Abstract
Purpose :
Many ocular wavefront sensors employ one or two telescopes to image the wavefront from an eye onto a Hartmann-Shack wavefront sensor (HSWS). Conventionally, the components are placed in a 4f setup, resulting in a rather long system, especially for an open field of view configuration. It would be useful to have a small and portable system that could be moved between different experimental setups. Therefore, this study tested the hypothesis that ocular wavefront sensors can be reduced in size by using a non-4f setup, while still maintaining the large eye-to-first-lens distance necessary for open-field sensing.
Methods :
The paraxial imaging properties of conventional 4f setups and non-4f setups were compared theoretically. The analyzed non-4f setups had the same eye-to-first-lens distance as the 4f setups, but with shorter focal lengths and with the eye and HSWS conjugate planes shifted away from the focal planes of the telescope lenses. Optical aberrations of the setups with different off-the-shelf lenses were also simulated in Zemax.
A wavefront sensor was built using one of the shortest developed non-4f designs. The HSWS was placed as close to the second lens of the telescope as possible. The correct position for the eye was found by utilizing a double eye model with refractive errors of +5 D and -5 D, which at the conjugate plane of the HSWS would yield equal detected pupil sizes on the HSWS. For verification of the system, trial lenses were then placed in front of the eye models and the wavefront was recorded as Zernike coefficients by the HSWS.
Results :
The paraxial calculations showed that the 4f and non-4f setups had identical imaging properties. Compared to a 4f setup, the maximum reduction in eye-to-HSWS distance was 50%, at which point the second-lens-to-HSWS distance was 0. At this limit, the first-lens-to-HSWS distance was reduced to 1/(M+2) of that of a 4f setup, where M is the angular magnification of the telescope. The Zemax simulations showed that the non-4f design could be made diffraction limited on-axis.
The measured defocus root-mean-square error for the eye models with trial lenses was 0.11 D, where the total defocus ranged from -5 D to +5 D in increments of 1 D.
Conclusions :
It is possible to construct a functional and compact ocular open-field wavefront sensor by using a non-4f setup. However, the initial alignment is more difficult, as the eye and the HSWS are not placed in the focal planes of the telescope lenses.
This is a 2021 ARVO Annual Meeting abstract.