Mozilla Skin*/?>

Conduit Node Reference

From ConduitWiki

Contents

1D Blur

The 1D Blur node performs a directional blur on the input image.

Parameters
Blur Amount controls the intensity of the effect. The size of the convolution kernal is related to the Blur Amount as follows. Kernal_width = (Blur Amount * 2) + 1.
Angle controls the direction of the blur. A value of 0 corresponds to a horizontal blur. A value of 90 corresponds to a vertical blur.
Example
1D Blur may be useful for simulating motion blur. It may also be useful when working with source material containing scaled color information (eg. 4:2:2 video). In that case, it's sometimes useful to temporarily convert the source from RGB to YUV (see RGB To YUV). Applying a horizontal blur to the U and V channels will soften the color information in the edges. Then, convert back to the RGB colorspace (see YUV To RGB).

Category: Image

2D MinMax

The 2D MinMax node expands or contracts the bright areas of the input image.

Parameters
Size controls the intensity of the effect.
Angle determines the direction in which the convolution kernel is calculated across the image.
Vertical aspect controls the ratio of the kernel's height to width. Values less than 1 yield a kernel that is wider than it is tall. Values greater than 1 yield a kernel that is taller than it is wide.
Example
2D MinMax is useful for creating procedural garbage mattes by expanding a high-contrast version of the raw, keyed matte.

Category: Image

2D Transform

The 2D Transform node translates and scales the input image. If you also need to rotate the image, use the Place Over node.

Parameters
X translation moves the input horizontally.
Y translation moves the input vertically.
X scale scales the input horizontally.
Y scale scales the input vertically.

Category: Image

3D Key

The 3D Key node creates an alpha channel for the input image, using the 3-dimensional distance (where RGB corresponds to XYZ) between input color and the input image color for each pixel.

Parameters
View allows you to choose between viewing the composite or the matte created by the operation.
Input black adjusts the input's black point.
Input white adjusts the input's white point.
Example
3D Key is useful for removing backgrounds in greenscreen or bluescreen footage.

Category: Composite

3D Transform

The 2D Transform node places the image onto a plane in 3D space. The plane can be translated, scaled and rotated.

Parameters
Antialiasing smooths the edges of the plane.
X translation moves the input horizontally.
Y translation moves the input vertically.
Z translation moves the input parallel to the camera axis.
X rotation rotates the input around the X axis.
Y rotation rotates the input around the Y axis.
Z rotation rotates the input around the Z axis.
X scale scales the input horizontally.
Y scale scales the input vertically.
Camera FOV determines the field of view of the camera. A low value is similar to a telephoto lens, while a high value is like a wide angle lens.

Absolute Value

The Absolute Value node calculates the absolute value of the input.

Parameters
Effect opacity varies the effect between 0% and 100%.

Category: Image

Add

The Add node calculates the sum of the inputs.

Parameters
Effect opacity varies the effect between 0% and 100%.

Category: Math

Bezier Curve

The Bezier Curve node modifies the input color according to a user-editable Bezier curve. This is a simple version of the Curves (RGBA) node and operates effects all channels.

The curve has four control points which determine how the input values on the X axis are mapped to the output values on the Y axis. For example, dragging the leftmost control point upwards will increase the brightness of the image shadows, and making an "S" shaped curve by modifying the two middle control points is a convenient way of modifying image contrast.

By default, this node clips the input values to stay within the 0-1 range, because the Bezier computation only works properly within that range. If you need to use this node on an HDR image, the recommended approach is to first scale your values to within the 0-1 range using a Levels node then feed the result to a Bezier Curve node. After the Bezier Curve node, scale the values back to the original range with another Levels node.

If you require more precise control points or control over individual channels, use the Curves (RGBA) node.

'Curve' in this context refers to adjusting the tone curve of an image. For drawing graphical curves, use the Shapes node.

Category: Image

Blend

The Blend node fades between two inputs.

Parameters
Bias adjusts the amount of each image that's passed to the output.
Effect opacity varies the effect between 0% and 100%.
Example
If Bias is set to 0.5, the result will be the average of the two inputs.

Category: Math

Canvas - scriptable

The Canvas - scriptable node renders a graphic image created by a JavaScript program. This node can be used to create almost any 2D graphics.

Tools created with this node can be packaged as custom plugins that appear like regular nodes in Conduit. To package a tool made using the Canvas node as a plugin, open the Scripts tab, click on the File button and choose "Save as Plugin in Conduit" from the menu.

Category: Special

Tutorial available: Canvas Tutorial

Channel Mixer

The Channel Mixer node mixes the input's RGB channel values according to the proportions given by the user-editable 3 x 3 matrix. This is a useful generic tool for many kinds of color effects. Mathematically, this operation performs a matrix multiplication.

The "Reset to identity" button returns the matrix to the default state where it has no effect on the input, i.e. each channel is given its original value.

When the "inverse transform" parameter is enabled, the inverse of the given matrix is computed. This can be useful for specific situations where you need to apply a linear transform and then reverse it eventually; e.g. transforming a color into a custom color space with a forward Channel Mixer operation, and then back into the original color space with an inverse Channel Mixer operation. Not all matrices are invertible, however. (You can use the Plot window as an aid in determining if the inverse Channel Mixer operation returns the original value.)

Category: Channels

Cineon to Linear

The Cineon to Linear node takes a Cineon colorspace input, where 0-1023 has been mapped to 0-1, and converts it to linear. The "black point" and "white point" parameters control the conversion range. The other parameters match those found in the application Shake which is commonly used to process Cineon images.

Category: Colorspace

Clip

The Clip node constrains the input value to within the selected range.

Parameters
Clip below sets the lower limit of the output.
Clip above sets the upper limit of the output.
Effect opacity varies the effect between 0% and 100%.

Category: Math

Color

The Color node produces an image consisting of a single, user-selected color.

Parameters
Red defines the value of the red channel.
Green defines the value of the green channel.
Blue defines the value of the blue channel.
Alpha defines the value of the alpha channel.

Category: Inputs

Color Difference Key

The Color Difference Key node performs a color difference keying operation on the input image by using the differences between the input channels to create a matte.

Parameters
View allows you to select between viewing the composite or the matte created by the operation.
Screen color defines the background color to be either blue or green. :Subtract chooses from the available difference algorithms.
Input black adjusts the input's black point.
Input white adjusts the input's white point.

Category: Composite

Color Picker

The Color Picker node brings in a color value selected by the host application's color picker. The Conduit plug-in provides up to four color pickers.

Select a color from the screen using the host's color picker and then bring the selected color into your composite using the Color Picker node.

Parameters
Host Picker chooses which of the plug-in's color picker inputs will be used by the node.

Category: Inputs

Color Range Key

The Color Range Key node performs a color range key operation on the input image. The user can define the range of values for each channel that will be subtracted from the input to create the matte.

Parameters
View allows you to select between viewing the composite or the matte created by the operation.
R input black adjusts the black point of the red channel.
R input white adjusts the white point of the red channel.
G input black adjusts the black point of the green channel.
G input white adjusts the white point of the green channel.
B input black adjusts the black point of the blue channel.
B input white adjusts the white point of the blue channel.

Category: Composite

Combine RGBA

The Combine RGBA combines four scalar values (red, green, blue and alpha) into an output vector to produce an RGBA image.

Category: Channels

Convert RGB Space

The Convert RGB Space node converts an RGB image from one color gamut to another. More precisely, this node converts between the RGB primaries that define the legal range of colors for a particular device.

Parameters
Source RGB space selects among five of the most commonly used RGB primaries: HD / sRGB (Rec.709), NTSC digital video (Rec.601), PAL (EBU 3213), ProPhoto RGB (common in high-end photography) and Adobe RGB 1998 (common in many graphic design workflows).
Source white point selects the source white point.
Output RGB space selects among five RGB primaries: HD / sRGB (Rec.709), NTSC digital video (Rec.601), PAL (EBU 3213), ProPhoto RGBand Adobe RGB 1998.
Output white point selects the output white point.
Example
It is common to have a HD image accidentally encoded with the standard-definition (NTSC or PAL) primaries, or vice versa. This node offers an easy way to repair this problem.
This node can also be used for converting between standard white points (D40 is the warmest and D70 the coolest).
For proper operation, the RGB image needs to be converted to linear color space first. Use the Video to Linear node to perform the conversion.

Category: Colorspace

Convolution (3x3)

The Convolution (3x3) node performs a convolution operation on the input image. Each pixel value is computed by combining the neighboring pixel values according to weight values, also known as the 'convolution kernel'. This node can be used for special sharpen, blur, or other effects that make use of nearby pixel values.

The group of 9 parameter values represents the 3 x 3 convolution kernel that is evaluated for each pixel in the input image. By default, the weights in the kernel are set to “identity”, meaning that the value of the output is exactly that of the input (i.e. a weight of 1 for the pixel being evaluated, and no weight for the surrounding pixels).

Modifying the weights of the parameters around the middle box will add the values of those pixels to the output for the evaluation pixel, based on the weights.

Note: The red, green and blue labels for the parameter rows are necessary for display of the parameter widgets and don't correspond to any per-channel convolution operations. The convolution works on all channels of the input image.

Category: Image

Crop

The Crop node crops the input image from any or all of the four sides (left, right, top, bottom). The cropped area is filled with transparent pixels.

Category: Image

Cubic Curve

The Cubic Curve node is basically the same as Bezier Curve, but is significantly less expensive to compute. In exchange, the curve is not as editable as a true Bezier curve. There are two parameters you can enable to make the curve more editable.

Parameters
Enable X scaling lets you move the start and end points of the curve in the X direction.
Enable X weighting lets you move the middle control points in the X direction.

The same caveat about input values outside the 0-1 range apply as with the Bezier Curve node.

'Curve' in this context refers to adjusting the tone curve of an image. For drawing graphical curves, use the Shapes node.

Category: Image

Curves (RGBA)

The Curves (RGBA) node is a powerful tool for adjusting the tonal range of an input image. It is equivalent to the Curves adjustment tool found in image editing programs. Each color channel can be adjusted separately. The number of points you may add to the curve is unlimited, for the highest precision.

The input and output values can be HDR, i.e. you can use this tool to create pixel values outside of the 0-1 standard range.

If you only need to do a simple tone curve adjustment to the whole image, consider using the Bezier Curve node instead because it renders faster.

'Curve' in this context refers to adjusting the tone curve of an image. For drawing graphical curves, use the Shapes node.

Category: Image

Darken/Min

The Darken/Min node compares two images on a channel by channel and pixel by pixel basis and outputs the pixel with the lowest value. This is equivalent to the "Darken" blending mode in many image editing programs.

Category: Math

Despill

Removes blue or green spill (reflected light from the green or bluescreen) on the foreground object. The amount parameter controls how much the despilled input is blended with the unaltered input.

Category: Composite

Displace / Sample

Moves the pixels in the first input image according to the pixel values of the second image. This node can be used for any kind of distortion effects.

The second image can be either a displacement map or a coordinate map. The "Mode" parameter is used to select the type of the input.

The Grid/Coordinates node can be used to produce coordinates needed to act as a coordinate map.

Category: Image

Distance 3D

Computes the distance between input 1 and input 2 in 3D space. The input color value is treated as 3-dimensional coordinates where RGB = XYZ and alpha is ignored. The output value is a scalar.

This operation is useful for driving a value based on the distance between two colors.

Category: Math

Divide

Divides the input value by the parameter value.

Dividing by zero is an invalid operation, and yields different results on different hardware: generally speaking, NVidia GPUs as well as general-purpose CPUs will return a special "not-a-number" value, whereas ATI GPUs simply return a very large number.

This difference may become significant if you use the result for any further calculations, because "not-a-number" is a meaningless value on which you can't perform any regular math operations. (When the image is finally displayed, any "not-a-number" values will be treated as very large numbers, i.e. they become white.)

Category: Math

Dot Product 3D

Computes the angle between two input vectors in 3D space. As with the Distance 3D node, the input color value is treated as 3-dimensional coordinates where RGB = XYZ (i.e. Cartesian X/Y/Z coordinates), and alpha is ignored. The output value is a scalar.

This operation is commonly used for lighting-related calculations. When working with synthetic 3D images, many rendering applications can be configured to output a "normal pass": instead of final lighting and texturing, the 3D surface normals are encoded in the image. Using the normal pass, you can perform post-production lighting of 3D models in Conduit. (This is sometimes called deferred rendering).

Category: Math

Embedded Image

Outputs an image loaded from an image file on disk. The image is stored as part of the conduit, so it is independent of the original image file once it's been loaded into this node.

Category: Inputs

Exposure

Modifies the image's brightness in a way that is similar to changing photographic exposure. The input image needs to be in linear color space first (use the Video to Linear node for this).

Category: Image

Gamma

Modifies the image's gamma. For full HDR gamma including negative values, disable the "Clip below 0" parameter.

Category: Image

Gaussian Blur

Performs a Gaussian blur operation on the input image.

Category: Image

Gradient

Renders a linear gradient. By default, the gradient consists of three colors. If you want a two-point gradient only, set "Position (color 2)" to one.

The gradient colors are determined by the three color inputs; thus the gradient doesn't have to be constant over the entire image, it can vary per pixel. The "horizontal" and "reverse" parameters determine the direction of the gradient.

By default, the gradient is applied according to the pixel's vertical or horizontal position within the output image. But this node can also be used to map arbitrary values into the gradient. To do this, connect a value to the Gradient node's "bias" input. A basic use for this functionality would be to create a two- or three-tone image by connecting a Luminance node to the bias input.

Category: Image

Grid / Coordinates

Renders a pattern image that is either a checkerboard-style grid of colors or a coordinate map. The "Tile style" parameter is used to select the type of the pattern.

Category: Image

Highlight Knee

Reduces the highlights in the input image. The amount parameter determines the strength of the effect. It can be useful to have the Plot window open and set to Plot View to see a visual representation of the “knee” curve.

Category: Image

HSV to RGB

Converts an input image (assumed to be in HSV colorspace, i.e. Hue / Saturation / Value) and converts it to RGB colorspace.

Category: Colorspace

Image

Brings one of the plug-in's image inputs into the composite. The image input selection parameter designates which of the eight available inputs should be connected to the node.

Category: Inputs

Inside

Composites input 1 over input 2, only outputting values within the alpha channel for input 2. The “Input 1 is premultiplied” parameter determines whether or not the node should unpremultiply input 1 before performing the composite.

Category: Composite

Invert

Inverts the image, i.e. produces a negative image. This is equivalent to subtracting the input from 1.0.

Category: Image

L*a*b* to XYZ

Takes a CIE L*a*b* input, and converts it to the CIE XYZ colorspace. Requires the reference illuminant of the Lab input (which is typically D50).

Note: XYZ spaces in Conduit always use D65 as the reference illuminant for user convenience (this matches typical video and digital images).

Category: Colorspace

Levels

Lets you modify the input and output ranges as well as the gamma of the input color. This is equivalent to the "Levels" tool in image editing programs.

The input black/white parameters determine the value range within the image that will be scaled into the output black/white range. Levels is HDR except for the "Gamma" control, which clips the input to 0-1. If an HDR gamma is required, use the "Gamma" node.

Category: Image

Lighten / Max

For each color channel, sets the output channel to the greatest of the two inputs. This is equivalent to the "Lighten" blending mode in many image editing programs.

Category: Math

Linear to Cineon

Takes a Linear input and converts it to Cineon, where 0-1023 has been mapped to 0-1. The "black point" and "white point" parameters control the conversion range. The other parameters match those found in the application Shake which is commonly used to process Cineon images.

Category: Colorspace

Linear to Video

Takes a Linear input (i.e. an image in linear light colorspace) and applies a gamma for proper viewing on a monitor or for outputting to a video file. For most video formats, the gamma parameter should stay at the default value of 2.2. For viewing and monitoring on a Macintosh system older than 2009, this may need to be changed to 1.8. (The default Macintosh gamma was changed to 2.2 in Mac OS X version 10.6).

Category: Colorspace

Linear to XYZ

Takes a Linear input and converts it to the CIE XYZ colorspace. You must provide the source's gamut/profile.

Note: XYZ spaces in Conduit always use D65 as the reference illuminant for user convenience (this matches typical video and digital images).

Category: Colorspace

Log

Computes the logarithm of the input value.

Category: Math

Look-Up Table

Applies a tone correction similar to the Curve nodes (Bezier Curve, Curves) but using a table of look-up values loaded from an external file. Look-up tables are commonly used for specific color correction tasks such as matching colors for a particular output device.

This node is compatible with the same types of files as the equivalent node in the Shake application.

Category: Image

Luma Key

Performs a luminance key operation on the input to produce a matte. The View parameter selects whether the composited image or the resulting matte is output from the node. The Conversion function parameter chooses which of the CIE recommendation functions will be used to compute the luminance values for the luminance map. The Input white and Input black sliders allow you to modify the white and black points for the resulting matte.

Category: Composite

Luminance

Calculates the luminance of the input RGB color, i.e. converts the image to greyscale. The accurate way to calculate luminance depends on the source image's color space, so this node offers the choice between two conversion functions: Rec. 709 (appropriate for HD video and standard computer images), and Rec.601 (appropriate for SD video).

Category: Channels

Multiply

Multiplies two input colors together.

Category: Math

Noise

Renders noise, i.e. a random video signal. This node can generate two types of noise: Gaussian and Uniform.

In Gaussian noise, the pixel values are distributed according to the so-called normal distribution. (See [Wikipedia: Gaussian noise] for a more detailed explanation.)

Gaussian noise is similar to many physical phenomena, so it can be useful to simulate e.g. the noise from a low-light image sensor, or a sensitive film stock.

Uniform noise is a random image where the pixel values are distributed evenly between zero and one. This type of noise looks very "digital", so unless that's the required effect, it's recommended to use Gaussian noise instead. (Gaussian noise is a new addition in Conduit 2.0. Older versions of Conduit only offered uniform noise.)

NOTE: The Gaussian noise is an advanced high dynamic range (HDR) image signal. There are both positive and negative values in the noise. Due to the normal distribution, most of the values will be close to zero, but there will be "outlier" pixels that have values close to -1 and 1. These may cause unexpected effects in your compositing. To ensure that your image doesn't contain HDR values, use the Clip node.

The Noise node will apply the noise over the image that's connected to its input. For Gaussian noise, the noise is blended using the Add operation. For Uniform noise, the Blend operation is used.

To create noise on its own, connect a Color node to the Noise node's input. (Note that when you apply Gaussian noise to a black color, the resulting image will contain also negative values, i.e. pixel values below zero! To make these values visible, you can either use a mid-grey color as the input, or use a Levels node after Noise to change the range of pixel values.)

To create smoothened noise, apply a Gaussian Blur node after Noise.

Category: Image

Normalize 3D

This node computes a normalized vector from the input color vector, i.e. a vector which has the same direction as the input vector but a length of exactly one.

As with the Distance 3D node, the input color value is treated as 3-dimensional coordinates where RGB = XYZ and alpha is ignored. If the input vector is zero (=black), this node will output zero because the input vector has no length and direction. This node also has a second scalar output, which gives the length of the original input vector.

Category: Math

Number

Outputs the designated number value (any floating-point value).

Category: Inputs

Opacity

Modifies the opacity of the input image. Zero is fully transparent, one is fully opaque.

Category: Composite

Outside

Composites input 1 over input 2, outputting just the portions of input 1 outside the matte for input 2.

Category: Composite

Over

Composites the foreground (first input) image over the background (third input), using the foreground's alpha value as its opacity. The middle input can be used to apply a matte for the foreground image.

Important note for Motion: RGBA images within Motion are premultiplied by default, but RGBA images in Conduit tend to be "straight." When compositing images straight from Motion, the "Premultiplied" parameter checkbox should be checked. Otherwise, if for example the alpha has been created in Conduit and applied to an image with Set Matte, you'd want to leave it unchecked.

Category: Image

Place Over

Like the Over node, this node composites the foreground (first input) image over the background (third input). The middle input can be used to apply a matte for the foreground image. This node also performs scaling, rotation and translation on the foreground image.

Category: Image

Premultiply

Premultiplies the input color against black, i.e. multiplies the RGB values with the alpha. Premultiplication is commonly applied to images with an alpha channel, because it's an efficient optimization and practical for most uses. However it destroys the color information in areas that are completely transparent (alpha is zero), so it's not appropriate for images where the alpha channel is used as a mask that would need to be edited later.

Most image formats in use on Mac OS X use premultiplied alpha. To remove premultiplication, use the Unpremultiply node.

Category: Image

Raise to Power

Raises the input value to the specified power.

Category: Math

Reciprocal

Divides 1 by the input value. This operation can be useful for making your conduit run faster on the GPU: division is an expensive operation, so if you find yourself performing several divisions by the same value, you should instead compute the value's reciprocal just once and do multiplies with the reciprocal value.

Category: Math

RGB to HSV

Converts the input from RGB (red / green / blue) to HSV (hue / saturation / value) colorspace.

Category: Colorspace

RGB to YUV

Converts the input from RGB (red / green / blue) to YUV colorspace.

YUV, sometimes more correctly called Y'CbCr, is a colorspace commonly used for video images. It is also known as component video. Of the three components, the first represents luminance, while the two others are color differences.

There are two widely used variants of YUV: Rec. 601 (used for NTSC/PAL footage) and Rec. 709 (used for HD footage). This node has a "Conversion function" parameter for selecting the variant of YUV for output.

Rotate Vector 3D

Rotates an input vector in 3D space around the origin. As with the Distance 3D node, the input color value is treated as 3-dimensional coordinates where RGB = XYZ and alpha is ignored. The rotation parameters are given as degrees.

To rotate an image, use the Place Over node.

Category: Math

Saturation

Adjusts the saturation of the input image based on one of two selected ICE recommendation algorithms. The Conversion function parameter determines which algorithm will be used to transform the image's saturation.

Category: Image

Scalar

Outputs the first channel of the image input as a scalar (single-channel) output. This node is particularly useful for converting a grayscale image into a true scalar value in Conduit, e.g. to use the image as a matte.

Category: Channels

Screen

Applies the foreground image (input 1) over the background image (input 2) using an operation which is the same as the "Screen" blending mode commonly found in image editing applications. Screen is similar to adding the two images together, except that it doesn't create overbright highlights so easily. Screen assumes that the colors are in 0-1 range, so it's not appropriate for HDR images.

Category: Image

Separate RGBA

Separates the red, green, blue and alpha (RGBA) values from the input color.

Category: Channels

Set Matte

Sets the alpha channel of an image (input 1) according to the specified scalar matte (input 2). If the matte input is a color, the matte is taken from its alpha channel.

Category: Channels

Shapes

Renders freeform graphic shapes. Shapes can contain any number of points, with either sharp or smooth corners.

Use on-screen controls in the Viewer to draw the shapes. (In Final Cut Pro and Motion, there is a button labelled "Show On-Screen Controls" to access the drawing tools).

Tutorial available: Shapes Tutorial

Category: Image

Slider

Brings one of the animatable slider values from the host application into the composite. The Host slider parameter selects which of the eight available slider values should be connected to the node.

Category: Inputs

Subtract

Subtracts the second input from the first input.

Category: Math

Supernode (scriptable)

Renders a custom effect. This node can be used to create custom effects of your own creation, packaged for convenient reuse. This is roughly similar to "macros" or "capsules" in some other compositing applications.

By default, this node renders an effect using an embedded conduit effect. To edit this conduit, open the Assets tab in the info view while the node is selected and click on the "Edit" button under the item "Asset 1" in the list.

Rendering can be completely customized using JavaScript. You don't need to write any scripts to make use of this node, however. Effects can be created by simply editing the default "conduit asset" of the supernode.

Effects created using Supernode can be packaged as plugins. These appear in Conduit Editor's Plugins menu, and can be used just like the built-in nodes in Conduit. To package a supernode as a plugin, open the Scripts tab, click on the File button and choose "Save as Plugin in Conduit" from the menu.

Tutorial available: Supernode Tutorial

Category: Special

Switch

Switches between two inputs. Outputs the value of input 1, unless the "use second input" parameter is enabled.

The Switch node can also be used to do per-pixel selection between the input values. To do this, connect a node to the "per- pixel select" parameter input. If this value is zero or negative, the Switch node will output the value from input 1, otherwise it outputs the value from input 2. If a color value is used to drive per-pixel selection, the selection is done for each color channel: e.g. an RGBA value where red is zero and the other channels are positive would mean that the Switch node's output will contain the red channel from input 1 and the other channels from input 2.

Category: Inputs

Switch Channels

Switches an image's channels, e.g. red to green, or blue to alpha. The full on and full off options set all pixels in the channel to on (white) or off (black).

Category: Channels

Unpremultiply

Removes a premultiplied black matte from the input color (see the Premultiply node for discussion on premultiplication).

Undoing premultiplication can be necessary e.g. for color correcting an image which includes an alpha channel that should not be modified by the color correction. In this case, all operations are performed on the unpremultiplied image, and the original alpha is reapplied using Set Matte and Premultiply nodes at the end.

Category: Channels

Unspill

Removes spill from selected portions of the input image by adding or subtracting color on a per-channel basis. Input 2 is the scalar spill map, which determines which parts of the image will be affected by the operation. The red, green and blue parameters determine the compensating hue shift that will be applied the output. If the alpha channel of the foreground is premultiplied, the Foreground is premultiplied checkbox can used to perform an unpremultiplication. The invert spill map checkbox will invert the spill map before performing the unspill operation.

Category: Composite

Video to Linear

Takes a "video" input (i.e. a standard video or computer image) and outputs a linear image. The gamma of the source footage must be known, and is usually 2.2. If the input image is a graphic created on the Mac before 2009, the gamma is probably 1.8. (Mac gamma was changed to the more standard value of 2.2 in Mac OS X version 10.6).

Category: Colorspace

xyY to XYZ

Takes a CIE xyY input and converts it to the CIE XYZ colorspace. This could be useful for converting white points (which are often in xyY format) to XYZ format.

Note: XYZ spaces in Conduit always use D65 as the reference illuminant for user convenience (this matches typical video and digital images).

Category: Colorspace

XYZ to L*a*b*

Takes a CIE XYZ input, and converts it to the CIE L*a*b* colorspace. Requires the reference illuminant of the Lab output (which is typically D50).

Note: XYZ spaces in Conduit always use D65 as the reference illuminant for user convenience (this matches typical video and digital images).

Category: Colorspace

XYZ to Linear

Takes a CIE XYZ input and outputs a linear RGB image. The user can select the destination RGB colorspace.

Note: XYZ spaces in Conduit always use D65 as the reference illuminant for user convenience (this matches typical video and digital images).

Category: Colorspace

XYZ to xyY

Takes a CIE XYZ input and converts it to the CIE xyY colorspace.

Note: XYZ spaces in Conduit always use D65 as the reference illuminant for user convenience (this matches typical video and digital images).

Category: Colorspace

YUV to RGB

Takes an YUV input and converts it to the RGB colorspace.

For a discussion of the YUV colorspace, please see the section about the RGB to YUV node.

Category: Colorspace