Despite promising results, Curvature wavefront sensing is usually not considered as an option for future AO systems such as AO systems for Extremely Large Telescopes (ELTs) or high order systems for the current generation of 8 to 10m telescopes. CWFS is generally thought to be useful only for low order systems, both for technical reasons (detector and DM technology) and fundamental reasons (noise propagation for high order curvature systems). I show in this paper that these worries are unjustified, and that, thanks to newly developed techniques and algorithms, CWFS is in fact much superior to more traditional Shack-Hartman wavefront sensing: (1) CWFS can be made extremely efficient, even for a high order system, thanks to a new "multi-stroke" curvature wavefront sensing mode (2) CWFS-based systems can efficiently utilize both piezo-stack type deformable mirrors and square pixel detector array, and there is therefore no reason to think that technological considerations limit CWFS-based systems to low-order correction (3) non-linear Fourier-based CWFS control algorithms can dramatically increase the performance of existing and future CWFS-based systems.
展开▼