one more question. does ImageCalibration only work if the output format is 32-bit float?
i have been experimenting with this and even if i convert my master flats/bias/darks to 16 bit integer, i can't get the output images to look right unless the master frames are 32-bit float and the output is set to 32-bit float.
