Speaker
Description
Deterministic control of light propagation through complex scattering media has enabled applications including optical communications, metrology, nonlinear optics, and, most notably, multimode fibre microscopy and imaging. Central to these applications is the generation of high-fidelity optical fields at the distal facet of the fibre using wavefront shaping elements such as liquid crystal spatial light modulators (SLMs). SLMs have recently gained renewed interest owing to their higher-diffraction efficiency compared to digital micromirror devices.
In this work, we introduce and develop a CGH algorithm for sampled operators based on gradient descent optimisation using Wirtinger calculus. We present a systematic comparison of our algorithm with heuristic error-reduction methods such as the Yang-Gu algorithm. Using experimentally measured transmission matrices, we evaluate each approach’s ability to generate arbitrary complex optical fields through multimode fibres and compare their convergence and reconstruction fidelity.
We show that appropriate hologram optimisation improves projected-field fidelity by more than 10%, corresponding to an approximately 50% reduction in normalised mean-squared error relative to conventional phase conjugation. Analysis of TM eigenmodes provides new insight into the fundamental limits of field projection through fibre systems, while simulations show excellent agreement with experiment across all algorithms. We further demonstrate the equivalence of the Yang–Gu algorithm and gradient descent for TM-based complex field generation. These results establish optimisation-based CGH as an effective framework for high-fidelity wavefront shaping through multimode fibres, enabling advanced imaging modalities including stimulated emission depletion, and light-sheet microscopy, while also supporting arbitrary beam synthesis, such as natural images, for holographic display and projection systems.
| I am the presenting author | Yes |
|---|