this is correct, for calibration the images should be in 'cfa' format which is a mono greyscale image. if you zoom in on it you'll see it looking like a screen door or checkerboard pattern. you're looking at the raw sensor data.
after calibration of the lights, the next step is to run the Debayer process on the images, which will convert them from 1-channel mono to 3-channel RGB.
rob