Problems in image calibration

[rfrancis]

For quite some time now I've been having problems in image calibration. Specifically, if I do not use the option to calibrate the reference frames (bias, dark, flat) I end up with a calibrated frame which seems to be reduced to essentially 2 values -- I attach 2 examples, one without calibration and the other with, after applying the screen transfer function.

I don't understand what's going on here -- calibrating the master frames should not create such radically different results, should it?

The original files are here, in case you want to experiment: https://www.dropbox.com/sh/uiywot30fviwf03/jrnQ4Yj-Kr

Thanks for any help,
Richard

[sreilly]

There are two processes that can be used for calibration, Batch Preprocessing and Image Calibration. Are you using Image Calibration? If so can you show a screen capture of your settings? How dd you make your master calibration frames? Are the darks of equal time and temperature as your light frames?

-Steve

[rfrancis]

Thanks for the response.

I have been using Image Calibration. I'm not now at the relevant computer, but the essence of the settings is in my text: for the Master Bias, the Master Dark and the Master Flat I either selected the calibrate check-box or not. If not I get the "bad" result, if yes, I get the "good" result.

There is also a checkbox and text entry window (defaulting to 1024) for Optimise in the Master Dark pane. If I leave that selected (default) then in this case the dark does not remove hot pixels. If I deselect it, the hot pixels are mostly removed.

The darks have the same duration and temperature as the lights (-20C and 600s in this case).

The Master frames were made using Image Integration, from more than 50 sub-frames (I forgot the exact numbers).

Thanks,
Richard

[sreilly]

My point was if you already calibrated the flats when you made the master frames you don't want to repeat that again. Same with the darks. If they are a match for the light frames do not check any boxes and don't use a bias frame as it isn't needed. I would also suggest trying the Batch Preprocess script and if you don't want to align and make master frames of the lights then check "calibrate only". Here's another good video from Harry on using the script http://www.harrysastroshed.com/Calibration.html

In case you haven't seen the video on making master calibration frames in PI see http://www.pixinsight.com/tutorials/master-frames/en.html

Steve

[jkmorse]

Richard,

Had the same problem recently but it is just a function of how STF sometimes works with uncalibrated images as far as I can tell.  Way I dealt with it was to simply open all of the images, make sure the setting are right in STF, then load the files normally in Image Calibration.  Seemed to clear the issue right up for me.

Good luck,

Jim 

[jkmorse]

Richard,

Have you been able to solve the problem you were having?  I am starting to suffer the same problem with select images and it is happening irrespective of my ImageCalibration settings.  Clearly it is not just a display issue because in my case PI will not run the ImageIntegration routine since the images are too degraded for the algorithm to work.  All I get is an error message and it stops.  I am not at my computer at the moment but will be later and can post examples to show what I am experiencing.  Interesting thing is that I took two separate sets of images over the weekend, a NB set of NGC 7380 on Thursday and a NB set of the Helix on Friday.  Using the same settings with both down the line, NGC 7380 is processing fine, but the Helix is suffering the same issues you described in your post. 

I tried a number of setting changes last night but had to give up since I was exhausted from the two prior nights of imaging.  Will try again tonight and report generally if the same issue continues.  Since it is happening to several of us, maybe there is a bug somewhere?  In case it matters, I am imaging using MaximDL and saving the images as 16bit fits files.

Thanks,

Jim

[MikeOates]

Jim,

Did you manage to trace the problem?

I also had a strange image after callibration and eventually tracked it down to an incorrect file being used, I think is was a flat. I forget now what was wrong with it, but it could have been a single non-linear flat file that I had looked at and accidently saved it with stf applied, or it was a completely wrong file like a dark or light that I had renamed incorrectly after taking but before processing. Anyway the result was like a solarized image.

So I suggest you go through every file to make sure they are all correct. Using blink can speed this up.

Mike

[jkmorse]

Mike,

Thanks for the suggestion.  I am currently on vacation but before i went, I shot new Dark and Bias frames and have all my original flats exposures to build new flats masters with.  Will re calibrate/integrate the whole set when I get back and expect that will solve all problems. 

Cheers,

Jim