Beta39-Beta41 Failing - errors, warnings, blank result

@astrogee ,

The first message shows that the integration field of view must be downgraded with a scale of 0.006. That indicates a severe blowing up of the field of view, so something must be way of with registration of the data. 

Are you trying to register a mosaic? If so, did you change the star amounts in 3) analyse stars and if so which values did you try? 

If it is a mosaic, with the new registration engine, 1000-2000 will be enough usually. More stars will make it slower and possibly less stable.

Mabula

Hi @mabula-admin, thank you for responding so quickly. No, this is a simple single session stack of a DSO, needle galaxy using only 13 of 66 images. There was a lot of wind so the session is a total loss but I deselected a few more lights that were heavily blurred or double image and I still get a blank result - I still saw the scale warning but not the others. Some of the pics are reversed to the others because of meridian flip but this is usually not a problem. The warnings I've never seen before with the same computer. I've aways used APP pretty much problem free but now all these warnings and the subject is the same as I have done before. I checked my last session (feb) and it was ok with beta39 so something changed? The OS  was updated to 26.3.

Hi Astrogee,

Okay, then i suspect that dynamic distortion correction becomes very unstable because of all the blurred and double stars.. Is all okay if you turn off dynamic distortion correction?

Mabula

Turns out that DDC was already turned off, so I turned it ON and re-registered, re-normalized and re-integrated and the result looks good.

Hi @astrogee,

Wow... that is really weird, so with Dynamic Distortion Correction on it works fine? I think that I need to test this dataset then if you are willing to share it? I really need to see why the error occured that you received. Can you upload it here:

https://upload.astropixelprocessor.com/

username: uploadData

password: uploadTestData

Please make a folder with your name and issue like: Astrogee-registration-issue-withoutDDC

And upload the data there and let me know once done 😉

I am glad that at least it works now and you get a great result here !

Mabula

Hi @mabula-admin, I've uploaded the files. It takes quite some time so I stopped short of sending everything. I think you can repeat the problem with the 13 lights I uploaded.  Let me know if you would like more files. Thank you for the help!

Hi @astrogee,

I have run the data that you uploaded, the 13 lights indeed are very bad captures, many double stars etc.. But if i process it without DDC all still runs and i do not get errors like yours.

Maybe I did not run it in exactly the same way? At least you got a result with DDC on, right?

What settings did you use to produce the error? 

Mabula

Hi @mabula-admin, I'm having trouble repeating the issue too. So I went back to beta39 - because this was what I was using at the start and I went to beta41 to see if it worked better but had the blank image also. I found something peculiar.

Firstly, I have reset all settings and menus to default. Then I unchecked Multi-Channel/Filter processing and Multi-Session processing. (All my results are with these settings.)

I found that the integration stalled on image 4, so I deselected it and the integration completed but the result was blank. I looked at the tool output window with all the analytical results and noticed that there was no results for registration for many of the images. 1 to 3 and 6 to 8 had no registration plot! But there is no indication that registration failed and all images were still included for integration.

I will go return to beta41 to see if I can repeat the issue.

So I tried beta41 again and I got a similar result to beta39. This time the stall occurred during normalization of the 8th frame. And some frames have no registration plots on the analytical results graph.

I realize that most of these images are horrible but without a blink utility its very hard for me to pre-select the bad images. APP seems to be succeeding randomly on this data. 

One thing I've always wondered about (when using APP in the past) is that all images are marked for integration while certainly some should be discarded due to very bad analytical results.

Hi @astrogee,

I will do more testing with your data. I did run it a couple of times and I did not get that weird behaviour, only frames that would not really register with a result with many double stars...

Certainly something is not okay, so I will work on it going forward.

Mabula