The purpose of the deconvolution is to digitally compensate for the blur introduced by the microscope. In 3D microscopy, deconvolution improves images on several points: - by increasing the resolution (along the axial direction in particular), - by reducing noise (especially at low flux), - by improving the contrast. This makes deconvolution a valuable tool for improving post-processing such as segmentation. This workshop proposes to demystify the deconvolution methods and offers a demonstration of open source deconvolution software. It will be in 4 parts: - a brief theoretical description, - the important points for a successful deconvolution, - comparison of classical methods with the DeconvolutionLab2 ImageJ plugin on simulated and real epifluorescence and confocal data. - in case the PSF is not known we will guide users in the use of EpiDEMIC, an Icy blind deconvolution plugin for epifluorescence.