Most types of microscope images could in principle be deconvolved. However, there are practical limitations, for example the image file sizes might be too large or imaging conditions might be dominated by effects other than blurring by the point spread function. If, for example, a sample has strong light scattering properties or if light is strongly absorbed by the sample, deconvolution becomes difficult or impossible.

Deconvolution works both in 2D as in 3D. The PSF is very small in 2D, so the improvements of deconvolving 2D images are usually not very significant. Its full power deconvolution can show when processing 3D image stacks which have been acquired according to the following general rules:

• Acquisition of images with enough pixel resolution by choosing objectives with numerical apertures >0.5 and using camera resolutions with small enough pixel sizes as recommended by the Nyquist criterion.
• Acquisition of Z-stacks with distance between individual planes not larger than recommended by the Nyquist criterion (2-fold oversampling of the theoretically resolvable information, Optimal button in the Z-stack tool).
• Acquisition of enough planes above and below the structure of interest. As a rule, acquiring about half the axial PSF size above and below is enough to also get restoration of the structures at the top and bottom of the structure of interest.
• Avoiding saturation of the detector.
• Choosing imaging conditions to avoid sample bleaching.
• Avoiding spherical aberrations by choosing objectives, which use an immersion medium with a refractive index as close as possible to the mounting medium of the sample (for example using water immersion objectives for cell cultures in aqueous medium).
• Choosing sample media with low background fluorescence (for example phenol red free culture media).

ZEN deconvolution is suitable for images from many different microscope types. The following list of image types have been tested and are supported by ZEN deconvolution: