TESTBA/CCD

The command does a series of tests on a catalogue of bias frames. Since this commands produce the bias offset that is needed in most of the other test commands ( e.g. TESTF/CCD and TESTD/CCD), it should be the first command to be executed. The whole test is split in five smaller tests, commands TESTB1/CCD to TESTB5/CCD that do the following:

1. Test B1: Creation of the combined bias frame. The result is loaded onto the display.
2. Test B2: Find the hot pixels. The combined bias frame is median filtered (using the parameter `fil_siz') and subtracted from the original. A plot is generated showing the position of the hot pixels and the affected columns. Hot pixels will only be searched for within the requested area and above the intensity level of (mean + 0.25*sigma + 5.), where mean is the mean int level, simga is the standard deviation.
3. Test B3: Inspection of single frames. From the combined bias frame rows and columns are averaged and plotted.
4. Test B4: The last frame in the catalogue is first corrected for hot pixels and then rebinned. A histogram of this rebinned frame is made.
5. Test B5: For each input frame in the catalogue determine the mean bias and standard deviation after hot pixel correction (using the hot pixel table determined in test B2), box averaging and median filtering. The keyword BIASMEAN and BIASSIGM are filled with the average values for the mean and sigma.

To avoid unnecessary computations the command checks for the presence of the combined bias frame and the median filtered hot pixel frame and does not recompute these frames if they are already present. In the case of subtests the command will (re)created these output frames.

The complete TESTBA/CCD command produces the following:

• A combined bias frame.
• A map of hot pixels in bias frames (obtained from a median stack of the raw bias frames);
• An ASCII and a MIDAS table containing the hot pixels;
• Plots of row and column averages of the mean bias;
• The mean bias level and standard deviations after hot pixel correction median filtering.