batch preprocessing light files problem

[jamesRC]

Hi,

I have been trying for a long time to get the script file "BatchPreprocessing" to work with my data set. So far, I get far enough for it to produce output folders of Calibrated, Logs, Master and Registered, I get that far enough before non correlation of my lights stops the script. I have tried this with my original lights, with those lights run through "StarAlignment", and with the lights in "Calibrated" from a previous run. Everything fails. Just what the heck kind of files is the script lights process expecting? I hope this is clear enough as information for helpers. Do the expected lights need to be already made perfect, in some way? This is not sarcasm, PI is unbelievably demanding sometimes!

James

[jamesRC]

a continuation:

I tried not using the BatchPreProcessing script. I took the individual steps described all the way thru 'Inside Pixinsight' pp16-24. This is the best way to refer to the long series of steps, which I carefully followed. I ended up with a set of "fully processed lights" as I named them. It was still necessary to run Image Registration/Star Alignment to get the stars lined up. Following this, I tried ImageIntegration to put all the "fully processed lights" into a final file.

I get the same error code, re no star points aligning. Sorry forget the exact words but thats the idea.

Now I have nothing left to try for sure!

Please help.

Thank you,
James

[ngc1535]

Hi James,

I am confused- you say below you had an error concerning "no star points aligning"- but you also imply you succeeded in registering the images since you proceed to ImageIntegration.

I suggest, with regards to BPP- that you first start with Calibrate only. Get that part to work. The other two steps of registering and combining (using ImageRegistration and ImageIntegration) should probably be done "by hand" anyway.

The issue with regards to "non correlation" of your lights (presumably not flats) deals with PixInsight trying to do dark frame scaling based on noise characteristics of your data and a master dark frame.

Having the exact language of the error can help. Seeing your data is likely even more helpful.

-Adam

[jamesRC]

Adam, thank you for responding.

I'm confused too :-) The lack of clarity in my question shows how uncertain I am in my understanding of PI.
I was trying to say: I was attempting to do image integration the manual "long" way vs the script "short" way. I can not get the script to run because of an error message I unfortunately can not remember. This message is to the effect that PI can not find star pairs in the images, meaning there isn't any acceptable correspondence in the light images. I still have no answer for this error, and eventually have to get back to it.

At this time, I just successfully integrated the images the "long" way, so have made some real progress.
Below is notes I made to myself on the way through this (following instructions in "Inside Pixinsight"):

1. StarAlignment applied to original subs. Output is a set of modded lights with "_r" suffix added to names;
2. ImageIntegration applied to original subs. Result is one integrated image with stars not registered.
3. ImageIntegration applied to StarAlignment-ed ), the new "_r' files. Result is an integration of all the subs, with the stars
aligned. This is a good result, but it does not make use of my bias and flats.
3a.Attempt to integrate bias to make a Superbias. Result is:
*** Error: L:/astronomy working 9 nb filters/IC 5070 Pelican Nebula/bias use for Oiii too/B_1-4000s_2017-09-30_19-09-44_Bin1x1.fit: Zero or insignificant signal detected (empty image?)
<* failed *>
3b. Turn off "evaluate noise" set to "no normalization" and turn off "noise evaluation" and it works. Result is bias-integrated image.
4. Run ImageCalibration/Superbias. Result is superbias image. Checked with Ctrl+A. OK. Saved.
5. Clear bias files, load darks. All settings unchanged. Result darks-integrated image. Saved.
6. Start ImageCalibration, nec. prelim for flats. Load flat files. Load superbias. Load Master Dark saved from integrating darks. Check Calibrate.
apply. Some confusion here calibration vs integration and both done for flats!
7. PI says 41 successful image integrations done with lights. Wow! A first!
This is not really a question.. It's more a description of how I eventually made my way through image integration. The question re using the script to do this, is still there but on the back burner for now.

The integration vs calibration processes get confusing, as well as the way some types of files are presented on the desktop and must be saved, vs others that are automatically written out.

Clear?  NAH!

James