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.
Any idea what would cause a mosaic to look like a fisheye projection - see attached. It was created from three Ha masters, one from a different camera. When I tried to include 3 O3 frames, I also got a message "Please supply more than one frame for integration"
Here are more details. Over the last couple years I've imaged the Crescent Nebula 3 times, twice with an ASI 1600 and once a QSI 690. All three times I used Ha and O3 filters. Once I got APP I decided to try to integrate the three sets, both as a mosaic and as a simple combination like the Rosette Nebula with 5 sets of data. I ran a Multiblend Multisession combine with over 300 lights and 200 cal frames. Worked fine, got 6 masters. Following the tutorial, I tried to then integrate the masters. I loaded all 6 masters into the Load module but it didn't recognize them as Lights. I had to turn off the Identify Masters and Integrations check box and then it worked. I ran a mosaic integration and everything rand fine - analysis, alignment, and normalization - on the 6 masters. Then I got a message "Can't integrate one frame", it seems APP was ignoring 2 of the O3 frames. I ran just the Ha and no problems, but opening the Integration it was clear the mosaic had run into a problem. See attached image. I did have same camera and optics turned off, the distortion correction turned on, and alignment scale set to 10 as recommended.
Edit: I did NOT use LNC on the individual frames, and I only had MBB set at 10%, when there's really 50% plus/minus. But in the big mosaic on the APP home page, some panels overlap entirely I think? Maybe just a normal integration would be better? but even so I'd need to understand what's happening to tell APP there's only one O3 panel.
A note: these really weren't a true mosaic - there was a lot of overlap, maybe 50% or more. Star detection only found a little over 500 stars per frame. But registration did run ok with less than .3 pixel error. Star sizes did vary quite a bit, up to 5 pixels. I was doing this just for practice but I would like to understand maybe what went wrong and why APP was ignoring 2 of the 3 O3 masters for Integration?
Mm, I may have seen this once before a long time ago. Just as a quick check, could some frames be flipped vertically? It shouldn't really cause this I think, but since you used various setups, it might.... ? You can put a checkmark on for that in in the "register" (4) tab. Maybe check it first with a few subs to test of both camera's. Since the overlap is so big, you can actually relax the alignment scale to, say, 8. If this all doesn't work I need to tag Mabula again. 😉
Yes one camera could be flipped. I'll try that. I'll turn on LNC 1/1 for the individual subs and set MBB higher 25% say. Also reduce the lights integrated t0 80%. But I still don't understand why APP couldn't use 2 of the O3 panels in Integration, unless it's due to one of the settings...
Figured most things out. APP didn’t recognize my O3 panels because the filter info wasn’t found in the header. Once I explicitly assigned the frames to O3 it was fine. I changed MBB to 40% and no more fisheye mosaics.
I did get sort of a seam in one area. It ran vertically about a third of the way across the frame. It was like a vertical band stretched across about 10 pixels. Stars were stretched horizontally across the band with points at each end, like little barbells. Same thing happened whether I registered Normal or Mosaic. Frankly I don’t even think it was on a border between frames, just a frame got stretched at one point. But I have to say it was cool to get it to work as well as it did, with no intent ever that I would merge these images.
Wayne
I think I've seen that stretching as well before, most likely it can be solved with a few little tweaks, Mabula (@mabula-admin) has a lot more experience with those cases. Problem for me to be able to give advice on it, is that it always worked. 🙂 It should definitely work really well, in cases when different data is combined and most certainly when you are actually going for a mosaic from the start. But yes, to get to results like this, it took me days with other software.