CCD Transfer Curves and Optimization

doi:10.1117/3.374903.ch2

Book: Scientific Charge-Coupled Devices

Author(s): James R. Janesick

Published: 2001

https://doi.org/10.1117/3.374903.ch2

Author Affiliations +

Abstract

Traditionally, CCD performance has been measured using commercial industry standards, which usually fall short of the requirements for many scientific digital applications. Also, scientific and commercial communities have been at odds with each other because each group uses different performance standards and units. This situation has created confusion as groups attempt to compare performance and data products. Commercial performance is often based on relative and photometric units (volts, lux, foot-candles, lumens, etc.), making it very difficult to know if absolute performance is being achieved. In contrast, scientific CCD camera systems are usually predicated on calibration standards that are based on absolute and physical radiometric units (photons, electrons, amps, watts, ergs, etc.). For example, it is physically meaningless to quote a noise or full well specification in relative units such as volts. However, these specifications become completely meaningful if specified in absolute units of electrons. In characterizing the capabilities of the CCD for scientific work, numerous absolute test tools have been developed that allow performance to be expressed in absolute units. Test results from these techniques are often presented in the form of a transfer curve. For example, the spectral sensitivity for the CCD is graphed as absolute sensitivity as a function of wavelength, an important transfer curve called QE transfer. Different transfer curves describe different CCD and camera performance parameters measured under various test conditions. Transfer curves have been organized into three different working categories: (1) CCD performance, (2) CCD camera performance and (3) CCD camera calibration.