15th Feb 2024: Astro Pixel Processor 2.0.0-beta29 released - macOS native File Chooser, macOS CMD-Q fixed, read-only Fits on network fixed and other bug fixes
7th December 2023: added payment option Alipay to purchase Astro Pixel Processor from China, Hong Kong, Macau, Taiwan, Korea, Japan and other countries where Alipay is used.
I have a 2-panel LRGB mosaic that is registering fine but one panel looks quite washed out compared with the other.
Set up each session as a different panel and I set up 1st degree and 4th degree LNC hoping that would even things out.
My normalization is set to Advanced/multiply-scale/BMMV with neutralize background off.
Anything else I should try to get the panels of the two images to better match up?
Thanks!
Mathew
So how does that panel look like when you integrate it first (before making the mosaic)? Is it quite different in quality somehow? What I usually do is to correct light pollution etc. per panel first and then I make the mosaic. Seems to work quite nicely, LNC then takes care of the rest.
here you can see the mosaic and then the session 1 and 2 individually, just with an stf in pixinsight
Oh that looks very strange. Can you give some details on how you made the mosaic and what the input data exactly is? Is it raw files or did you already process them somehow?
Would be nice to have the 2 panels as well, I can have a quick look then as to what might be up.
Sure thing - should I share with your fb acct or is there a way to DM links to the raw file?
You can upload it to our server, or.. since it's 2 panels; maybe share a wetransfer and e-mail it to vincent@astropixelprocessor.com. Whatever works best for you, if you want to upload to our server;
Go to https://upload.astropixelprocessor.com and use upload2 as username and upload2 as password.
Create a directory named “mr1337-mosaic” and upload in there. Thank you!
Sent along some files for reference. and here are the rest of my settings. Let me know if you need any more detail:
Register
Pattern Recognition Descriptors: Quadrilaterals
Scale Start: 5
Scale Stop: 10
[checked] Use dynamic distortion correction
[checked] Same camera and optics
Registration Mode: mosaic
Projection Model: projective
Normalize
Mode: advanced
method: multiply-scale
scale: BWMV
Integrate
integrate: automatic
LNC degree: 1st (also tried 4th) degree LNC
LNC iterations: 1 (also tried 3)
[checked] enable MBB
10%
Ok, thanks for the data, working on it now. Looking at the mosaic, it seems like you didn't use outlier rejection for the panels, is that correct? Or at least I'm still seeing satellite trails. What calibration data do you use? Did you use LNC also for creating the panels? If not, that may already improve things.
Furthermore, I didn't get the washed out effect, but the only difference is that I turned off "same camera/optics" and switched off dynamic correction (didn't work switched on in this case, which can happen if the data doesn't need too much correction). The registration RMS wasn't great though, so I'm wondering also how you made the mosaic, what kind of overlap etc?
Overlap is around 10%
Tried manually using the outlier rejection and that helped balance the frames out a great deal. I saw registration issues after doing that, specifically on the right side. After reintegrating with dynamic correction off I still saw some issues and went in and removed a few jumpy frames from the lights. After doing that and going back and turning on dynamic correction again I got a really nice mosaic. I think the only difference now is a few frames that had moved a bit and not using the automatic integration option. I guess that will be my go-to now is just to make sure I don't use that automatic feature for mosaics since that is pretty much the only change I ended up making in the end.
Ok, interesting still as normally that would work pretty much automatically. But, good to see you got it working nicely!