Pixinsight - Stretching using Pixelmath
Vložit
- čas přidán 20. 07. 2024
- Update-Version 5! - I have updated the stretching maths for both Linked and Unlinked in which I have added a function to modify the stretch curve shape, so a curve value of 1 is the same as a STF midtone stretch curve (normal stretch). If you increase the value to say 1.05, you will get more of an "arcsin" style curve which will help prevent blowing out the nightlights. I also added shadow clip protection. Last, I removed the GHS pixelmath script because it was difficult to use, and the actual GHS process within Pixinsight is so much better.
Here is the download link:
drive.google.com/drive/folder...
Note: The PixelMath script within this download link will be updated if changes are needed.
Enjoy!!!
More amazing tools Bill! Thank you so much for sharing all of these tools with us. I know you didn't create these, but you brought it to my attention and I'd never seen them before. I'm quickly amassing a great folder full of tools from your channel, looking forward to adding more.
Thanks Joe! Wait until you see what I post next week 😎
Thanks Bill for creating this nice and easy to use Pixelmath version of the Generalised Hyperbolic Stretch formula that I wasn't aware of.
Thanks Logan, I hope they help your workflow 😎
Thank you Mr Bill. I have been using the GHS since its direct use in PixelMath until its last update with the UI interface. Currently it is the only process that I use when stretching since it has Archsinh incorporated for color. I have used it on all kinds of objects, including globular clusters in which it has given me excellent results.
Ty so much Bill. It must be admitted that stretching is a pain, as there are numerous methods, techniques, and a sea of information that can make it difficult for a beginner to choose the best one.
Hey Bill, great video. I took your Median idea and modified it a little so I now have a pixel math process that outputs an image called "Reference" from the smallest Median value of Red, Green or Blue images (I pretty much always set my identifiers this way for my RGB images). This makes it easy to set a process with "Reference" already entered as the reference for Linear Fit. Here is the expression "iif(Med(Red)
Another great tutorial Bill, you are the MASTER !!!
Lol, thanks buddy!
Thank you Bill for sharing your knowledge and your tools!🙂
My pleasure!
Thank Bill, you saved me once again. PI through a wrench in my processing routine with the latest update. I had been using DarkArchons EZ Processing Stretch for quite a while in my normal stretching routine - I know I am lazy, but Dark is unable to keep supporting it so I had to find a new quick method which normalizes the background. Your PixelMath routines are very easy to use. Thanks Again. Cheers Kurt
Kurt, EZ Suite works in 1.8.9-2 thanks to Mike Cranfield revising the code. You just have to download the script files and install yourself. Here is the link regarding this topic. pixinsight.com/forum/index.php?threads/ez_softstretch-resurrected.21314/page-2
Step-1: Download files and unzip.
Step-2: Copy "EZ_Suite" folder (windows PC) into Pixinsight scripts folder: C:\Program Files\PixInsight\src\scripts
Step-3: Open Pixinsight and go to Scripts, Feature Scripts, click add and select EZ file. Done.
Thanks for the info Bill, I will try to download the links, not so much for for EZ Soft Stretch since I like the method you and others developed but to see if some of the other EZ processes work as well. Cheers
@@anotherastrochannel2173
Aloha Bill, great tip and thanks for throwing this (and the Star Reduction) tools out there!! Seems like I was spending 30 hours trying to process color RGB or SHO images. In the past, I would use PixInsight and stop after doing a mild stretch to non-linear and then kick it over to and finish in Photoshop. I'll have to add these tools into my new workflow. Thanks again!!
Thanks!!!
Thanks for sharing, Bill. That's cool.
Thank you so kindly, Bill, for sharing all this with the astrophotography community.
Thank you for such a nice comment!!!
Thank you, Bill.
Once again thank you for these tools. They are awesome. I do have one question as I may have got myself a bit confused. For those of us shooting with OSC, are we using the unlinked or linked stretch? I of course use SPCC first. I have been using the unlinked, but just re-watched this video and confused myself.
You are a legend, thanks very much..
Lol, thanks!!!
Just recently saw your video on star reduction, and it is fantastic. Completely changes the way I've been working and thinking about post-processing.
Do you happen to have a magic pixelmath button to correct oblong and wind-shifted stars? :D :D :D
Thanks for your contributions!
Glad you liked it!!!
Super Info after watching Sascha Wyss great....
Thank you so much 🙂
I think I'll see if I can bring this into the world of Siril too. It also has pixel math and you can now store presets.
I see the link to the Google Drive but how do I get that download into Pixinsite to use? I’m trying to figure this out for your one shot color to HOO took as well
Hi Bill. I am a complete beginner in Pixinsight, although I've been astroimaging for a while now. Your Stretching using PixelMath looks to be a great tool, but unfortunately I can't get it to work. I dragged the .xpsm file onto the PI workspace and got this message: "2 new icon(s) had to be renamed due to conflicting identifiers with previously existing icons. Original identifiers have been suffixed as necessary to make them unique.". This seems to indicate it has done something but there are no icons on the screen and I cannot find where they are in PI. Can you advise?
Many thanks
So, is the reason for using these to save using STF manually? The 3rd way sounds good.
Thank you for this, my processing icons are being replaced by pixel math icons!
Thanks for these! I've had good success with the HT scripts, but I get strange results with the GH script. The image looks properly stretched, but has a foggy film over it. Is the script good for OSC? Any ideas?
Same here. Before applying bills math, I had used the script and just couldn’t get it to work. Had a weird ordeal with the histogram as well.
Ronald, the GHS is a bit different as it is not "automatic" like the HT stretches. First, make sure you have done a background neutralization (BN) before running the script (Note: If you do PCC with BN on, this is done automatically). This is very important and I should have mentioned this in the video otherwise the backgrounds wont be the same. Next, try starting off with a "b" value around 0.25 to 1 range where as anything less than 1 is near the same stretch curve as the HT stretch is (1 is default and should be fine to start for all images). Next, adjust the "D" value to control the stretch amount (the higher the value the more the stretch). If you have a nebula core being blow out, then you will need to increase the "b" to be more arcsin style. Using this math does take some time to dial in but works well when it is, but again, make sure your background has been neutralized.
@@bbllaakkeeee Blake, I responded to Ronald if you want to give this a try again.
@@anotherastrochannel2173 I just tried your suggestions. As far as the b & HP channel are concerned, any adjustments there are negligible. The D channel creates a white film that remains unless the value is dropped to at least 170. At that point the stars and background looks good but the target (in my case a small galaxy) has almost no luminance.
@@bbllaakkeeee Can you send me your data via googledrive download link or simular, then I can test here? or you can email me at "anotherastrochannel at gmail"
Hi Bill, fantastic work (As usual) I have just downloaded the scripts, however the GHS script does not seem to be there. Has it been removed? Regards Mark.
Yeah, I removed it a while back after the GHS script/process came out as the actual GHS process works so much better.
@@billblanshan3021 Thanks for the quick reply, hope all is well. Regards, Mark
Thanks for your effort. I downloaded the script but GHS stretch is not included :(
Yeah, the GHS pixelmath script was a bit difficult to use and the actual GHS process in pixinsight is much better so I removed it.
Hi, I downloaded the script, I tried to drag it to the pixinsight window but the three HT and GH stretch work windows don't appear, what am I doing wrong? Thank you
Try and right click on your Pixinsight workspace and select arrange icons, that should allow you to see icons that where off the screen.
@@anotherastrochannel2173 ok solved, thanks
Hello I used the method but I got a totally white image out of it. I tried lowering down the value but still it is either too bright or too dark. any help please?
Your image isl white because your stf stretch is still on so after you stretch your image using the math make sure you turn off the stf stretch
@@billblanshan3021 I'm sure it's me because I'm new with PI but I reset the STF and the image is totally white!
@@magedsyehia can you send me your unprocessed data via download link and I will run through it and see what's going on. AnotherAstroChannel at gmail
Maged, here is a short video I made for you using your data, I hope this helps explains a few things for you regarding my stretching maths and also the display STF.
drive.google.com/file/d/1OT4hIPOIKthCSdEHpT958EqU_bkFXhh6/view?usp=share_link
Hi , is it normal to have the image washed out completely when using the Pixel math process icon, i can see the rosette neb but its over blown by a lot, i did not change anything in the script. ( GH_Stretch)
here is the link to the pix example.
drive.google.com/file/d/1m0II5KMcFnb0h9EzITBpYh2CTbD0cOpV/view?usp=sharing
Hey Mitch! Please make sure that you run background neutralization before running ghs math. I should have mentioned that in the video but I was assuming that everyone would have neutralize a background before stretching anyways. Give it a shot and let me know if it works
@@anotherastrochannel2173 will do. 🙂
@@anotherastrochannel2173 Bingo !!!! Thanks...
@@anotherastrochannel2173 Same problem as Mitch here but after running Background Neutralization I still get that white image.
@@eugennaiman1195 did you crop your image before stretch and BG?