Pulse-shape dependent time averaging effects in interferometric nonlinear phase shift measurements

When optical pulses of temporal width less than or comparable to a detector's response time are used to interferometrically measure intensity-dependent phase shifts, it becomes necessary to use time-averaging protocols to deduce the material nonlinearity. Relations between the actual peak nonlinear phase shifts and the measured values are numerically evaluated for Gaussian and sech square pulse shapes. We believe that these results are also very important to understand the pulse break-up effect in Mach-Zehnder type all-optical switching applications.