ImageMerge

This node is a generic merge node that is able to perform all the other merge operations supported by Katana.

Connection Type

Connection Name

Function

Input

bg

The background image.

fg

The foreground image.

out_mask

An optional image to use as a mask. By default, the merge is limited to the non-black areas of the mask.

 

Control (UI)

Default Value

Function

[2D node controls]

N/A

Set the controls for the stereo view.

For more information, refer to the Common 2D Nodes Widget Type in Common Parameter Widgets.

ImageMerge parameters continued

operation

Merge

If you don't want layer the images together using the Merge compositing operation, select the operation to use instead.

The following conventions apply to the below operation descriptions:

F refers to the fg input.

f refers to the fg input's alpha channel.

B refers to the bg input.

b refers to the bg input's alpha channel.

operation (continued)

 

 

 

 

The available operations (based on the Porter & Duff paper) are:

Atop - Fb+B(1-f). This shows the shape of the background, with the foreground covering the background where the images overlap.

Average - (F+B)/2. This produces the average of the two images. The result is darker than the original images, but accentuates highlights.

Difference - abs(F-B). This shows how much the pixels differ and is useful for comparing two very similar images.

Divide - B/F. This divides the background values by the foreground values.

Exclusion - F+B-2FB. This is a more photographic form of Difference.

From - F-B. This subtracts the background from the foreground. For subtracting the foreground from the background instead, see Minus.

Geometric - 2FB/(F+B). This is another way of averaging two images. Visually, it's close to Min.

Hypot - sqrt(F*F+B*B). This resembles the Add and Screen operations. The result is not as bright as Add, but brighter than Screen. Hypot works with values above 1. It can be is useful for adding reflections, as an alternative to Screen.

In - Bf. This only shows the areas of the background that overlap with the alpha of the foreground. It can be useful for combining mattes.

Matte - Ff*B(1-f). This is a premultiplied Over. Use unpremultiplied images with this operation.

Max - max(F,B). This takes the maximum values of both images. This is a good way to combine mattes and useful for bringing aspects like bright hair detail through.

operation (continued)

 

 

 

Min - min(F,B). This takes the minimum values of both images.

Minus - B-F. This subtracts the foreground from the background. For subtracting the background from the foreground instead, see From.

Multiply - FB. This multiplies the values of the foreground by the values of the background. It can be used to composite darker values from the foreground with the background image - dark gray smoke shot against a white background, for example.

Out - B(1-f). This only shows the areas of the background that do not overlap with the alpha of the foreground. This can be useful for combining mattes.

Over - F+B(1-f). This layers the foreground over the background according to the alpha of the foreground. This is the most commonly used operation. It's used when layering a foreground element over a background plate.

Plus - F+B. This produces the sum of the foreground and background. Note that the add algorithm may result in pixel values higher than 1.0.

Screen - F+B-FB. This is similar to Hypot, but clamps pixel values to 1.0. This is mostly useful for combining mattes.

Under - F(1-b)+B. This is the reverse of the Over operation. It layers the background over the foreground according to the alpha of the background.

amount

1

Dissolves between the bg image at 0 and the full merge effect at 1.

displayWindow

Background

The frame size to output in the event that the fg and bg inputs are different sizes:

Background - output the frame size of the bg input.

Foreground - output the frame size of the fg input.

Union - output a combination of the bg and fg inputs' frame sizes.

Intersection - output an intersection of the bg and fg inputs' frame sizes. This restricts the output to the area where the two frame sizes overlap.

clampAlpha

enabled

When enabled, the output alpha channel is clamped to the 0-1 range. Color channels (RGB) are not affected.