Present alignment methods already have an accuracy of some microns, allowing in general the fairly precise assembly of multi element optical systems. Nevertheless, they suffer decisive drawbacks, such as the necessity of an iterative process, stepping through all optical surfaces of the system when using autocollimation telescopes. In contrast to these limitations, the wavefront based alignment offers an elegant approach to potentially reach sub-µm accuracy in the alignment within a highly efficient process, that simultaneously acquires and evaluates the best optical solution possible. However, the practical use of these capabilities in corresponding alignment devices needs to take real sensor behavior into account. This publication will especially elaborate on the influence of the sensor properties in relation to the alignment process. The first dominant requirement is a highly stable measurement, since tiny perturbations in the optical system will have an also tiny influence on the wavefront. Secondly, the lateral sampling of the measured wavefront is supposed to be as high as possible, in order to be able to extract higher order Zernike coefficients reliable. The resulting necessity of using the largest sensor area possible conflicts with the requirement to allow a certain lateral displacement of the measured spot, indicating a perturbation. A movement of the sensor with suitable stages in turn leads to additional uncertainties connected to the actuators. Further factors include the SNR-ratio of the sensor as well as multiple measurements, in order to improve data repeatability. This publication will present a procedure of dealing with these relevant influence factors. Depending on the optical system and its properties the optimal adjustment of these parameters is derived.