Investigation of adaptive laser pulse shaping by direct estimation of group delay profile

We investigate the benefits of estimating the group delay spectrum (as opposed to the phase spectrum) on the performance of an adaptive laser pulse shaping experiment. We use both numerical and experimental approaches, and consider different figures of merit such as efficiency, noise robustness, convergence speed, repeatability, fitness sensitivity, and parameter insensitivity. While primarily focused on the process of broadband second harmonic generation, we also study a resonant third order nonlinear process. Our results show an overall superiority of delay-based formulation in the cases under study (compared to a conventional phase-based formulation) for adaptive laser pulse shaping.