PixInsight: StarAlignment, Debayering and Hot Pixels!

[@u2pilotjt]

As usual, excellent explanation of what's going on in the background. Thanks!

[@AdamBlock]

Thanks!

[@marcusandrea3477]

Thank you for these complete and interesting explanations

[@billtinsley5385]

I really appreciate how you explain and demostraste the reasoning behind the process. Great work!!

[@AdamBlock]

Thanks Bill!

[@frankbraker]

So helpful! This has had me scratching my head for the longest time!

[@AdamBlock]

Thank for watching!

[@DebashishGhoshOfficial]

Your approach is very scientific. I really liked it

[@Aerostar509]

Very informative, I was hoping you would have zoomed into the hot pixels and then blinked but we know they were the reference points. I am always amazed that people are smart enough to write these actions to process the data with so many variables that affect the outcome. Looking forward to more teaching videos from you. Tim McCollum in Georgia

[@AdamBlock]

Yes, for completeness I agree. The console is saying this with no translation- but I could have blinked the images to show the hot pixels did not move. And...don't wait for more videos... Join my website today for hundreds of hours of more "informative" lessons. :)

[@stevesolon6710]

Many many thanks, Adam. I was beginning to pull out what's left of my hair trying to align my CMOS data. I did not know about the SuperPixel method. I had Calibrated correctly, Cosmetically Corrected, (but unfortunately did not click CFA), then Debayered. I even painstakingly went through every one of 67 ten-minute subs and used the Clone tool to remove Outliers, etc., and still couldn't get the images (all perfectly centered) to align. After watching your lesson, I re-used CosCorr, and clicked CFA, then switched to the SuperPixel method, and was able to align all the images WITHOUT having to manipulate Star Detection, etc..So my problem, as you suggested, was 1.) Clicking CFA, and 2.) using SuperPixel. I echo other reviews: Very well explained. I thank you. - Steve

[@jonathanpearceff]

Great video, Adam. Very informative. And what a super piece of software PI is. The more I see of it (demo’s such as these), the more I think “why use anything else?”

[@michaelogle638]

Adam: After going thru the FastTrack WBPP/StarAlignment/Image Integration process on my Heart Nebula data, I ended up with stars that were pretty good except that the stars in the lower right quadrant were elongated. Then I remember this video. Could using VNG for the CFA debayering in WBPP be the source of my elongated stars ONLY in the lower right quadrant? I am using a modified Canon T5 (IR blocking filter removed) with an Optolong LPF_l-enhance_clip-in filter with a 200mm telephoto lens mounted on a SkyWatcher EQ6R Pro mount. Tracking was around 1" during the 35 180sec run. Michael Ogle

[@AdamBlock]

No... VNG is just an interpolation scheme between pixels. It is symmetric and global...it cannot distort anything. Lens can certainly have all kinds of aberrations that appear to occur on one-side of an image (usually because of the centering of the light beam plus the aberrations).

[@michaelogle638]

Thanks Adam

[@Hidden.Light.Photography]

Can you use the cosmetic correction with CFA checked and super pixel then pull those files into WBPP and run the process from there?

[@AdamBlock]

If you are using WBPP.. you do not need to check the CFA box. WBPP will know what to do since you tell it the images are CFA. No need to do any additional steps.

[@steveweintraub1113]

Thanks again Adam for the tutorial, it really clears things up. Question; to avoid most of what you covered, if you shoot/add dark(s) with the OSC or CMOS data, will this eliminate the hot pixels? Or at least the majority of them so this process can be avoided?

[@AdamBlock]

That is definitely the idea. Uncooled sensors are tricky though- so understanding that hot pixels are part of the game even in perfectly calibrated data is important to understand.

[@Steve_The_Ignorant_Astronomer]

Good video , wish I had your knowledge.... Pix is like an onion , so many layers

[@BrokenPik]

great video.. Q how are you blinking 2 frames 11:27 on top of each other . keystrokes ?

[@AdamBlock]

In my lessons at AdamBlockStudios.com I demonstrate a handful of ways to blink images in PixInsight. You really should join! :) The answer to your question is CRTL+PGDN .

[@BrokenPik]

@@AdamBlock oh thank you i myself have 60 videos on KZbin for absolute beginners and some intermediate. and still i will join your lessons , because this hobby never ends. just waiting on some $$ LOL ... oh and im in CAD so the price is way higher since your tutorials are in US Funds.

[@Steve_The_Ignorant_Astronomer]

Can I ask a question , I have a osc asi294 color.. When I tried the hot pixel with and without the cfa checked the hot pixel was detected. Should I always keep cfa checked ?

[@AdamBlock]

Yes. It will handle the neighboring pixel calculation properly when it knows there is a CFA. Otherwise, the substituted value- though dimmer, might not be the correct color. This is a subtle thing- but not doing this right will just add color noise to your images.

[@Steve_The_Ignorant_Astronomer]

@@AdamBlock Thanks for getting back to me.. Love your videos... Will have to check the box from now on

[@davepastern]

tldr - cosmeticcorrection first, then debayer, then star align with colour cameras that don't have calibration frames...got it!!!

[@AdamBlock]

Almost. I used this as an extreme example. Use Cosemetic Correction even with proper calibration frames. It will still help!

[@jimvariadakis]

I have a problem with my process! I follow your steps but the cosmetic corrected image have a problem with the debayer and the colors are strange. In some areas it is purple and in some others it is green while the nebula in the center is desaturated, meaning black and white!

[@jimvariadakis]

I tried it with the CFA and without the CFA and the color patterns are different after the debayer

[@AdamBlock]

@@jimvariadakis It sounds like this solution might not be a perfect fit. If you are a member of my site- I offer help when I can. (Members can submit questions to the forum on my site at AdamBlockStudios.com)

[@jimvariadakis]

Ok then! Thanks...! I will send you there!

[@astromatz]

Excellent! Thank you very much! Very well explained :-)