Improved modulation transfer function evaluation method of a camera at any field point with a scanning tilted knife edge

Abstract

Modulating Transfer Function (MTF) has always been very important and useful for objectives quality definition and focal plane settings. This measurand provides the most relevant information on the optimized design and manufacturing of an optical system or the correct focus of a camera. MTF also gives out essential information on which defaults or aberrations appear on an optical objective, and so enables to diagnose potential design or manufacturing issues on a production line or R&D prototype. Test benches and algorithms have been defined and developed in order to satisfy the growing needs in optical objectives qualification as the latter become more and more critical in their accuracy and quality specification. Many methods are used to evaluate the Modulating Transfer Function. Slit imaging and scanning on a camera, MTF evaluation thanks to wavefront measurement or imaging fixed slanted knife edge on the detector of the camera. All these methods have pros and cons, some lack in resolution, accuracy or don’t enable to compare simulated MTF curves with real measured data. These methods are firstly reminded in this paper. HGH has recently developed an improved and mixed version of a scanning technique used on a slanted knife edge giving a more accurate, ergonomic, high resolution and precise Line Spread Function (LSF) and one axis MTF measurement of a camera. A selected single pixel corresponding to a precise field point of the camera is scanned with sub pixelic resolution by the tilted knife edge thus enabling an optimized accuracy for LSF and MTF curves. The experimental protocol which requires a high-performance collimator, a scanning wheel device and a camera set up is detailed in this paper. Explained simulations are done to prove the under 1% accuracy of this method with regards to the different characteristics of the camera. All the parameters of this improved measurement technique are described and their effect criticized to give out all the result influence of these variables. These simulations and the algorithms used are then confronted to real measurements on a camera thanks to a mirror-based collimator and a scanning wheel device equipped with a slanted knife edge target.