Real-time, dither-free Shack-Hartmann wavefront sensor optical gain measurement technique

Abstract

Past research in adaptive optics (AO) has demonstrated the link between apparent beacon extent and wavefront gradient estimation sensitivity, or optical gain, of a classical Shack-Hartmann (SH) subaperture when using quad-cell detector regions. Pixel diffusion and residual wavefront error broaden the effective subaperture point spread functions as the atmospheric seeing varies in time. Although the AO community has generally shifted toward resolved subapertures to combat these interlinked issues, the quad-cell subaperture design offers efficient light usage for dim beacons, integrating less pixel noise while also reducing sensor readout latency. Particularly for telescopes in poor seeing conditions, in order to reduce beacon magnitude requirements, a quad-cell SH design, coupled with the proposed algorithm, can be an enabling solution. We present research conducted at the Starfire Optical Range over the past 8 years in implementing a robust approach that measures the real-time sensitivity on the site’s natural guidestar and laser beacon AO systems at the 3.5- and 1.5-m telescopes. Emphasis is given toward the practical aspects that must be considered beyond the pure theory, which has been presented in several prior works. A high-signal-to-noise strategy has been implemented that estimates the aperture-averaged subaperture sensitivity (related to beacon size) by exploiting the null space of the least-squares wavefront reconstructor. Careful consideration has gone into the implementation of this estimation method to avoid unintended effects, particularly at low-light levels. Unfortunately, this solution does not in itself address aperture-variant effects, such as sodium beacon elongation for extremely large telescopes.