BatchPreprocessing calibration not flattening light frames

[sarmen]

I think this may have started after the main PI upate for me. Normally I use BPP for calibration and integration of LRGB frames. This has worked wonderfully for a number of years. All of the sudden, BPP fails to flatten my light frames. I don't see any failure or warning messages flying by. BPP is creating master flat frames which look fine. The light frames are not being flattened. What is going on?

I use SGP for image capture. Looking at the fits headers, filter name for example would be 'Blue  ', type 'Light  '. Same goes for raw flat frames. I notice after flat calibration, the fits header for flats is 'Flat' with no trailing spaces. Wondering if this makes any difference.

I remember in the distant past some issue with Fits header text not matching and calibration just failing with error message. I would manually fix the header text for "Filter" I think.

When I manually calibrate the same light frames with Image cal process and the same master flat, it works just fine.

BPP was so nice and quick and performed good calibration for me. Now cannot use it! Any other PI - BPP users experiencing similar behavior?

If it may help, attached is the last log created by PI with final attempt with BPP using the master flats generated from previous BPP run, failed also.

Thanks in advance for any responses.

[andreasmax]

hi,
according to your log-file you are using masterfiles for BIAS, DARKS and FLATS.

when using the BPPscript DO NOT use masterdarks and masterflats for your stackings. it will end up not flattening your light frames and not correcting the dust donuts!

why?
the script will calibrate all these masterfiles (DARK, FLAT) a second time, therefore overcorrecting your lightframes! it took me several days and lots of postings to finally got the problem by my own.

when using the BPPscript just use the masterbias. instead of the other masterfiles use the original calibrating images and you will get your result.

give it a try. hope this helps.

cheers from austria,
andreas

[pfile]

you must have some other problem... or else someone fundamentally changed how BPP works recently. keep in mind you have to check "use master bias / use master dark / use master flat" in order to make it treat the masters you have loaded as masters and not re-calibrate them as though they were calibration subs.

rob

[sarmen]

Hello Andreas,

Thank you very much for responding. I believe I understand what you suggest, however I have been able to use master bias and master darks previously in BPP without problem. The reason in that log file master flat was shown was it got created in my first attempt on this data set w/BPP. The master flat was created and looked correct to me so I started trying to input that as a master in later attempts. The results were the same with the raw or master flats.

I will try again as you suggested though with only master bias, then raw darks and flats.

-Steve

[sarmen]

Hello Rob,

Thank you for responding. I do check the master (bias,dark,flat) button as needed in BPP when using masters. As mentioned, this was all working fine in the past. Either something changed with my FITs headers that is confusing BPP, or BPP or PI changed in someway to break this method.

One thing that would make BPP really high power would be capability to calibrate multiple nights worth of images in one shot. Ninety nine percent of the time I have multiple nights of data to work with. What I mean is separate flats per night with lights, and all use the same master dark and bias.

-Steve

[pfile]

so one thing with FITS is that there's nothing in the file which indicates what corner of the image that byte 0 corresponds to. if you look at the fits reader settings, you'll see that there are 2 origin settings - up-bottom and bottom-up - which accommodate the possible FITS origins.

it could be that you had it set one way (which matches how SGP is writing out the files) but during the update this setting got reset to default. BPP actually has a checkbox for this, so you can set it there on a run-by-run basis rather than changing it in the FITS file handler. you might try ticking that checkbox and see if anything is different.

rob