How to Calibrate Your CCD in the Dark? a Precise CCD Calibration Method

Abstract

Single particle tracking and super-resolution microscopy experiments require good knowledge of CCD pixels’ gains in terms of photoelectrons per ADU (Analog to Digital Unit, output of the CCD camera) registered by the read-out electronics of the CCD camera. This is because the observed CCD output in a given time period produced by light with constant intensity follows, ignoring other noise sources, statistics of photoelectrons produced by incident photons. It is common practice to calibrate the gain of a CCD camera using a constant intensity light source to get the gain in photoelectron per ADU. This procedure requires a stable light source and it is not always possible to achieve a truly uniform and stable illumination. We propose an easier calibration method that uses dark noise as the calibration source for a CCD camera; the method requires that the dark noise level is stable during the calibration data taking. The proposed method works as demonstrated under assumptions that statistics of photoelectrons and thermal electrons follow Poisson distribution and CCD pixel gains in electrons per ADU are the same for the both sources of electrons. We will show that these assumptions are valid and will demonstrate that it is possible to calibrate individual gain of each pixel at ∼1% level or better using real and simulated data. As a byproduct, the proposed calibration method provides properties of individual pixel, not only its gain but also other properties such as linearity in low photon intensity and dark noise level. This information gives assessment of quality of the CCD camera. All the necessary software is developed using freely available software package Octave similar to Matlab. This method might improve the localization of fluorescent molecules in super-resolution and particle tracking measurements.