Next: TESTD/CCD
Up: Test Commands
Previous: TESTFA/CCD
The command does a series of tests on a catalogue of flat frames.
The flat fields in the catalogue should be grouped in pairs with
the same exposure time. Most ideally, one should be of two groups of
the order of 8 frames each - the first with increasing exposure
times and the second with decreasing exposure times, interleaved
with those of the first group. In this way, trends observed in the
CCD response that are probably caused by the effect of temperature
variations on the light source can be rejected.
The command requires a value for the mean bias level and the
standard deviation in the keywords BIASMEAN and BIASSIGM
to be filled and hence should be executed after the command
TESTB/CCD. If no value or the value zero is found no bias
offset will be subtracted.
The whole test is split in three smaller tests, commands TESTT1/CCD
to TESTT3/CCD that do the following:
- 1.
- Test T1: Creation of the transfer/linearity table.
The table will contain 5 columns: column 1 for the exposure
time of the first of each sets (frames 1) (label :Exp_tim1);
column 2 the exposure time of the second frames (frames 2)
(:Exp_tim2); column 3 the median pixel intensities over the
selected frame sections in frames 1 (:Med_cnt1); column 4
the median pixel intensities over the selected area in
frames 2 (:Med_cnt2) ; column 5 the variance of the
difference of the frames 1 and 2 (:Variance).
- 2.
- Test T2: Determination of linearity curve and the shutter error and
the shutter offset.
Entries in the linearity table not fullfilling the selection
criteria will now be selected out. From the remaining
entries in the table a linear fit is done to determine the
linearity curve for frames 1 and 2 and the shutter error.
Using the linearity data the fractional count rates are
plotted against the median counts, applying a shutter offset
in the measured exposure times. The real shutter offset is
determined by the value for which the fit give the minimum
mean residual.
- 3.
- Test T3: Determination of the transfer curve.
From the selected entries of the table a linear regression
analysis is done to determine the analog to digital
conversion factor and the electronic readout noise. The
readout noise is determined by the inverse of the slope
between the median and the variance multiplied by the sigma
of the bias (determined by TESTBA/CCD OR TESTB5/CCD and
stored in keyword BIASSIGM).
The complete command produces:
- A table containing the exposure time of the first of each
sets (frames 1); the exposure time of the second frames
(frames 2); the median pixel intensities over the selected
frame sections in frames 1; the median pixel intensities
over the selected area in frames 2; the variance of the
difference of the frames 1 and 2
- Two linearity curves, expressed as count rate versus true
exposure time. The mechanical shutter delay is determined
either by linear extrapolation of the normal linearity curve
(observed counts versus exposure time), thus assuming the
response of the CCD is linear, or by adjusting the exposure
times such that the count rate curve is closest to a
straight line, thus allowing for a first-order nonlinearity
in the response of the CCD.
- A transfer curve (Janesick et al., 1987) generated
for any window onto the images obtained.
The linearity and the transfer curves may be generated for
any section of the images.
Next: TESTD/CCD
Up: Test Commands
Previous: TESTFA/CCD
Petra Nass
1999-06-15