Main Page | Modules | Data Structures | File List | Data Fields | Globals | Related Pages

find_mpup

Analysis of micro-pupil frame

Parameters:
-in none Input micropupils frame
-nbin 100 Number of bins in histograms (relative)
-threshold 0.95 Frac of pixel population considered as mpups
-fracdist 0.1,0.1 1/2-sizes (x & y) of the analysis area around mpup [interlens]
-analysis NULL Output analysis table
-mask mask Output mask
-nggeo 3 Gaussian components in the geometric PSF (3|5|7)
-ngpsf 3 Gaussian components in the global PSF (>0)
-ngloc 1 Gaussian components in the local PSF (<=1)
  • If <0, global PSF is discarded, and Core PSF = Geometric PSF
  • If >0, Core PSF = Geometric PSF $\otimes$ Global PSF
-lambda none Reference wavelength of the input micropupils
-table NULL Input 1st-order mpup-position table
-offset NULL Manual offset subtracted to input coordinates
Returns:
Error status
Input table
Describe the use and the input columns (from ad-hoc analysis): :X1,:Y1 (FLOAT) The input columns were previously :XCEN,:YCEN (DOUBLE) from Midas.
Warning:
The simple $r^2$-distortion model used here do not fit properly the pseudo mpup position (structured discrepancy up to 4.5 px). Therefore, one cannot compute the "real undistorted" position, needed to compute the sky coordinates.

Todo:
Test the existence and use the variance extension in the preproc images for X-disp. profile fits (see USE_PROFERR). For the moment, analytic computation of the variance from photon noise and RoN.

Deprecated:
Since there's nothing close to a mpup exposure, find_mpup will soon disappear. The cross-dispersion profile analysis and the first estimate of the mask parameters will be done with prepare_mask.
Warning:
It's difficult to properly normalize the intensity of the mpup during the global fit, since it will depend on the strength of the line used for analysis.

Generated on Thu Oct 27 00:00:24 2005 for Snifs by doxygen 1.3.5