Phase retrieval in the focal plane

Phase retrieval in the focal plane is a very appealing technique, which would simplify the optomechanics of an Adaptive Optics (AO) system a lot and could gain sensitivity under certain conditions. For conventional AO systems the limiting magnitude of the system does not depend on the diameter of the telescope, since any wave front sensor splits the light into sub-apertures, which are in number related to the telescope diameter. Having this in mind the phase retrieval technique looks promising as it breaks this paradigm in the diffraction limited case and thus yields some gain in limiting magnitude with larger telescope diameter. Until now this path was not followed deeply in astronomical AO systems, as the solution of the inversion is non unique and demands much higher calculation power as in conventional AO. This might change with state of the art computers. We give a short overview of some existing techniques and algorithms of focal plane AO and report results of other groups, which tested them in laboratory and on sky. To solve the drawback of the large computational demands and to increase the sensitivity we propose a bootstrapping process with dynamical binning