11.36
...
·
9
likes
|
|---|
|
Hi all, I have released a new tutorial about PixInsight. This time about a procedure that I developed for my last project: https://www.astrobin.com/43ckmd/?nc=user You can check the details on the original article here: https://aiastro.wordpress.com/2021/03/10/projects/ This methodology takes profit from the functionalities of projects in PixInsight to use a Process Container produced from the History Explorer. Process Container is a tool that I always knew about, but I never used them, but once I discovered how they can store a certain workflow, they become a powerfull tool for this purposes. The idea is quite simple. Once we have processed an image, we can extract the entire workflow and put it in a Process Container, allowing to apply the whole process to another unprocessed image. The key point of this methodology is that we can modify this workflow in the Process Container to try different processing approaches. The article describes some of the functionalities that we have avaliable when using projects, but I will share here simply an example of application. Let's assume we have processed an image, giving the following result:  If we open its History Explorer we will see the following processes:  We applied Morphological Transformation (Star Erosion), Color Saturation, and Histogram Transformation (stretch). We can easily agree that the image has too much saturation, but if we inspect the image carefully we will see that there's room for improvement also in the erosion process, as well as in the stretching. The article explains the procedure more in detail, so I'm only going to explain how we can take control of the workflow refinement procedure. First step is to drag the History Explorer to our workspace. It will create an instance of the History Explorer as a Process Container. Note that we have kept the unprocessed image, that we will clone to apply the improved workflow.  Now double clic on the process to be tweaked (or drag it to the workspace). An active instance will be opened. Then we will simply modify the process parameters to taste and drag again to the process Container. At this point I would suggest to previously disable the original process, as it can be confusing to have two processes side by side (the processes inside a Process Container or a History Explorer loss their description).  Once done, we can apply the modified Process Container to the unprocessed image. We may wait to all the processes to be applied to the image.  Left is the original processing workflow, while right is the improved one. We notice a clear improvement on the saturation amount, and how is it influenced by the rest of the processes. It is important to note that this is an example of an unmasked process modification. Masked processes are quite different. This is because the processes doesn't keep (as far as I know) the information from being masked, so to achieve our goal, we must proceed in another way. Let's modify the Morphological Transformation (Star Erosion), that is masked. If we do the previous procedure, the result will be a new process included in our Process container, but it will appear unmasked. I realized that processes dragged from a History Explorer effectively keep the masking information so we will use a Dummy Image (same as the target image, but black), to apply the masked process to it (using an improved mask), and from its History Explorer, to the Process Container.  Then we should reapply this modified Process Container to another unprocessed image and this is the result:  The stars are now less muted, with a better profile and less artifacts. Note the change of the mask in the Morphological Transformation. This methodology is not only useful to improve our workflow, but also to previously analize it. Que can analize, for example: -Compare differente noise reduction methods (MMT vs MLT vs TGV...) -Group iterated process in one (for example SCNR) -See the influende of a certain process over the whole workflow (for example star reduction) -and so on... As always with PixInsight, the limit is our imagination. As an example, this is an image of the influence of the star reduction application on the image I published recently.  Hope it helps. Clear Skies! Alberto |
