We developed a novel deconvolution software that stunningly increases the visibility of minute details in light sheet microscopy and confocal microscopy recordings. Prior to deconvolution the software can optionally apply adaptive background subtraction by rolling ball filtering along the three principal axes . As demonstrated by the presented 3Dreconstructions this approach stunningly increases the level of detail in data sets depicting small structures (as e.g. dendrites or nerve fibers) located on a inhomogeneous, diffuse background. By automatic block-wise processing the Wiener Deconvolution Tools can process image stacks of virtually unlimited size even on small computers with only 8 or 16 GB RAM. Via consequent usage of parallelization as well as optional GPU-acceleration the software works with high speed: On a PC equipped with a state of the art NVIDIA graphic board a stack containing about 1 billion voxels can be deconvolved in less than 10 minutes. Our deconvolution algorithm uses flux-preserving regularization that does not modify the photogrammetry of the image data. To our knowledge, this regularization approach has not been used for deconvolving microscopy data before. Preservation of photogrammetry is especially helpful for performing quantitative comparisons of signal intensities between different images or regions. For non-commercial use a fully functional command line version of the Wiener Deconvolution Tools can be downloaded as free software. An additional graphical user interface (GUI) providing further features as batch processing, destriping of light sheet microscopy data by directional frequency filtering and basic 3D visualization is commercially available from the Technical University of Vienna.