In OYSTER, the visibility amplitudes and phases of the program stars are calibrated using calibrator stars, for which those variables are modeled and the resulting model coefficients transferred to the program star data.
In practical terms, the software keeps four ``copies'' of the amplitudes (squared and triple), which correspond to the uncalibrated amplitudes, the uncalibrated but normalized amplitudes, the calibrated amplitudes, and the normalized and calibrated amplitudes. Normalization is based on a model of the star, currently a single disk with a diameter taken from a file compiled with a STARBASE procedure. When calibrating the data, the user should begin by looking at the calibrated and normalized amplitudes first to see whether the data has already been calibrated (for example when a .cha file was read). In addition, this data type is the only one which should approach unity for the calibrator stars after calibration. (In the case of visibility phases, there are no normalized quantities.)
When writing calibrated data into a .cha output file, both the uncalibrated and calibrated visibilities are written, but not the normalized data. This is because normalization is done every time the data is read from the output file. The calibration table, which can be used to undo specific calibration and which is updated during the calibration process is only needed for this purpose; any calibration can be reset without it and thus start over for a new calibration.