![]() ![]() ![]() So, how does this all apply to PixInsight? Why? Because it's pretty much a "How-to" book! That's why it worked for me! I've pretty much devoured Ron Wodaski's book. Now that I'm on the chapters for image processing, I'm filled with lots of information on *why* (formulas, formulas, formulas!) and very little how! Those first chapters of AIP4WIN were now explaining the *why* part of image acquisition and calibration. ![]() I actually devoured the first few chapters of the book, but those chapters were talking about image acquisition and calibration: things that I already know *how* to do. For some reason, seeing formulas shuts my brain down, even though I know such knowledge is good for me.Īs I'm typing this, I'm thinking to myself "Wait, that's not how I took AIP4WIN." But it just dawned on me: I'm kinda frozen on the image processing part of the book. Which happens to be written the reverse of the way I like to operate: Why (SGBNR processes and formulas), then How. I went straight to the help topic on SGBNR. When I downloaded PixInsight, I was after the one "How": how to use SGBNR to smooth my images. Once I see progress with my exercises, then I go back and start reading the "Why". Let me explain: my workflow/learning curve with any new software is to first learn the "How" it works. You just made an honest man out of me when you asked if I read the documentation deeply. When I downloaded and started using PixInsight, my gut reaction was "What?!? I have to learn yet another huge software package just so I can smooth out my images?!?!?" This is on top of still climbing the learning curve on packages like CCDSoft, TheSky6, Photoshop, AIP4WIN. Truth was, I was looking for a relatively quick and easy to use application. If you recall, I heard of PixInsight in an earlier post in this forum where I was asking about noise reduction programs. Next, I must confess that I have yet to fully explore either the software or the help file. You are going to have a lot of freedom to customize PI!įirst of all, I want to extend my apologies for anything I have said that may have criticized PixInsight too harshly. It will be avalaible only in the STD release, and every process will be builded as a module using that API system. That would be very nice! The API will work based on C++ code, and will include a visual interfase to let you create the enviroment. >I'm also in the software industry (business applications >software), so I would also interested in helping write a >plug-in, time permitting! Anyway, almost everything is covered in the documentation. We are aware of that, and we'll try to develop more tutorials on how to play with the interfase elements. >I like the potential of PixInsight, and hate for its >interface/help to be its downfall. Also you can try a loop between wavelets/sgbnr to remove the noise in steps, while you enhance some detail layers. The best we can do is to show you examples with specific images to let you see how each paramether work, and how to adjust it to find the optimal paramether.īTW, we found that SGBNR makes much better results when used after wavelet's noise reduction. It is impossible to give such default values for specific object types becouse each image is so different from another that there is no correlation between object type and SNR between different people's pics. If it varies a lot from the highlights to the shadows, it is convenient to use a inversed luminance mask. You must vary both values untill only the noise is blured, and that depends from the signal to noise ratio. The most critical paramether is the edges protection threshold. ![]() For galaxies, start with these >parameters," etc. >If I may make a suggestion as to how to make the interface >or help a little friendlier, how about some suggested >parameters to start with? I've noticed in other software >literatures for image processing that it has been a huge >big first step to say, e.g., "For nebulae, try these >parameters for SGBNR. We'll try to upgrade our documentation, examples and tutorials to help people use all the processes, from novice to advanced users, and we need your imput to do that! That means that there will be no "one buttom" processes in the short/medium term. Our politics will be to encourage you to join into the state-of-the-art techniques and use them in your benefit. > It also doesn't help that I look for "quick fixes". >but as Dave mentioned the help and the user interface > definitely could use some additional friendliness.Ĭould you be more explicit about that? It is very important for us to know your impressions. >I'm jumping in here as well, if you don't mind. ![]()
0 Comments
Leave a Reply. |