Calibration parameters

Which data are used to calculate the bias values is controlled by the calType parameter.

calType = 0     The bias values are read from a file calData.input which must be in the working directory. The file format is 1 row per channel, with the WB values first. The bias values are given in the order BX, BY, BN, Brn, k. A file in the correct format called calData.input can be produced by using the saveCalData parameter. Using this parameter will destroy any previous versions of the file.

calType = 1     The bias values are calculated from averages of all the DARK and BRIGHT data present in the archive data file. The entire file will be searched, not just the time range specified on the command line.

calType = 2     Bias values for BX, BY, BN, Brn are calculated from the average of each BACKGROUND instance and for k using an average of all the BRIGHT data. For each block, the set of bias values is chosen from the BACKGROUND data closest in time.

calType = 3      Bias values for BX, BY, BN, Brn are calculated from the average of each BACKGROUND instance and for k using an average of all the BRIGHT data. For each block, a set of 3 bias value sets is chosen based on closest time and each bias value is determined separately as the median value from those 3 sets.

calType = 4      As for calType = 3, except BX and BY are calculated from the FOREGROUND observations.

calType = 5     As for calType = 3, except BX, BY, Brn are calculated from the FOREGROUND observations. This method results in the removal of an estimated scintillation term from the data.

By default, for calType = 1 to 5 the entire input data file is searched for calibration data, regardless of the start and stop times specified. This loop through the data can take many minutes for data files lasting for several hours. If setCalTime is set to 1 (in config.params), the calibration processing will only use the data in the time range specified on the command line.

If no BRIGHT data are present, the values for k are set to the default value, which currently (May, 2003) is 0.26. This value can be changed using the DNPERELECTRON variable in vis_main.cc.

A memo describing how the default values of these parameters were determined using KI observations is available at http://msc.caltech.edu/KISupport/dataMemos/index.html.