`dcr.par` parameter variation for Red camera binning 2x2

Open simontorres opened this issue 6 years ago • 0 comments

Method

This has been done by hand changing some parameter and comparing with the initial results. If it improved this was set and another parameter was modified and the best was kept as reference.

Final set of parameters

- THRESH  = 3.0   // Threshold (in STDDEV)
+ THRESH  = 4.0   // Threshold (in STDDEV)
- XRAD    = 9     // x-radius of the box (size = 2 * radius)
+ XRAD    = 7     // x-radius of the box (size = 2 * radius)
YRAD    = 9     // y-radius of the box (size = 2 * radius)
NPASS   = 5     // Maximum number of cleaning passes
DIAXIS  = 0     // Dispersion axis: 0 - no dispersion, 1 - X, 2 - Y
LRAD    = 1     // Lower radius of region for replacement statistics
URAD    = 3     // Upper radius of region for replacement statistics
- GRAD    = 1     // Growing radius
+ GRAD    = 0     // Growing radius
VERBOSE = 1     // Verbose level [0,1,2]
END

Changing DIAXIS between 0 and 1 had no effect but definitely 2 is not the right one.