ISOCAM Interactive Analysis User's Manual Version 5.0 - ISO - ESA

20.6. CVF ANALYSIS 245
1. Determine which of the EXPOSUREs of the BS PDS field .IMAGE are source pointings
and which are reference pointings. (The source and reference EXPOSURES are indexed
by .REF IMAGE and .SRC IMAGE respectively.)
2. The source and reference EXPOSURES are coadded (with reduce cube) takinginto
account .NPIX. The coadded reference image is subtracted from the coadded source image.
The result is the beam-switch MOSAIC.
3. .RASTER 1 is filled with the beam-switch MOSAIC. .NPIXRASTER is filled with the
total number of EXPOSURE pixels that are used to compute each pixel in .RASTER.
.RMSRASTER is filled with the RMS image of .RASTER.
20.6 CVF analysis
Dedicated CVF analysis is described in this section.
20.6.1 Sensitivity and straylight correction
Sensitivity correction, or more simply conversion from ADU to milli-janskys (mJy), can be performed
by dividing the EXPOSUREs in .IMAGE by the sensitivity correction factors stored in
.RESPONSE[i].SENSITIV. When a CVF PDS is initially created with get sscdcvf, the sensitivity
and straylight correction factors are taken from a CDS and placed in .RESPONSE.SENSITIV.
This is similar in principle to the way the CAL-G FLATs and DARKs are handled.
conv flux applies the correction and converts the EXPOSURE pixels to mJy.
CIA> conv_flux, cvf_pds, /image
The EXPOSURE pixels are now converted mJy. The CVF PDSfield.IMAGEUNIT is
updated to reflect this. Note that this operation is not reversible. To re-perform sensitivity
correction use reduce to recreate the EXPOSUREs and hence refill .IMAGE.
20.6.2 Photometry on faint point sources
For a CVF observation, the source positions moves, depending on the wavelength, slightly within
each CVF segment. For changes from one segment to another (there are two LW segments),
or a switch of the detector channel, the source position can jump several pixels. This behavior
makes photometry on CVFs more difficult. To overcome this, compute spectrum computes
for each wavelength (or CVF step) for a point source the best fitting of a PSF, and uses this
information to compute a shift-corrected flux spectrum.
After a the CVF observation has been fully reduced, and the approximate position of a point
source has been determined, e.g. by cvf display, cvf spectrum can be called:
CIA> compute_spectrum, cvf_pds, x_pos, y_pos, $
psf_dir="SAPI01$DKA200:[CIA.DATA.PSF]", $
sort_wave, flux, est_flux, est_flux_error
CIA> plot, est_wave, est_flux
1 .RASTER is not a likely name for a beam-switch MOSAIC. It was chosen early in the development of the
beam-switch analysis routines in order to make the BS PDS compatible with the raster PDS.