 Okay, this is part 2C of my series on Lagoon Nebula without a Star Tracker. In this part, we're going to look at Pix Insight, which is a paid program available on Linux, Windows, and Mac. I'm going to be using it on my Mac. I realize there is a newer version that includes StarNet, but I haven't updated that version yet because it doesn't work with the Mac operating system that I'm using, which is an older version of macOS. So it's just a personal thing, but if you are on the latest Mac operating system, I would recommend updating to the latest Pix Insight version where StarNet is included, and they've made some improvements too. But if you're like me and you're on an older version of Pix Insight, you can always just Google StarNet++ and download it from the Sourceforge site and follow the instructions to get it working on your operating system. That's right here under Pix Insight module. I did that. I'm on macOS 10.13, and I have StarNet working just fine within Pix Insight here under the processes, so that's an option too. Okay, anyways, enough about setting this up. We're going to be processing some files of the Lagoon Nebula. This was done without a Star Tracker, just on a tripod. I'm going to be using a script to do the processing, specifically the batch preprocessing script. I have tried weighted batch preprocessing, which is the newer update of batch preprocessing. For me, it doesn't work so well because I usually have lots of problem files, and so I haven't figured out a way to weight those appropriately. I need to do batch preprocessing with calibrate only on and then look through my files in something like subframe selector or blink before moving on to registration and integration. This might work for some people, but I'm still on batch preprocessing. The first thing I do is turn on this calibrate only. I typically don't register or stack using this because I don't know what my files look like yet. I like to look through them and make sure I'm stacking the best ones. It's pretty simple to use. You just add your lights, flats, bias, dark frames into these tabs here, just using the buttons down here at the bottom. I'll go ahead and do that. Everything is on my desktop here in this folder called Lagoon. If you want to follow along, you can download these files from my website. The link will be in the description. But basically I'm just selecting my files and adding them using the appropriate buttons here. I'll just go ahead and speed this part up since it's a little boring. I've added everything now. I'm just going to leave most things on their default parameters. For the most part, the defaults in batch preprocessing script are just fine. These are CFA images, meaning that they have a color filter array, specifically a Bayer array in front of the sensor. We're not using any master bias, master dark, master flat, but if you did have a master bias or something like that, you would just only add the master bias here and then just click this little checkbox next to use master bias in order to use that. We're not going to be doing any registration, so I don't have to add a registration reference image, but I do have to specify an output directory, so I'm just going to make a new folder called BPP, save to that, and then we can go ahead and click Run. It will take a while because we're doing hundreds of frames, so it has to create all those master calibration files and then calibrate all of our lights using those. We'll get a little time here, and then when it's all done, we'll move on to the next step. The batch preprocessing script finished, and I went ahead and closed out of it. Next up, we're going to go to the Process menu and go down to All Processes, Subframe Selector, and I use this process to determine which frames have the best focus, so finding a good registration frame, and it also will tell you if there's any frames that seem really odd and out of whack. So I'm just going to enter in a few little parameters here. Our subframe scale was around, I think, 17 arc seconds per pixel. I don't know where our camera is, so I'll just leave it on 1, change our camera resolution to 14-bit, and then I'm just going to click up here where it says Subframes, I'm going to click on the Add Files button, and in the BPP folder, there's another folder called Calibrated, then Light, then Debared, and so basically it creates logs, it creates masters of your calibration files, and then this calibrated folder is for where it actually calibrates your flats and your lights. We don't want to use these ones because these are in black and white, we want to use the debared ones where they're in full color, and then I'm just going to click and Shift click to select all of these files and click Open. Okay, and now I have 282 subframes ready to analyze, and so we're going to go ahead and analyze them or measure them using this process, and it does take a while, I'm going to go ahead and start it just by hitting the little circle down here, and you can see it starts over here, it does about eight subframes at a time, if for some reason you have to stop the process and then pick it back up, it does save it into memory, so I've often had to like, it takes up a lot of your computing resource and your memory, so if for some reason you have to stop it and do something else like some other work, it's fine, you can always pause it and then come back later, but it does take a while, so I'm going to go ahead and cut the video here and then we'll see what it says when it's all done measuring. All right, the subframe selector finished analyzing my frames here, I have 282 lights that it analyzed, and the first thing we want to look at is this measurements window, so bring that to the front, and don't pay too much attention to the actual numbers that I'm going to be showing you, because these are going to be wildly different based on your conditions, your focal length, all kinds of things, the main thing that we're trying to do with this is just look for outliers, so obviously there's at least, or there is two outliers here in my 282 frames in terms of full width half maximum, which is a good correlator to focus, so basically two frames are way, way out of focus and we should get rid of those frames. If I look at eccentricity, it's the same two frames that are way off, so obviously there's something wrong with those two frames, I'm guessing probably they were when I moved the camera to reframe, maybe it was still going or something, or they were just test frames when I was testing focus, so in any case I want to get rid of those. The other thing that I want to do with this script is find the frame that has the best focus, so I'm just going to do that by sorting up here, so if I sort by, instead of sorting by index, which is just the number of frames in order, I'm going to sort by FWHM, full width half maximum, and if I sort ascending, that will give me, at the top here, the frame that has the best focus. So I'm just going to write this down, it's image 1028. There are ways that we could sort them and then output them and you know eliminate ones we don't want using these expressions and the output option. I'm not going to show that right now just because with so many frames that's going to take too long and take up so much space on my hard drive, so I'm not going to use those features, but maybe in another video I can show how this expressions window works. So I'm just going to write down I want to use 1028 when I do image registration because that's the frame with best focus. Okay, then I'm going to sort FWHM full width half maximum descending. And from here, I can see that the two frames that I want to delete are 0904 and 1000. Yeah, the two frames that I want to delete are 0904, this one with a huge full width half maximum, and number 1000. So I'll just write those two down 904 1000. And then I'm going to do is just go into my Lagoon folder, go into the batch preprocessing folder calibrated light, debared, and just find 904 and delete it, and then find frame 1000 and delete that. Now, it doesn't show up here. This doesn't this doesn't update live that I've gotten rid of those. But I know that those are the two frames that were huge outliers that were there were big problems, just to double check, I can sort by a centricity here. And yep, it's the same two ones 904 and 1000. So getting rid of those has given me a nice fresh start here with all clean data. And then the only other one that I am paying attention to here is 1028. That's the frame that I'm going to use to register against. And just to double check that it's there's nothing wrong with that frame, I like to open it up and look at it. So you can just do file open. Again, you're going to go into that batch preprocessing calibrated light, debared, and find it in here 1028. Open it up, get this out of the way. Just apply it apply an auto stretch to this and zoom in a little bit. And yeah, this looks good. I don't see anything wrong with this frame. So I'm going to use this to register all of my lights to so we can go ahead and move on to that. That's all that I'm going to show in subframe selector today. Maybe again, maybe in other videos, I can show more about it because it is a useful tool. And we're going to go right on to image registration. So if you go to processes, image registration and choose star alignment. This is a process that will let you automatically align using a reference image and match star patterns. So I'm going to choose up here where it says reference image, I'm going to switch it from view, meaning that would be if we had a view open, we could use that, but I already closed the image. So we're going to choose file. And then I'm just going to pick that one that I just had open 1028. Click open. I'm going to leave all of these settings alone. The only thing that I'm going to change from the default is I'm going to turn off generate drizzle data that would take a long time if I left it on and I'm not going to drizzle this data because it's not dithered. And then where it says target images here, I'm just going to click add files and pick everything left in my debaired folder since we deleted those two problem files. Okay, and then I'm going to create an output directory, meaning just a folder. I'll put it just in my lagoon folder and I'll call it R E G for registered. Okay, and that's it. I usually just leave all of these other settings alone. The only time you would need to change them is if you were having issues. So for instance, if you were having issues with it not being able to find stars in your image, you could turn up or down the detection scales, you could change the noise scales, hot pixel removal, all these different things sensitivity to try to pick up more stars. Basically, if you hover over any of these, they do tell you a bit about them and how you would use them. But I think the defaults are going to work for me. So I'm just going to go ahead and leave all of those alone and just hit the circle icon down here to apply this process globally on all the target images. And this does take some time. The more cores you have in your computer CPU, the faster it will go, because it is a very multi threaded process. But for my five or six year old laptop here, it'll probably take, I don't know, an hour or something. So we'll just leave it alone and come back when it's all done. Okay, it finished aligning all of our images using the star alignment process. If I go over here and hover over process council, it tells me that all 280 images succeeded, zero failed, zero skipped, and it took about 26 minutes. Okay, so next, what we're going to do is go ahead and close out of that all the files that we just aligned with star alignment or in that reg folder, and we'll open up another process called image integration. So I'm just going to go to the process menu, go down to image integration, and then image integration again. I'm just going to go ahead and click this little reset button down here in a lower right, just to make sure all the settings are the defaults. Then I'm going to click add files, go into lagoon, reg, and select all the files in the registration folder, click open. I'll just make sure that I have 280 here, I do. Then in the image integration tab, I'm going to open that. And the only thing that I am going to that I'm going to open the image integration tab, I'm not going to change anything in here, but I just want to show you these are all the settings that we want to use average noise evaluation, evaluate noise, you know, generate integrated image, all these default settings are what we want. Under pixel rejection, one, we want to change this, we don't want the rejection algorithm set to no rejection. That wouldn't be smart. What we want is to change it to Windsorized Sigma clipping. And what that will do is anything that are outliers, anything that aren't in the central part of the standard deviation, meaning hot pixels, any other kinds of noise will get thrown out when we do the averaging. So we definitely want to turn on a rejection algorithm. I would recommend turning on clip high range in addition to all of these other clipping options right here. Okay, with that set, I can go ahead and click the circle button down here to apply this globally, meaning that it's going to apply it to all the frames in our input images list. And there's no place to save them because what it's going to do is after it integrates all these images, it will create the image right here in the window. And then we can keep working on it. So again, this is going to take a little time. So I'm going to speed up the video. Okay, that process is finished. We have the integrated file here. And to see what it looks like, we can turn on an STF auto stretch, you can do that in one of two ways, you can either just hit the little radioactive button up here, or you can open up the process called screen transfer function, which gives you a few more options. So I'm going to go ahead and open up the process. And what I want to point out in screen transfer function is that normally the channels are linked, meaning that it's going to stretch all three equally. But if you unlink the channels, then it will apply different stretches to each channel and you'll get a better color balance in your preview. So I'm going to go ahead and unlink the channels and then hit this radioactive icon to auto stretch it. And that shows me what we have here. So you can see this is a nice wide field shot of the Milky Way. You can see there's some color balance issues on each side where the signal to noise ratio goes down. There's also some registration artifacts down here at the bottom. So the first thing that I'm going to do is I'm actually just going to crop in on this picture so that we don't deal with these low signal high noise areas on the sides and at the bottom. And to do that, I'm going to go and open up the process called dynamic crop. If you go to process menu and go down to geometry, it's called dynamic crop. And it just works like any other crop tool, you can just draw out a box here, something like that, and then hit the green check mark to crop the image. Okay, that's done. Next thing I'm going to do is apply a dynamic background extraction to this. And so that's also under processes, it's the second option down background modelization and then dynamic background extraction. And you first want to click on the image. And then just click into areas where you should have a fairly neutral black. So I'm mostly just going to sort of pick areas here where there's either dark sky like in this corner, and this corner, or dark nebulae. You don't want to click on, you know, anything bright or a star, you just want to pick the more neutral black parts of the image that could be considered background. And I'm just sort of trying to add them with some coverage. So getting each corner getting each side and the middle, and usually about 10 or 12 will do it. You can see that this one down here is red. And that means that it wouldn't be included if we tried to run this process now. So what you can do to include that sample is increase the tolerance. So it's by default at 0.5, I'm just going to try to raise it to one. And then under sample generation, click resize all. And then you can see that sample down there changed from red to white, meaning it's now included. Okay, with that done, I'm going to go ahead and open up the target image correction menu right down here, and choose subtraction. And then hit the green checkmark. What it then does is it creates a background model, using the samples that we just gave it to create a model of the background, which you can see right here. It looks a little funny if I don't up the bit depth, there we go. So this is the background model. And then we can see what the image looks like with that background model subtracted. Okay, good, I'm going to go ahead and close that process. I'm going to close the background model. And I'm just going to put these pictures side by side. So you can see what it did. So here's the before. And here's the after. And hopefully you can see on your screen that on this one, the sides especially look a bit washed out and like low contrast. And in this one, because that background model was subtracted, it looks a lot better. So I'm just going to minimize this one, and we'll put it over to the side. And we'll keep working on this one. At this point, I'm going to go ahead and save this just so we have something to fall back to in case there's a program crash or something like that. I'll just save it to the desktop as an XISF file, which is the default format in pics insight. Okay, the next thing I'm going to do is I'm going to stretch this. But before I stretch it, I'm just going to increase saturation just a little bit with the curves transformation. So I'm going to open up curves. I'm going to switch to the saturation mode, which is this S over here on the right. And just pick a spot about two thirds up on the curves line, and just stretch that up like that, and apply it just by hitting the square right here. Okay, and hopefully you can see that increase the saturation a little bit. I'm going to keep this curves transformation open because I might want to do it a couple times more as we're stretching the image. I'm then going to take off the auto stretch, I'm going to open up histogram transformation, which is under intensity transformations in the process menu. And I'm going to start stretching the image. So I'm going to open up, I'm going to say I want this to apply to integration underscore DBE. That's our background extracted image right here. And I'm going to take this middle slider and move it pretty far over to the left, and then hit the square to apply. You can see even with that much stretch, we're just barely seeing some stars. So I'm going to apply it again. Okay, this time I'm going to do a little bit less of a stretch, something like that, and hit the square. Okay, now we can start seeing the image. And so I'm just going to keep backing off the stretch just a little bit. And now we have enough breathing room here that I could take this shadow slider and just move that over a little bit to the right. You want to be careful here, down here it tells you how many pixels you'd be clipping completely to black. And so right now it says I'd be clipping 172 pixels to black, I'm comfortable with that amount. But if it said like 20,000 pixels are clipping the black, that'd be way too much. And so I would I would want to back off that a little bit. Okay, I'm going to apply this. Actually, before I apply that last stretch, let's go ahead and just add some more saturation just by applying this saturation curve. Okay, cool. And then I'm going to apply the stretch. Perfect. Okay, and honestly, you know, this is already looking really nice. We might not really have to do much else to it. You might be happy with just that amount of processing and picks insight. All we did was a background extraction, a crop and a and a stretch. There are a few things though that I'm noticing that could be improved on. So I'm going to close these two processes. One is that we have a little bit of green noise over on this side of the picture. So I'm just going to try to run a noise reduction scnr green. And I'm going to start it at something like point three, because I don't want to run it at full strength or it usually takes out too much green and screws up your color balance. But I'm just going to try to take out a little bit of that green noise. Okay, I think that did a pretty good job. Here's the before you can see this this side of the picture, especially really as this green cast to it. And there's after and it looks a lot more naturalistic. But we could try running it again if you think that it needs it. So you can always just apply it again. Okay, and it took out even more green. Here's the first, you can see it's really green on the side. Here's the first application of scnr green. And here's the second application. And yeah, I think I like two applications. So I'm going to leave it like that. Okay. The next thing we can do is remove the the stars from the image and boost the Milky Way a little bit and then add the stars back in. To do this, we use a process called star net plus plus. If you have the latest version of pics inside installed, it comes with pics inside now as a main process. You can always update to the latest version to get it. Personally, I'm on an older version just because of the age of my Mac. But for most people, you can just update to the latest version and get it that way. Okay, for a wide field image like this, I don't think the stride of 128 is going to do a very good job. So I would usually lower it to 32 or even 16. Let's try 32 and see how it does. So to apply it, all you do is either just hit this apply button or drag the new instance icon onto your image and let go. And it does, of course, like everything in processing, take a while. So we'll let it do its thing and catch up when it's done. Here's the result of the pixel math operation. It took 49 minutes with a stride of 32. One thing I forgot to do is make a copy but we can always do that now. And then I'm just going to undo the star removal on this one. So now we have a starless image and the image with stars. And let's just name them appropriately. So I'm just going to type in stars there and double click on this tab and call this starless. And I'll go ahead and save these as well. Okay, then what I'm going to do is on my starless image, I'm going to apply some curves. So I'm going to go to process intensity transformation curves transformation. And I'll leave this saturation one up like that. But I'm also going to just do the RGBK and just do a slight S curve like this. And let's go ahead and hit the preview, the real time preview to see what this looks like. Oh, and that's the real time preview of the star image. That's not what we want. So let's click on this one and then do the real time preview. And I think that's a little bit too much on the saturation. So I'm just gonna edge off that a little bit. And I'm just going to add a little bit more shadow definition just by making this S curve just a little bit more extreme. Okay, let's go ahead and apply that. Okay, so here was the before a little bit washed out. And here's the after just a lot more extreme contrast. I think that it made the overall picture just a little bit too dark. But the thing is when we combine these two, the picture is going to get a lot brighter. So that's fine for now. So I'm just going to leave it like that. And let's go ahead and now combine the two images back together. And we're gonna do that with pixel math. So you can open up pixel math just by going to process pixel math, pixel math. And we can use a single RGBK expression. What we want to do, this is a little bit funky, and I'm not going to be able to explain it exactly what it's doing. But the tilde key, which is under the escape key on your keyboard, it looks like a little sort of curve dash, that's an inversion of all the pixel values. So we want to do tilde parentheses, tilde stars, times, which is the asterisk, tilde, starless, and parentheses. Okay, let me. Okay, and what this is is this is the equivalent of a screen blending mode in Photoshop, but in pixel insights pixel math. So let's go ahead and see how that looks. Okay, so I have that typed in, then what I'm going to do is I'm going to open up the destination tab and change it from replace target image, because I don't want to do a replacement to create new image. And I wanted in the RGB color space, and then I'm just going to hit the square to apply. Okay. And we get this image. And hopefully you can see other than just being brighter than our original image, it also made these nebulae pop out a bit more. And a lot of the dust, the dark lanes in the Milky Way as well. To make those dark areas of the Milky Way pop even more, there's actually a script in in picks insight, it's under the script menu up here. And it's under utilities. And it's called dark structure enhanced script. So let's try running that as well. Our new image five, I'll just hit okay. Okay, so this is the before. That's the after you can just see that the dark dust lanes throughout the Milky Way got a little bit enhanced through that. The next thing we want to do though is reset the black point on this image. And we can do that again with curves. So I'm going to open up curves transformation again. Reset it and open up a real time preview. And then I'm just going to drag down here over in the shadows area. Okay, so now we can see what the starless boost did to the image. Hopefully you can see that it's pretty subtle, but it definitely boosted up the the nebula regions a little bit. And I think sort of improved just sort of the drabness of this one, by making it a little bit punchier. We can also zoom in on some feature and see how it looks. Okay, so there's our lagoon and triphid, I'm just going to apply that same view over here to the just stars image. And hopefully you can see in this view that it just it minimizes the sort of the dominance of the star field a little bit and makes the nebulae I think pop out a little bit more and appear just a little bit punchier. Let's look at another feature here. So here's the star cloud in our new one. And here it is in the old one. So it still looks like a really beautiful star cloud, but it just I think it just looks a little bit more refined over this on this one. And then here's the Omega and the Eagle. And again, I just think they pop out a bit more in this one. But basically, I'd be I'd be happy with either image. So if for some reason, star net isn't working for you, I think, I think this looks just fine. But this just looks a little bit more finished to me. Okay, and then for saving, we can always save it as a pics insight file. So I can just call it lagoon PI final. You can save it off as a TIFF file. And that's good if you want to bring it into another program like GIMP or Photoshop for some final fixes. But we're going to call this done. And so I'm just going to save it as a JPEG. And I'll go ahead and increase the quality to 100. Okay, and then here's our final result out of pics insight. This is my final video in this lagoon no tracker series. So I can now sort of evaluate the different pictures we've produced. I have to say, without doing any color calibration, I think that the stars and colors in the pics insight version are just really good. I didn't really have to do anything with it. And I think it also did a pretty good job of retaining detail. I didn't try any noise reduction, which I think is is good for this image. I don't think it can really hold up to much noise reduction. I think that was actually a mistake when I did it in Photoshop, because there is there is this is the stock DSLR, if I haven't said that. And there is quite a bit of detail you can see the cluster in the core of the lagoon and you can see sort of the lobes of the lagoon and how it's more bluish in the middle and then is pink towards the edges. I just think everything came out really nicely in this pics insight version. So for me, pics insight is still really the best processing software for astro images that I have. I understand why some people though don't want to buy it because it is a bit expensive. But to me it is it is worth having in my toolkit. So this has been Nico Carver from NebulaFotos.com. I hope you enjoyed this video can always leave questions and comments below. And till next time, clear skies.