We use a detailed modeling formalism based on numerical simulations of local calcium release events where the blurring of the image, the presence of diffusional barriers provided by large organelles situated close to the release site, as well as the variable position of the scan line with respect to the release site are taken into consideration. We have investigated the effect of the fluorescence noise fluctuations on the accuracy in computing the signal mass from linescan recordings and obtained a quantitative description of both the signal mass and the local increase in the free Ca(2+) level as a function of the release current, the release duration and the orientation of the scan line, for three different levels of noise magnitudes. The model could provide a very good fit to a wide set of available experimental data regarding the signal mass of puffs visualized by fluorescence microscopy in the Xenopus oocyte loaded with 40 μM Oregon Green-1 in the absence of the calcium chelator EGTA. Numerical simulations also predict the amplitude and the kinetics of calcium signals evolving in the absence of the indicator, and indicate that sub-maximal activation of IP(3) receptors could produce in average levels of about 2 μM and 0.4 μM free Ca(2+) close to a release site located in the animal or in the vegetal hemisphere, respectively, whereas the maximal levels reached in more rare events could be 11 μM and 4 μM, respectively.