AUTHORS : A. Martel, G. Hartig, J. McCann.
The main goal is to measure the count rates in flat fields acquired with the internal T2 (WFC), T4 (HRC), and D2 (HRC) lamps through all the filters with no cross filters. These will serve as references for future calibration campaigns and for CEI verification.
The bias frames and flat fields were acquired as part of the dust filter search with the internal tungsten and deuterieum lamps (T2 for WFC and T4, D2 for HRC) and WFC Build#3 (amplifiers B and C, gain=1) and HRC Build#1 (amplifier C, gain=1). All the images were acquired with SMS procedures (JGCW27A-E for WFC and JGCH27F for HRC) and include all the filters, thus providing a wide range in wavelengths and exposure levels. The F892N filter and the polarizing filters were read out as 1/4-field subarrays and all others as full arrays. The F220W and F250W filters were imaged with a D2 lamp and HRC. Only the count rates of the flats obtained at the nominal filter positions (FWoffset1=0, FWoffset2=0) are considered.
The reduction is fully explained in the HRC and WFC Bias Subtraction and Overscan Analysis and we summarize it here. We first subtracted the bias frame from the individual flat fields and then removed any residual in the leading physical overscan (WFC) or trailing physical overscan (HRC). The minimum (MIN) and maximum (MAX) pixel values in the data regions were evaluated from the histogram of each image while the median (MED) was calculated with the 'msstat' task in IRAF after constructing pixel masks for each chip over the bad pixels/columns. The count rates are simply obtained by dividing the counts by the exposure time.
In Tables 1 and 2, the count rates are tabulated for each filter imaged with the WFC and HRC, respectively, after subtraction of the bias frames. For the WFC narrow-, medium-, and broad-band filters, we take the average values for Chips 1 and 2. The ramp filters were imaged at three consecutive wheel positions on the WFC to cover most of the surface of the three segments. Since each image shows great variations in intensity between each segment, average values between the two chips would be misleading, hence we list the MIN, MAX, and MED count rates of the full mosaiced WFC field at each wheel position.
The full well of the WFC Build#3 CCDs is >78000 e- (see WFC Builds) and its gain of 1 corresponds to 1.7 e-/DN, so saturation sets in around ~46000 DN. The polarization filters are very near this limit and in fact become non-linear in their central parts around 38000 DN. This is apparent in the ratioed images shown in the dust filter search, which are identically equal to one in the middle. The POL flat fields will have to be taken again with shorter exposure times for the same WFC detector.
The exposure time required to reach a median count of 20000 DN/pixel is indicated in the last column of each table. This value is a useful guide since no HRC or WFC detectors saturate at this level at gains of 1 or 2 and sufficient S/N ratios are reached.
Table 1 : WFC Build#3
Table 2 : HRC Build#1
CEI SPECIFICATIONS :
The count rates were determined for the different filters and internal lamps. Measurements for the polarization filters need to be repeated with non-saturated images.