Batch Preprocessing with various different length exposures

[Mick G]

I have taken images of M42  at 5, 15, 45, 90, 180, 300, 480 seconds (5 of each).  I am trying to get detail in both the bright center and the dimmer outer regions.  I am also trying to figure out a method to best process this type of data for future use.  It was suggested to me to do each group of 5 seperately, but use the same registration image for all despite the difference of exposure length.  Then combine these images.  The problem is I am getting an error message  "Cannot integrate less than three frames".  In the section on the left (telling what is going on, I don't know what this is called) it says "unable to find a valid set of star pair matches".  Is this the right methodology,  any idea what I am doing wrong, and what tool would you use to combine the images?

Any help will be greatly appreciated.

[pfile]

if all the frames are registered to a common frame, you can try HDRComposition to merge them together.

the thing on the left is the 'console'.

or are you saying that you were unable to register the images at all, meaning right now you have a bunch of calibrated but unregistered images?

rob

[Mick G]

Sorry, I still just learning.  I'm not sure the difference between calibrated or registered.  In the past I have entered all my lights darks and flats in Batch preprocessing hit "run" then in the master file I find a light image to continue processing.  These masters are what I had planned to combine in HDR combination.    In looking at the results of one of the failed processings there are all five 480 second images calibrated and only two are registered. In this case the reference image was a single 300 second image. 

[pfile]

ok no problem.

unfortunately when something like this happens the only way to debug it is to use the StarAlignment process on it's own to see how many stars were detected in your various images.

StarAlignment is very good at finding stars, so when this happens it's generally due to soft focus, or various tracking problems, such that the stars are not very "star like" and then StarAlignment has trouble finding them without tweaking the parameters.

an alternate possibility is that the calibration frames were not good and so the calibrated frames are essentially corrupted, making it impossible for staralignment to find stars. you'll have to open up your _c files and take a look at them.

rob

[Mick G]

SO, If I understand this correctly, I should use the Batch Preprocessing script to combine each group of like time frame images with a registration reference image from that specific set of images separately.  Then the stacked lights from the master files produced from each set should be put in the target images part of the Star Alignment tool.  For the reference image I would think one of the lights from the middle of the time ranges would be best.  Agree??  Then click apply.   Are there any settings that should be altered from the default settings?

Bye the way Jedi Master the force IS with you.

[pfile]

there are 2 ways of doing it. one is to simply put all the images into BPP. the script will calibrate, debayer (when "CFA images" is checked) and register all the images to the reference image. assuming that succeeds, you'll have a bunch of c_r.fits images (or c_d_r.fits if debayering was used) that are all aligned to the same reference image. BPP will also integrate all the images, which may or may not be what you want. but you can throw the integrated result away of course.

then you can stack the different exposure length groups together using ImageIntegration. and since they are all already aligned to one another, you can then take those masters and put them in HDRComposition.

however, if you do as you're describing, yeah, you'll have multiple integrated images which are not aligned to each other. so then you have to register those using StarAlignment, before proceeding to HDRComposition.

but in your original message, it sounded like BPP failed to register all the images. so i was suggesting ways to try to debug what happened. SA has a mode where it shows you where the stars are that it detected in your image. you can use this mode to see if it's picking up on hot pixels instead of stars, or if finds too few stars, or what. the dropdown menu in one of the upper panes of SA has a mode selection. "detected stars" is the special mode that just shows you what stars were detected in the target image.

rob