<title>Adaptive optics on a shoe string</title>

Abstract

ABSTRACT There are two main ways to mitigate the effects of atmospheric turbulence on an imaging system. A postfacto approach, where data are opportunely acquired and processed in order to increase the overallresolution attainable by the optical system, speckle imaging is an example of such technique. The otherapproach is to use an adaptive optics system that will compensate for atmospheric effects before the dataare recorded. Of course, the situation is not sharply distinct. Hybrid approaches have been proposed anddemonstrated. Other approaches that are a mid-way between the two are also possible. The basic idea ofstatic and dynamic pupil masking will be presented. Experimental results based on point sources andextended objects will be presented. Advantages and liniitauon of such technique will be discussed. Finallysome new ideas involving fiber optics and liquid crystals will be presented.Keywords: pupil masking, liquid crystals, fiber optics. 1. INTRODUCTION The mitigation ofthe effects of atmospheric turbulence on imaging systems is a wide field of interest.Many authors have been involved in a variety of approaches demonstrating a wide range of capabilities.Even the difference between on-line and post-facto correction has been, somewhat, blurred by the use ofdeconvolution schemes on data coming from adaptive optics systems2. Even though, a growing body ofexperimental evidence is showing that the use of adaptive optics is superior, in fmal data quality, to a post-facto approach3 only. However, the bottom line criterion to establish what kind of approach is more suitablefor a specific application is, more often than not, economic considerations. In this framework is interestingto note that simpler, andmuch less expensive, techniques have not been carefully studied and very oftenhave been dismissed on basis like lack oftotal flux collection capabilities. While such capability may bevery important for applications like spectroscopy or photometry, it is more questionable for imaging.The idea of apodizing the entrance pupil of an optical system has been proposed by several authors4'5'6'7'8,and experimental evidence of such effects has been noted even by amateur astronomers a while ago7. Fewattempts to study the SNR characteristics of such systems have been made4. Even though the fewexperimental data show that the use of apodized pupils may generate near diffraction limited imagery theuse of such inexpensive techniques have found a large disinterest within the community. In order torekindle the interest in such technique we will go through some of the theoretical and experimental workthat has been accomplished so far.One ofthe major obstacle to overcome is the profound and deep rooted believe, especially in the visibleimaging community, that the monolithic aperture is always superior to an apodized one as far as photonflux and SNR characteristics are concerned. We claim, in a little provocative way, that most of this believeis ofpsychological and historical origin more than being substantiated by any real physical reason. Theradio community has been producing high fidelity imagery with sparsely sampled pupil for decades.All that we are going to describe and argue from now on will implicitly assume, if not explicitly stated,that the imaging system is affected by a random aberrator, like the Earth's atmosphere. However, some of