In the chapter 3, we showed that, under some given conditions, the correlation function of ONFS images can be used to derive the intensity of the light scattered by a sample. The calcultions were performed in the ideal case, in which the scattered light comes only from the sample. The presence of non ideal lenses and optical elements introduces an amount of undesired scattered light. This problem is common to every kind of scattering measurement; the undesired light, often referred to as stray light, is generally scattered at small angles.
In standard scattering measurements, the effect of the undesired light is additive. It can be subtracted, since the stray light can be measured by a blank measurement.
Dynamic scattering gives a way to distinguish the effect of the light scattered from elements that evolve with time from stationary ones. If the stray light comes from stationary elements, such as imperfections of the optical elements, its effect is to increase the correlation function, with no dependence on the delay. Thus the time dependent informations on the sample will be given by the bell shaped part of the correlation function, while the pedestal will contain informations on both the statically and dynamically scattered light.
If a blank measurement is possible, a more refined subtraction of the stray light becomes possible [18], provided that the stray light constitutes a speckle field, that is, the field is gaussian. Such a data processing can be extended to ONFS too. In the following sections, we will find a way to subtract the effect of the stray light, first considering an unlimited number of images, taken at different times, and then a finite set of images. Then, we will describe the whole data processing algorithm.