2010PASP..122...63P

This paper presents the on-sky performance of the unmodulated infrared pyramid wavefront sensor PYRAMIR mounted on the ALFA adaptive optics system at the 3.5 m telescope of the Calar Alto Observatory. The performance of the system is compared with the performance of the Shack-Hartmann wavefront sensor of the ALFA system. We carried out a series of measurements to characterize the performance of PYRAMIR under open-loop tip-tilt compensation, and high-order closed-loop conditions, using stars of different magnitudes. We measured the tip-tilt jitter by following the centroid position of a stellar image on a fast series of frames. Additionally from the pyramid wavefront sensor data we could estimate the tip-tilt jitter in closed-loop. Under closed-loop conditions we also measured the long-exposure Strehl ratio. We compared the results of the wavefront sensor measurements with those of the Shack-Hartmann sensor on the same telescope, especially regarding the distribution of the error budged over the Karhunen-Loève modes, and the power spectral density. Our first finding is that we can indeed start up this nonmodulated pyramid system, even under bad seeing conditions. Under good conditions the Strehl ratio reaches in band. We found that the minimum signal-to-noise ratio (S/N) in each subaperture required to close the high-order loop is only 0.4. This is a surprisingly low number. To compare the performance to existing systems, we introduce the S/N per subaperture per loop cycle as a device-independent measure. Using this scheme, we find that the ratio between the low-order residuals and the high-order residuals in the case of PYRAMIR is lower than that of the Shack-Hartmann system, especially in the faint flux regime. This is an important finding because it means that the pyramid-based system removes the halo, i.e., light scattered by the atmosphere, around the target star better than a Shack-Hartmann sensor-based system. A comparison of the power spectral density of the PYRAMIR and the Shack-Hartmann sensor measurements, and a comparison of the noise propagation coefficients of PYRAMIR with theoretical predictions from the literature, confirm this superiority of the pyramid over the Shack-Hartmann sensor.