easy target, however, as with many similar objects there is a balance to be found between burning out the galactic core yet capturing the fine detail in the dust lanes that give any galaxy it's individual character. It is for these reasons that I have deliberately used this object for a basic worked example and have used images taken with a commonly used DSLR, the Canon EOS 300D and a 200mm lens as many owners buy an 80 to 210mm zoom lens to go with this camera.
The final image consists of 11 images of 120 seconds exposure each plus a dark frame of the same exposure length, all taken at ISO 800. This is a typical maximum exposure time and ISO value for a shot like this for several reasons:-
1. it is within the capabilities of many motorised mounts to track adequately without guiding,
2. it will capture enough of the finer dark detail without 'burning out' the bright core too much,
3. it is not long enough to introduce too much sensor noise and
4. it is hopefully below the 'fog' limit for light pollution! No light pollution filter was used for this image as I wanted to show how this number-one-enemy of the astro-imager can be dealt with relatively easily.
The following image is one of the 11 images used for this project, just as it was downloaded from the camera:-
This image could yeald a reasonable finished image but much more detail can be extracted by 'stacking' several images and carrying out a dark frame subtraction. This was carried out using automatic settings in Registax V3.0 which is readily available free of charge. The following image is the result of stacking the 11 images:-
There is now more detail to work with but that light pollution is spoiling the image so this is the first task to tackle. A useful by-product of this process is that we will immediately have a better view of what we have captured and what other adjustments will need to be carried out. Selecting Image/Adjustments/Levels we will be presented with a dialogue box showing a histogram of the image and some very powerful adjustment tools. However, for now we are going to do the single most important adjustment - that of setting the background level to something approaching the night sky as we would like to see it! There are three 'pipettes' (also known as 'droppers') within this box (outlined in red below) and they are used to select the Black, Grey and White levels within the image but for now, we are only concerned with the first one.
Clicking on the black pipette changes the cursor to a pipette shape and moving this onto the canvas and selecting a part of the sky with NO stars, nebulosity or other objects results is a magical improvement in the image. Note how the histogram has changed to reflect the change that we have made. Sometimes this simple expedient can be a little savage as you don't want a completely black sky as this is not how you will ever see it, however, later 'level' adjustments will bring some brightness back into the background making it appear more 'natural'. Please note that this is a rather 'down and dirty' method and I explain a much better method in my book - 'Making Every Photon Count'.
We can now see more clearly what has been captured as well as achieving a more 'natural' looking image. As we have a very black background now, we can afford to 'stretch' the image a little to reveal more detail and this is done using the middle of the three sliders underneath the histogram (outlined in red below). Sliding this to the right will decrease brightness and sliding to the left will increase brightness and this is what we want to do. The default Channel is the RGB channel and this is the one to use for now. However, selecting the Red, Green or Blue channel will allow you to enhance or diminish just this colour rather than all three colours together.
As you can see, this has made a huge change to the image again and much additional detail has been revealed. The image is now ready for 'cropping' to improve the composition of the image. To maintain the existing aspect ratio of the image we will use the 'Canvas Size' tool. Select Image/Canvas Size then choose 'percent' and 'lock' the section than we want to remain unchanged - in this example, I have locked the top right hand corner of the image so that any reduction will take place at the left and base of the image. I chose 85% to centre the galaxy in the frame.
Having centred the image to my satisfaction, I then performed the operation a second time with no 'locking' to concentrate the image on the galaxy and not the surrounding stars. The following is the final image cropped and sized to 600 x 400 pixels suitable for inclusion on my website:-
To show just how far the image has changed with just these basic adjustments, the last image is a 'before and after' composite:-
Copyright Steve Richards 2008