Metrology of optical components (including 'classical' optics, x-ray optics and neutron optics) is usually performed in two steps: (1) Actual measurement process, resulting in a raw profile of the component under test; (2) Numerical processing of the above data, to obtain actual behavior of the optical beam. The first step can be performed in many ways: Slope profilometry (LTP), Height profilometry (including interferometry), Shack-Hartmann . . . In many cases, the desired accuracy poses a challenge to the experimentalist: Temperature, vibration, reference surfaces, etc. The second step presents a different challenge: In order to extract useful information from the data, this data should be processed according to the actual propagation law of the beam. We may illustrate this in a simplified manner: According to the wavelength of the incident beam, a given defect type may be considered 'roughness' (diffraction), or 'geometric.' In reality, the optical profile is made of an 'infinity' of lateral defect sizes. Also, the coherence area of the individual photon may not cover the whole optical surface. Therefore, one may not simply apply pure low amplitude diffraction, or pure geometrical processing to the raw data. Clearly, more complex numerical processing is required. In this paper, we use the properties of synchrotron X-ray beams to show the proper handling of X-ray mirror raw profile data. These properties include, in particular, the spatial coherence area of the individual photons. We suggest the use of a coherence function in the processing of raw data, to get even nearer the behavior of the optics in real use. This paper is (hopefully) the first of a series of papers on metrology aspects of x-ray mirrors.