| Version 2 (modified by , 15 years ago) ( diff ) |
|---|
Steps for masking
Generation of the automasks
First, process 100 i band 3pis using DEFAULT_DARKTEST as the reduction. Then run the following ppMerge:
ppMerge dark.XY02.mdc dark.XY02 -recipe PPMERGE PPMERGE_FLAT -Db VARIANCES F -Di ITER 3 -Df FRACLOW 0.0 -Df FRACHIGH 0.0 -D COMBINE MEDIAN -type flat'
dark.XY02.mdc looks like this:
rawImfile MULTI
rawImfile METADATA
IMAGE STR neb://ipp005.0/gpc1/darktest.20101213/o5367g0082o.184503/o5367g0082o.184503.ch.168800.XY02.fits
END
(but with all 100 images)
Then, when ppMerge finishes, run 'makemasks.pl' from ~heather/staticmask. This has various cuts based on the median stddev, sigma stddev and %images used, and uses a kernel to reject bad pixels that do not also have bad neighbors.
Combining the automask with cte, dark, flat masks
go.maskmerge merges together the cte,dark, flat, and automask for each chip.
Masking
same as before, use 'fixmask XY??' to mask a chip, 'fixmask XY?? xy??' to start masking at a certain cell.
Merging/culling cells that were masked twice
When masking, it is possible to exit out and restart a cell. This leads to duplicate cell entries in the static mask. 'fix.single.test XY??' will take the latest cell in a mask as the correct one, as well as find masks that are inconsistent.
Special Chips
While masking, we discovered that the old CTE mask was not valid for certain chips. Chris generated a new CTE mask, and I removed the old CTE mask and added the new CTE mask to the manual masks.
XY11 -- Because of CTE issues, had not been masked yet.
XY14 -- required the new dark as well. This is the only chip with the new dark (the old dark was masking the first and 3rd column)
Camera Changes
Changes (since the previous mask)
- 2010-02-28 OTA14 improves
- 2010-07-23 OTA11 improves (possibly before, this was the date it was noticed)
- 2010-09-12 "camera voltage change"
- 2010-09-27 bathtub rings visible in phase 2 devices
For this mask (2010-12-15) I chose a start date of 07-23-2010
ingesting the chips
see /data/ipp004.0/gpc1/detrend/gpcmask.20101215/register.mask.20101215.sh for the script
