3.5  Corrections for flat-fielding, burn-in and nonlinear effects, using VDS_CALIB

Details are given in CDS software note 51.

The flat field of the detector presents variations of only a few percent. However, before the spectra are divided by the appropriate parts of the flat field, the burn-in correction is added to the flat field. The VDS detector sensitivity decreases with time in the most strongly illuminated areas (i.e. in the centres of the brightest lines), because of local gain depression, a well-known effect in NIS (called burn-in). This has been monitored since launch using wide slit observations.

In fact, the burn-in of all the isolated lines can be accurately estimated from the images obtained when the instrument is operated in spectroheliogram mode with the 90x240 arc sec slit. Such images show a marked depression of the line intensity at the core of the line.

Figure 9: [Figure from Bill Thompson] A recent (July 2000) 'flat-field' exposure at the wavelength of HeI 584 Å. On the left is the raw data, showing the pre-loss and the current (to the right) burn in of that line. The profile of the image is overlayed - notice that the post-recovery burn in is already larger (at about 50%) than pre-loss. The image on the right shows the same data after the standard corrections have been applied.

The line profiles are reconstructed, adding to the flat field some gaussian `absorption' profiles with a predicted width, position, and normalized peak intensity. Table  3.5 shows some of the lines that need such a correction. Note how the correction changes with time, and how considerable it is. For example, the He I 584 Å line, the brightest in these detectors, had already in September 1997 a burn-in of 40%. This correction is accurate (to a few percent) only for the bright and un-blended lines.

Table 1: The burn-in factors for some lines, as applied by the standard CDS software, for two dates one year apart. Note the increase of the correction with time and that in 1997 the brightest line in NIS 2 already had a 40% in peak intensity depression. The wavelengths correspond to the predicted position of the lines for the observations on 1st Sept 1997 (there is a time-dependent wavelength calibration).

Line	Detector	1st Sept 1997	1st Sept 1996
Fe XVI (bl) 335.407 Å	NIS 1	0.141	0.095
Fe XI 341.117 Å	NIS 1	0.005	-
Si X 347.390 Å	NIS 1	0.080	-
Si X 356.023 Å	NIS 1	0.081	0.006
Fe XVI 360.763 Å	NIS 1	0.165	0.093
Fe XII 364.456 Å	NIS 1	0.096	0.051
Mg IX 368.086 Å	NIS 1	0.226	0.146
O III 525.863 Å	NIS 1	0.104	0.050
He I 537.108 Å	NIS 2	0.161	0.112
O IV 554.499 Å	NIS 2	0.236	0.146
He I 584.370 Å	NIS 2	0.419	0.321
O III 599.620 Å	NIS 2	0.124	0.082
Mg X (bl O IV) 609.818 Å	NIS 2	0.268	0.158
Mg X 624.990 Å	NIS 2	0.189	0.099
O V 629.819 Å	NIS 2	0.369	0.274

3.5.1  Other corrections applied by VDS_CALIB

The NIS1 readout electronics create a low-level fixed-pattern effect in the spectrum. Every fourth pixel appears to have a lower count than expected. The effect is corrected by the routine VDS_CALIB.

VDS_CALIB, after dividing the spectra by the flat field, corrects the exposure time and divides it into the data to give the ADC count rate. The correction to the exposure time is due to the electronic shuttering, which makes each exposure slightly longer (0.1165 seconds for the normal VDS operating voltage).

VDS_CALIB also applies a correction for non-linear effects at high count rates (ADC counts per photon, a function of the voltage). These corrections are large for high count rates.

Then, the VDS throughput is divided into the data to convert from DEBIASED-ADC count-rate to photon-events/pixel/s. The term ``photon-events'' refers to those EUV photons which actually interact with the detector.

For subsequent data analysis, it can be useful to have the spectra in photon-events/pixel and so the output from VDS_CALIB should be multiplied by the `corrected' exposure time.

Exposure durations:

The exposure durations are given in the parameter QLDS.HEADER.EXPTIME in the quick-look data structure. Strictly speaking, one should really multiply by the corrected exposure time, which is slightly longer than that given in the header. However, this correction is very small (0.1165 seconds for the normal VDS operating voltage), and can safely be ignored in most cases.

vds_calib, qlds, exptime_corr=exptime_corr,THROUGHPUT=THROUGHPUT

; QLDS now has units  PHOT-EVENTS/PIXEL/SECONDS
;if you want COUNTS/PIXEL/S  multiply for THROUGHPUT

;if you want PHOT-EVENTS/PIXEL multiply by exptime_corr