Run the 
Deconvolution > 3D Deconvolution command to display the following window.
Figure 584.
Dialog window settings
When the Common Settings tab is selected, any changes made in the dialog are automatically applied to all image channels. To adjust settings for a specific channel, check the box next to its name (or select an already checked channel) and modify the values accordingly. Changes made within individual channel tabs are marked as “<per channel>” in the Common Settings tab. In most cases, the user sets common settings once, selects the deconvolution channels, and processes the image without configuring each channel separately. Deconvolution is applied only to the channels that are selected.
Select which deconvolution method is applied. See Deconvolution methods for further details.
To handle the out-of-focus planes correctly, it is important to know how exactly the image sequence has been acquired. Select the proper microscopic modality from the combo box.
If SFC/Slit mode is selected in the Modality field, define the slit orientation (horizontal, vertical).
Enter the refraction index of the immersion medium used. Predefined refraction indexes of different media can be selected from the pull-down menu.
Specify an estimation of the amount of noise present in the image.
Note
For the Richardson-Lucy deconvolution method a precise SNR (signal to noise ratio) value can be entered into the edit box when Custom noise level is selected. This value ranges from 1 (noisy image) to 50 (image almost without noise).
For the Richardson-Lucy, Landweber and Blind deconvolution methods an Automatic noise level recognition can be set.
Blind, Landweber and Richardson-Lucy deconvolution are iterative methods. More iterations (repetitions) take more computation time but usually improve quality of the result. Specify the number of iterations.
This function can be used to subtract the background value from the image. Deconvolution process is performed on the image without the background and once the process is finished, the background is placed back into the deconvolved image to its original place. By default the average background value is estimated automatically (Automatic). It can also be switched Off or defined manually (Manual) - click 
to define the background manually, the Background Picker dialog window appears (see: Background Picker).
If dark artifacts occur in the deconvolved image, it could be caused by single high intensity pixels which are sometimes produced by a confocal scanner. This option removes these pixels from the image before deconvolution.
If turned on, spherical aberration of the PSF is calculated and used. Enter the Acquisition Depth (depth defining the center of the sample where the PSF is generated). Default value of the Acquisition Depth is taken from image metadata as a distance from the Z-Stack edge to the Z-Stack Home position. This distance is in most cases 1/2 of the Z-Stack depth. The only exception is the Asymmetric Z-Stack mode where the home position is explicitly defined by the experiment setting.
Adjust the number of Layers if you want to use the depth variant deconvolution. If you set Layers to a number bigger than one, the stack is internally split in the Z dimension into several sub-stacks (layers) and an aberrated PSF is computed separately for each one. For example a Z stack having a total range of 10 µm, 5 layers, and acquisition depth (set in the deconvolution dialog) of 10 µm will generate PSFs for the sub-volumes at 6 µm, 8 µm, 10 µm, 12 µm, and 14 µm.
Finally set the Sample Ref. Index value (refractive index of the medium used for sample acquisition). Default Sample Refractive Index is 1.35 (for water).
If Use Spherical Aberration Correction is turned off, PSF without any aberration is used.
Clicking enables the user to import the Point Spread Function image. If this is done, the deconvolution parameters will be disabled.
Importing PSF can be of great help especially if you have determined the PSF image by capturing a nano-bead. It is reasonable to import the PSF for fast, blind and Richardson-Lucy deconvolution either. See Determining the Point Spread Function (PSF).
The current image is overwritten with the result of the deconvolution process.
This option enables the user to convert 3D deconvolved images into floating point image. When Create New Document is checked, this conversion is done automatically.
Deconvolution can be performed just on a portion of the image - region of interest. Check Run on ROI and define the ROI parameters. Click to draw the ROI directly into your image and confirm it by clicking the secondary mouse click. Or fill in the rectangle values manually into the edit boxes.
Check this item to see the changes and the result immediately in the image.
Note
When computing the preview image, deconvolution is applied to the current view. For example, having a mono timelapse ND file, the deconvolution is applied to the currently displayed frame. Total time needed for deconvolution of the whole file can be then estimated as Preview Time x Number of Time Frames. If you process a single-frame image, the preview time equals the deconvolution time.
Figure 585.
Dialog window settings
Represents the name of the channel.
Background value for each channel from the selection (Current Frame, All Frames). Use the buttons to gain the appropriate values.
Click to draw a rectangle working as a restriction area for getting the background values. Confirm it with the 
right mouse button.
Specifies how the background value of the ROI is calculated.
By clicking this button all the background values for all the channels available in the current image are automatically calculated and filled into the table.