Calibration and align batch operations

[mmirot]

I'll wait for the movie then.

Max 8)

[mmirot]

You might make another video on the blink module

[pfile]

bug report: 20 DSLR images, Ha filter. calibrated with master dark and master bias only, no flats.

1) is the default value of the fits direction flag in the script consistent with the pixinsight fits reader default direction? my master calibration frames gave me the dreaded "no correlation between dark and light frame" until i unchecked the fits direction box. all my masters had been previously prepared by hand and i don't think i've changed the fits reader settings.

2) the calibrated and debayered frames could not be registered by the script, however, if i put all the calibrated/debayered frames into StarAlignment manually, they were registered with no problem using the default StarAlignment parameters. i can only guess here that the script is calling StarAlignment with some different set of parameters, possibly the noise thresholds.

should i upload the CR2 files to the Wookie planet?

yes, please upload these files.

Kai

i uploaded my master bias, master dark, and 4 CR2 files to the "forum shared files" area, under the folder called "pfile_unsolvable_image". i wanted to change the name of the folder but it seems that you can't. anyway have a look at those files. i have tracking problems so please don't laugh at the star shapes

[georg.viehoever]

.... i have tracking problems so please don't laugh at the star shapes

I can assure you: You are not alone   

[Harry page]

Hi

Video as promised   http://www.harrysastroshed.com/Calibration.html

Harry

[kwiechen]

bug report: 20 DSLR images, Ha filter. calibrated with master dark and master bias only, no flats.

2) the calibrated and debayered frames could not be registered by the script, however, if i put all the calibrated/debayered frames into StarAlignment manually, they were registered with no problem using the default StarAlignment parameters. i can only guess here that the script is calling StarAlignment with some different set of parameters, possibly the noise thresholds.

I have put the master bias/dark and  the 4 light frames into the v1.0 script - no problem to calibrate and register. Can you please send an icon instance of the script with your settings?

Kai

[pfile]

very, very strange. i just tried to use the same calibration masters (together with a new master flat that i calibrated and integrated myself) on some lights that i captured last night.

now when i run, the script says that the master dark has no CFA pattern, and warns that this is contrary to the settings. it then tries to apply the dark and i only see one channel (K0) with a crazy scaling factor.

however, if i use the same calibration masters in ImageCalibration, the fits reader says it detected a CFA pattern in the dark.

i will try to reproduce the 1st problem and make an icon instance, as well as an icon instance for this 2nd problem.

[uli t]

Hi,

tried that new script w/o having darks. Does not work. I get the following error: no master calibration frames or overscan regions have been specified.

Does ist only work with darks?

Uli

[Fco. Bosch]

tried that new script w/o having darks. Does not work. I get the following error: no master calibration frames or overscan regions have been specified.

Does ist only work with darks?

Yes, nowadays it only works with darks. Mind that this is a script for "calibration", aligment and integration, and there isn't calibration without darks, that you can build at home with the time exposure and temperature that you wish or need, and to preserve in a library for future use. So the images will be better; this is the aim of the script.

[Harry page]

Hi

The script works without darks as long as you have a bias selected -- it will use this

Harry

[uli t]

Hi Harry,

thank you, works fine.

Yes, nowadays it only works with darks. Mind that this is a script for "calibration", aligment and integration, and there isn't calibration without darks, that you can build at home with the time exposure and temperature that you wish or need, and to preserve in a library for future use. So the images will be better; this is the aim of the script.

I am dithering, so I think, I don't need Darks. (DSLR EOS 450Da). At least, I try at the moment. There had been great discussions about that topic.

Maybe later I'll try with Darks to see the difference.
Thanks
Uli

[Harry page]

Hi

Juan are you going to give this script a push , would be well worth your time 

Harry

[cs_pixinsight]

I ran a BatchProprocessing run tonight and noticed this in the console output:

Loading master calibration frames:
Loading calibration frame image:
O:/fits/20120322_pi_master_gen/master/bias-BINNING_1.fit
Reading FITS: 32-bit floating point, 1 channel(s), 3906x2602 pixels: 100%
Loading calibration frame image:
O:/fits/20120424_rho_ophiuchi2/master/dark-BINNING_1-EXPTIME_180_55F_100subs.fit
** Ignoring 2 additional image(s) in master calibration frame.
Reading FITS: 32-bit floating point, 1 channel(s), 3906x2602 pixels: 100%
** Warning: The file format reports no CFA pattern for the master dark frame, but it is being forced as per process instance parameters.
Loading calibration frame image:
O:/fits/20120424_rho_ophiuchi2/master/flat-BINNING_1_35mm_f56_100subs.fit
** Ignoring 2 additional image(s) in master calibration frame.
Reading FITS: 32-bit floating point, 1 channel(s), 3906x2602 pixels: 100%

Notice the statement I bolded.  These subs are from a Canon 400D DSLR, so I checked the CFA Images box, but it appears the master dark is being generated as a monochrome image instead of RGB.  Is that by design or should the masters be generated as RAW RGB images instead for DSLR images?

Craig

[pfile]

yeah that's what was happening to me as well. i flaked out on the full bug report since i've been too busy with work.

[jeffweiss9]

    I tried the neat  CosmeticCorrection Module only, since I was in mid-course on some processing when I found this thread and watched Harry's new video, but I have the following question:
    How does one determine the "right" number of hot pixels to remove in operating the sliders in the Module?
    I thought that would be simply a matter of using the module to count the number of hot pixels in the reference MasterDark and adjust the slider by whatever means to get roughly that count.  However, all the choices I tried had me ending up using 2.5 to 3 sigma on "autodetect" since it seemed to make visible hot pixels in a sample imaging sub go away (like in the movie).  That must have been way too many false hot pixels, however, since when the RGB was later combined I ended up with a lot of random uni-color stars suggesting a lot of those 'hot pixels' I removed must have been, in fact, faint, smaller stars (unless I removed to few and suddenly got an additional batch of hot pixels showing up).   Any suggestions for a reliable way to gauge those settings, as I obviously didn't find one ? 
  I previously have used CosmeticCorrection SCRIPT with its default settings and then hand-edited; i.e. cloned, the several (5-20, say) residual hot pixels out of the L, R, G, B images before continuing by comparing them to a test RGB to spot the uni-colored stars ). The video gave me the idea that the MODULE would do a better job.  I know I really should be using dithering to reduce this problem in the first place (on deck to happen).
 Thanks very much.
-Jeff