Comparison of calibration bias for five calibration methods using observations of MWC349. The measured-to-expected flux-density ratio is shown as a function of the atmospheric transmission for the Baseline method (first row), as well as for methods using the taumeter (second row) and skydip (third) opacity corrections, and for methods resorting to the PC-demo (fourth) and PC-point (fifth) photometric corrections. Dashed lines show the flux-density ratio rms errors.