The input image, this can include any number of channels and layers - if true Z defocus is required one of the channels should contain the Z depth information.
An optional input image that represents the shape of the defocus kernel. Although any image can be used it is most effective when reference stills have been created for the lens in question. This is usually achieved by rolling a few frames of a pin light shining directly into the lens in question.
An optional Nuke 3d camera that is used to override the focal plane, aperture etc. of the Bokeh node.
An optional Deep Data input, this can be just deep opacity values or full RGBA data. Although you’ve attached a deep input the plug-in still requires a main input, if you have just deep data it’s recommended that you use a DeepToImage and use the output of that.
An optional input that can be used to control various parameters with an image matte.
Used to select the desired channels/layers to apply the defocus effect too.
Used to select the desired Z channel, by default this is depth but any can be used.
Varying output types:
Used to limit the rendered output to all, front or back focus.
Used to control the scale of the defocus effect, multiplier is ignored when real world lens has been selected but both front and back are utilized.
Depth values are stored in a variety of ways, this lets you select how Bokeh interporates z values of the input image ( this is not relevant for deep inputs ).
The plane in focus.
Only available when real world lens is turned off, this expands the region in focus.
Controls how the effect falls off over the depth.
An optional matte channel ( supplied via the Mattes input ) to modulate the defocus amount
Used to normalize the output values when Visualize Depth is used
In some cases artifacts can arise from integrating all the slices together, this will separate the front and back focus areas and integrate them separate to help avoid these.
Turning on Real World Lens Simulation gives you control over the focal length, f-stop and other parameters required to match physical lens characteristics.
Focal length of the lens.
Aperture or f-stop of the lens.
The scale of the world units, you should set this to the scale provided by CG renders - Maya generally defaults to cm.
This is used as a multiplier on top of the world scale, as an example - if your scene units are based on 10cms then you can set the World Scale to centimeters and this parameter to 10.
Presents used to set the desired filmback, this will set the correct aperture width and height for the selected format. Custom can be used to enter your own values.
Automatically set by the Film Format selected above, it’s possible to override via Custom if a desired format isn’t available.
Will cause bright values in the input image ( defined by the bloom threshold ) to create a brighter lens bloom.
Used to define how much lens curvature there is, this will modulate the Bloom based on where in the lens the light is entering the lens.
Determines how bright the input must before before blooming is applied, by default this is 0.7 which means any input values above that will cause a blooming effect.
An optional matte channel ( supplied via the Mattes input ) to modulate the level of lens Bloom.
Control the spherical aberration of the lens.
Introduce chromic aberration artifact into the defocused image.
Used to adjust the refraction offset of each colour for the chromic abberation.
An optional matte channel ( supplied via the Mattes input ) to modulate the level of Chromic Aberration
The maximum size a kernel is allowed to be, this can be used to optimize renders if need be at the expense of clamping the defocus effect. It’s rare that you would actually need this.
Provides the means of controlling the aspect ratio of the Bokeh, this is useful for matching anamorphic effects
How soft the generated kernel will be.
The number of actual aperture blades, or “sides” of the shape.
Rotation of the polygonal kernel.
Artifacts are common in post processing effects that rely on depth due to the difference in fidelity and filtering between the Z depth and RGB images generated by most rendering applications. One of these artifacts is how a final image is composited due to varying depth values - as you can see in the image provided below it appears that the second sphere eats into the first.
Bokeh provides a means of correcting this by managing how the images depth values are composited - the volume that is sliced is defined by the focal plane and the focus region size.
By adjusting the Corrective Slices value it’s possible to remove most signs of the artifacts.
Be aware that due to the nature of this corrective stage the over all Bokeh level may need to be increased to return to the desired effect.
If selected the parameter below can be used to define the exact volume to be sliced vs. inferring it from the focal plane and region size.