Transform images
Add transformations to your image policy to automate processes like cropping, rotating, resizing, and applying visual effects to your images.
Overview
Image and Video Manager automatically applies a default policy named .auto
with baseline settings that determine how to generate derivative images unless a request includes a parameter indicating a valid custom policy name.
You can can modify the default policy to select different image transformations, but you can't delete it. The preset values in the policy set the image quality at 85
and width values for derivative images at 720
, 1080
, 1366
, 1920
, and 5000
pixels.
Create or update an image policy
- Review the information in the Managing processing section to understand the basics about image transformation application and processing.
- Use the image policy data source to set up your transformations, building out the data block with the transformation arguments here.
- Set the output of the data source to a JSON file or use the output as a variable for the value of the
json
argument in the policy image resource.
Manage processing
Application of image effects occurs in both initial processing and post-production and on the areas within the break point of an image and outside.
- List the effects to apply to the image area inside the break point in the
transformations
argument. - Use the
post_breakpoint_transformations
argument to list the effects to apply to the image area outside the break point.
Use Variables
Many Image and Video Manager arguments let you specify a variable object instead of a string, number, or boolean value.
When using variables, define the variable name in an argument that ends in _var
. Use the variable as you would in any other argument.
Organize
Control and organize how to apply image transformations.
Argument | Description |
---|---|
compound | Groups an ordered sequence of transformations together as a single transformation. |
if_dimension | Chooses a transformation depending on the dimensions of the source image. Contains:
|
if_orientation | Chooses a transformation depending on the orientation of the source image. Contains:
|
im_query | Applies artistic transformations on a per-image basis by specifying transformations with a query string appended to the image URL. If used, all arguments required. Contains:
|
Manipulation
Cut or combine elements from multiple images to reduce size, direct focus, or create new images.
- All transformations are available for use during initial processing.
- A subset are available for use in post-processing to fine-tune and polish your images post-production.
Append
Use the append
block with its child arguments to place a specified image beside the source image by a major dimension first and then on the minor.
Transparent pixels fill any area not covered by either image.
Argument | Required | Description |
---|---|---|
image | ✔️ | The image type. Value is one of:
|
gravity | The placement of the image relative to the source image. The default centers the image. Value is one of:
gravity_var instead. | |
gravity_priority | Determines the exact placement of the image when gravity is Center or a diagonal direction. Value is one of:
gravity_priority_var instead. | |
preserve_minor_dimension | Whether to preserve the source image's minor dimension. The default is false .
preserve_minor_dimension_var instead. |
Aspect crop
Use the aspect
block with its child arguments to change the height or width of an image's aspect ratio by either cropping or expanding the area.
All arguments optional.
Argument | Description |
---|---|
allow_expansion | Increases the size of the image canvas to achieve the requested aspect ratio instead of cropping the image. If setting a variable for this argument, use allow_expansion_var instead. |
height | The height of the aspect ratio to crop. If setting a variable for this argument, use height_var instead. |
width | The width of the aspect ratio to crop. If setting a variable for this argument, use width_var instead. |
x_position | The horizontal portion of the image you want to keep when applying aspect ratio cropping. If using allow_expansion , this setting defines the horizontal position of the image on the new expanded image canvas. If setting a variable for this argument, use x_position_var instead. |
y_position | The horizontal portion of the image you want to keep when applying aspect ratio cropping. If using allow_expansion , this setting defines the horizontal position of the image on the new expanded image canvas. If setting a variable for this argument, use y_position_var instead. |
Crop
Use the crop
block with its child arguments to reduce the area of an image.
All arguments optional.
Argument | Description |
---|---|
allow_expansion | If cropping an area outside of the existing canvas, expands the image canvas. If setting a variable for this argument, use allow_expansion_var instead. |
gravity | The placement of the crop relative to the source image. The available values include eight cardinal directions (North , South , East , West , NorthEast , NorthWest , SouthEast , SouthWest ) and Center by default. If setting a variable for this argument, use gravity_var instead. |
height | The number of pixels to crop along the y-axis. If setting a variable for this argument, use height_var instead. |
width | The number of pixels to crop along the x-axis. If setting a variable for this argument, use width_var instead. |
x_position | The x-axis position to crop from. If setting a variable for this argument, use x_position_var instead. |
y_position | The y-axis position to crop from. If setting a variable for this argument, use y_position_var instead. |
Face crop
Use the face_crop
block with its child arguments to detect faces in the source image and applies the rectangular crop on either the biggest
face or all
of the faces detected.
Image and Video Manager tries to preserve faces in the image instead of using specified crop coordinates.
All arguments optional.
Argument | Description |
---|---|
algorithm | Specifies the type of algorithm used to detect faces in the image: cascade (default) for the cascade classifier algorithm, or dnn for the deep neural network algorithm. If setting a variable for this argument, use algorithm_var instead. |
confidence | With algorithm set to dnn , the minimum confidence needed to detect faces in the image. Values range from 0 to 1 for increased confidence, and possibly fewer faces detected. If setting a variable for this argument, use confidence_var instead. |
fail_gravity | Controls placement of the crop if no faces are detected in the image. Directions are relative to the edges of the image being transformed. The available values represent the eight cardinal directions (North , South , East , West , NorthEast , NorthWest , SouthEast , SouthWest ) and Center by default. If setting a variable for this argument, use fail_gravity_var instead. |
focus | The focus of the crop rectangle, which is either around biggestFace or allFaces . This is all by default. If setting a variable for this argument, use focus_var instead. |
gravity | The placement of the crop relative to the faces, plus padding. The available values represent the eight cardinal directions (North , South , East , West , NorthEast , NorthWest , SouthEast , SouthWest ) and Center by default. If setting a variable for this argument, use gravity_var instead. |
height | The height of the output image in pixels relative to the specified style value. If setting a variable for this argument, use height_var instead. |
padding | The padding ratio based on the dimensions of the biggest face detected, This is 0.5 by default. Larger values increase padding. If setting a variable for this argument, use padding_var instead. |
style | to crop, or scale a crop area, for the faces detected in the source image. This is zoom by default. The output image is resized to the specified width and height values. A value of crop places a raw crop around the faces relative to the specified gravity value. A value of fill scales the crop area to include as much of the image and faces as possible, relative to the specified width and height values. A value of zoom scales the crop area as small as possible to fit the faces, relative to the specified width and height values. If setting a variable for this argument, use style_var instead. |
width | The width of the output image in pixels relative to the specified style value. If setting a variable for this argument, use width_var instead. |
Feature crop
Use the face_crop
block with its child arguments to identify prominent features of the source image to crop around relative to the specified width
and height
values.
All arguments optional.
Argument | Description |
---|---|
fail_gravity | The placement of the crop if Image and Video Manager does not detect any features in the image. Directions are relative to the edges of the image being transformed. The available values represent the eight cardinal directions (North , South , East , West , NorthEast , NorthWest , SouthEast , SouthWest ) and Center by default. If setting a variable for this argument, use fail_gravity_var instead. |
feature_radius | The size in pixels of the important features to search for. If identified, two features never appear closer together than this value which is 8.0 by default. If setting a variable for this argument, use feature_radius_var instead. |
gravity | The placement of the crop relative to the region of interest plus padding. The available values represent the eight cardinal directions (North , South , East , West , NorthEast , NorthWest , SouthEast , SouthWest ) and Center by default. If setting a variable for this argument, use gravity_var instead. |
height | The height in pixels of the output image relative to the specified style value. If setting a variable for this argument, use height_var instead. |
max_features | The maximum number of features to identify as important features. This maximum number of features is 32 by default. The strongest features are always chosen. If setting a variable for this argument, use max_features_var instead. |
min_feature_quality | The minimum quality level of the feature identified. To be considered important, the feature needs to surpass this value. Image and Video Manager measures quality on a scale from 0 (lowest) to 1 (highest). This is 0.1 by default. If setting a variable for this argument, use min_feature_quality_var instead. |
padding | Adds space around the region of interest. The amount of padding added is directly related to the size of the bounding box of the selected features. Specifically, the region of interest is expanded in all directions by the largest dimension of the bounding box of the selected features multiplied by this value. If setting a variable for this argument, use padding_var instead. |
style | Specifies how to crop or scale a crop area for the features identified in the source image. This is fill by default. The output image resizes to the specified width and height values. A value of crop performs a raw crop around the features relative to the specified gravity value. A value of fill scales the crop area to include as much of the image and features as possible, relative to the specified width and height values. A value of zoom scales the crop area as small as possible to fit the features, relative to the specified width and height values. If setting a variable for this argument, use style_var instead. |
width | The width in pixels of the output image relative to the specified style value. If setting a variable for this argument, use width_var instead. |
Fit and fill
Use the fit_and_fill
block with its child arguments to resize an image to fit within a specific size box and use a fill of that same image to cover any transparent space at the edges.
By default, the fill image has a blur transformation with a sigma value of 8
applied. You can use the fill_transformation
argument to customize the transformation.
All arguments optional.
Argument | Description |
---|---|
fill_transformation | Used to customize the sigma value for the fill image. |
height | The height value of the resized image. If setting a variable for this argument, use height_var instead. |
width | The width value of the resized image. If setting a variable for this argument, use width_var instead. |
Mirror
Use the mirror
block with its child arguments to flip an image horizontally, vertically, or both.
All arguments optional.
Argument | Description |
---|---|
horizontal | Flips the image horizontally. If setting a variable for this argument, use horizontal_var instead. |
vertical | the image vertically. If setting a variable for this argument, use vertical_var instead. |
Region of interest crop
Use the region_of_interest_crop
block with its child arguments to crop around a specified area of interest (region) relative to the specified width
and height
values.
All arguments optional.
Argument | Required | Description |
---|---|---|
region_of_interest | ✔️ | The bounding shape of the important features to search for. Value is one of:
|
gravity | The placement of the crop area relative to the specified area of interest. The available values represent the eight cardinal directions (North , South , East , West , NorthEast , NorthWest , SouthEast , SouthWest ) and Center by default. If setting a variable for this argument, use gravity_var instead. | |
height | The height in pixels of the output image relative to the specified style value. If setting a variable for this argument, use height_var instead. | |
style | Specifies how to crop, or scale a crop area, for the specified area of interest in the source image. The output image resizes to the specified width and height values. A value of crop places raw crop around the point of interest, relative to the specified gravity value. A value of fill scales the crop area to include as much of the image and point of interest as possible, relative to the specified width and height values. A value of zoom (the default) scales the crop area as small as possible to fit the point of interest, relative to the specified width and height values. If setting a variable for this argument, use style_var instead. | |
width | The width in pixels of the output image relative to the specified style value. If setting a variable for this argument, use width_var instead. |
Relative crop
Use the relative_crop
block with its child arguments to shrink or expand an image relative to the image's specified dimensions.
- Transparent pixels fill any area not covered in the expanded areas.
- Positive values shrink the side.
- Negative values expand it.
All arguments optional.
Argument | Description |
---|---|
east | The number of pixels to shrink or expand the right side of the image. If setting a variable for this argument, use east_var instead. |
north | number of pixels to shrink or expand the top side of the image. If setting a variable for this argument, use north_var instead. |
south | The number of pixels to shrink or expand the bottom side of the image. If setting a variable for this argument, use south_var instead. |
west | The number of pixels to shrink or expand the left side of the image. If setting a variable for this argument, use west_var instead. |
Resize
Use the resize
block with its child arguments to resize an image to an absolute dimension.
This transformation uses the fit
aspect preservation mode to select a value for the missing dimension that preserves the image's aspect unless you pass values for height
and width
.
All arguments optional.
Argument | Description |
---|---|
aspect | Preserves the aspect ratio. Select fit to make the image fit entirely within the selected width and height. When using fit , the resulting image has the largest possible size for the specified dimensions. Select fill to size the image so it both completely fills the dimensions and has the smallest possible file size. Otherwise ignore changes the original aspect ratio to fit within an arbitrarily shaped rectangle. If setting a variable for this argument, use aspect_var instead. |
height | The height to which to resize the source image. Must be set if width is not specified. If setting a variable for this argument, use height_var instead. |
type | The type of constraints on the image resize. Select normal to resize in all cases, either increasing or decreasing the dimensions. Select downsize to ignore this transformation if the result would be larger than the original. Select upsize to ignore this transformation if the result would be smaller. If setting a variable for this argument, use type_var instead. |
width | The width to which to resize the source image. Must be set if height is not specified. If setting a variable for this argument, use width_var instead. |
Rotate
Use the rotate
block with the child argument degrees
to rotate an image around its center by the given the degrees.
- Positive values rotate clockwise.
- Negative values rotate counter-clockwise.
If setting a variable for this argument, use degrees_var
instead.
Scale
Use the scale
block with its child arguments to change the derivative image's dimensions relative to the original image’s.
All arguments optional.
Argument | Description |
---|---|
height | The scaling factor for the input height that is used to determine the output height of the image. Values between 0 and 1 decrease image height. Values greater than 1 increase the image height. A value of 1 leaves the height unchanged. Image dimensions need to be non-zero positive numbers. If setting a variable for this argument, use height_var instead. |
width | The scaling factor for the input width that is used to determine the output width of the image. Values between 0 and 1 decrease image width. Values greater than 1 increase the image width. A value of 1 leaves the width unchanged. Image dimensions need to be non-zero positive numbers. If setting a variable for this argument, use width_var instead. |
Shear
Use the shear
block with its child arguments to slant an image into a parallelogram as a percent of the starting dimension in decimal format.
- Specify at least one axis argument.
- Transparent pixels fill empty areas around the sheared image as needed. To counter this, use a
background_color
transformation for these areas.
All arguments optional.
Argument | Description |
---|---|
x_shear | The amount to shear along the x-axis, measured in multiples of the image's width. Must be set if y_shear isn't specified. If setting a variable for this argument, use x_shear_var instead. |
y_shear | The amount to shear along the y-axis, measured in multiples of the image's height. Must be set if x_shear isn't specified. If setting a variable for this argument, use y_shear_var instead. |
Smart crop
Use the smart_crop
block with its child arguments to crop around features in an area of interest relative to the given width
and height
values.
This transformation combines the capabilities of the Face Crop and Feature Crop transformations to automatically detect faces and pronounced features in an image.
- When Smart Crop finds more than one cace. it crops around them, preserving as many complete faces as possible within the specified dimensions.
- If it does not find any faces, the crop is around the most prominent feature.
All arguments optional.
Argument | Description |
---|---|
debug | Whether to execute the transformation. The default is false .
|
height | The height in pixels for an image relative to the style value. |
sloppy | Whether to sacrifice any image fidelity for transformation performance. If setting a variable for this argument, use sloppy_var instead. |
style | Specifies how to crop or scale a crop area for an area of interest in the source image. Value is one of:
|
width | The width in pixels for an image relative to the style value. |
Trim
Use the trim
block with its child arguments to automatically crops the background uniformly from the edges of an image.
All arguments optional.
Argument | Description |
---|---|
fuzz | The fuzz tolerance of the trim expressed as a value between 0 and 1 . This determines the acceptable amount of background variation before trimming stops. If setting a variable for this argument, use fuzz_var instead. |
padding | The amount of padding in pixels to add to the trimmed image. If setting a variable for this argument, use padding_var instead. |
Enhancement
Improve the overall quality of your images.
- All transformations are available for use during initial processing.
- A subset are available for use in post-processing to fine-tune and polish your images post-production.
Background color
Use the background_color
block with its child argument color
to set the hexadecimal CSS color value for the background.
If setting a variable for this argument, use color_var
instead.
Blur
Use the blur
block with its child argument sigma
to apply a Gaussian blur to a given value.
- The value of
sigma
scatters original pixels to create the blur effect. - The resulting image may be larger than the original as some pixels may scatter outside the image's original dimensions.
If setting a variable for this argument, use sigma_var
instead.
Chroma key
Use the chroma_key
block with its child arguments to apply a green-screen technique and change any color in an image within the specified volume of the HSL colorspace to transparent or semitransparent.
All arguments optional.
Argument | Description |
---|---|
hue | The hue to remove. Enter the degrees of rotation (between 0 and 360) around the color wheel. By default chroma_key removes a green hue which is at 120 degrees on the color wheel. If setting a variable for this argument, use hue_var instead. |
hue_feather | How much additional hue to make semi-transparent beyond the hue_tolerance . The default is 0.083 which applies semi-transparency to hues 30 degrees around the Hue Tolerance. If setting a variable for this argument, use hue_feather_var instead. |
hue_tolerance | How close a color's hue needs to be to the selected hue for it to be changed to fully transparent. If you enter the maximum value of 1.0 the entire image is made transparent. The default is approximately 0.083 or 8.3% of the color wheel. This value corresponds to 30 degrees around the specified hue. If setting a variable for this argument, use hue_tolerance_var instead. |
lightness_feather | How much additional lightness to make semi-transparent beyond the lightness_tolerance . The default value of 0.1 corresponds to 10% away from the tolerated lightness towards full black or full white. If setting a variable for this argument, use lightness_feather_var instead. |
lightness_tolerance | How much of the lightest part and darkest part of a color to preserve. You can space this value out from the middle (for example, 0.5 lightness or full color) to help preserve the splash lighting impact in the image. You can define how close the color needs to be to the full color to remove it from your image. The default value of 0.75 means that a color must be within 75% of the full color to full white or full black for full removal. If setting a variable for this argument, use lightness_tolerance_var instead. |
saturation_feather | How much additional saturation to make semi-transparent beyond the saturation tolerance. The default is 0.1 which applies semi-transparency to hues 10% below the saturation_tolerance . If setting a variable for this argument, use saturation_feather_var instead. |
saturation_tolerance | How close a color's saturation needs to be to full saturation for it to be changed to fully transparent. For example, you can define how green the color needs to be to remove it from your image. The default value of 0.75 means that a color must be within 75% of full saturation in order to be made fully transparent. If setting a variable for this argument, use saturation_tolerance_var instead. |
Composite
Use the composite
block with its child arguments to apply another image to the source image as an overlay or underlay.
The image underneath is visible in areas that are beyond the edges of the top image or that are less than 100% opaque.
A common use of the overlay composite is to add a watermark.
Argument | Required | Description |
---|---|---|
image | ✔️ | The image type. Value is one of:
|
gravity | The placement of the image relative to the source image. The available values include eight cardinal directions (North , South , East , West , NorthEast , NorthWest , SouthEast , SouthWest ) and Center by default. If setting a variable for this argument, use gravity_var instead. | |
placement | The placement of the applied image either on top of or underneath the base image. Watermarks are usually applied on top. Backgrounds are usually applied underneath. If setting a variable for this argument, use placement_var instead. | |
scale | A multiplier to resize the applied image relative to the source image while preserving the aspect ratio (1 by default). Set the scale_dimension to calculate the scale from the source image's width or height. If setting a variable for this argument, use scale_var instead. | |
scale_dimension | The dimension, either width or height , of the source image to scale. If setting a variable for this argument, use scale_dimension_var instead. | |
x_position | The x-axis position of the image to apply. If setting a variable for this argument, use x_position_var instead. | |
y_position | The y-axis position of the image to apply. If setting a variable for this argument, use y_position_var instead. |
Contrast
Use the contrast
block with its child arguments to adjust contrast and brightness of an image.
All arguments optional.
Argument | Description |
---|---|
brightness | The brightness of the image. Positive values increase brightness and negative values decrease brightness. A value of 1 produces a white image. A value of -1 produces a black image. The default value is 0 , which leaves the image unchanged. The acceptable value range is -1.0 to 1.0 . Values outside of the acceptable range clamp to this range. If setting a variable for this argument, use brightness_var instead. |
contrast | The contrast of the image. Expressed as a range from -1 to 1 . Positive values increase contrast, negative values decrease contrast, while 0 leaves the image unchanged. Values outside of the -1 to 1 range clamp to this range. If setting a variable for this argument, use contrast_var instead. |
Goop
Use the goop
block with its child arguments to distort an image by randomly repositioning a set of control points along a specified grid. The transformed image appears goopy.
Adjust the density of the grid and the degree of randomness. You can use this transformation to create watermarks for use in security.
All arguments optional.
Argument | Description |
---|---|
chaos | The greatest distance control points may move from their original position. A value of 1.0 shifts control points over as far as the next one in the original grid. A value of 0.0 leaves the image unchanged. Values under 0.5 work better for subtle distortions, otherwise control points may pass each other and cause a twisting effect. If setting a variable for this argument, use chaos_var instead. |
density | The density of control points used to distort the image. The largest dimension of the input image is divided up to fit this number of control points. A grid of points is extended on the smaller dimension such that each row and column of control points is equidistant from each adjacent row or column. This parameter strongly affects transformation performance. Be careful choosing values above the default if you expect to transform medium to large sized images. If setting a variable for this argument, use density_var instead. |
power | By default, the distortion algorithm relies on inverse squares to calculate distance but this allows you to change the exponent. You shouldn't need to vary the default value of 2.0 . If setting a variable for this argument, use power_var instead. |
seed | Your own seed value as an alternative to the default, which is subject to variability. This allows for reproducible and deterministic distortions. If all parameters are kept equal and a constant seed is used, goop distorts an input image consistently over many transformations. By default, this value is set to the current epoch time measured in milliseconds, which provides inconsistent transformation output. If setting a variable for this argument, use seed_var instead. |
Grayscale
Use the grayscale
block with its child argument type
to restrict image color to shades of gray. The default value for type
is Rec709
. Supported values:
Brightness
Lightness
Rec601
Rec709
If setting a variable for this argument, use type_var
instead.
HSL
Use the hsl
block with its child arguments to adjust the hue, saturation, and lightness (HSL) of an image.
Notes:
HSL and HSV are similar. Where HSL's
lightness
attribute controls the vibrancy of a color, HSV'svalue
attribute darkens a color.Other transformations can also affect color, such as
grayscale
andmax_colors
.
If you're using more than one of these transformations, consider the order of application for the desired results as the provided order is the processing order.
All arguments optional.
Argument | Description |
---|---|
hue | The number of degrees to rotate colors around the color wheel. The default is 0 . If setting a variable for this argument, use hue_var instead. |
lightness | A multiplier to adjust the lightness of colors in the image. Note that lightness is distinct from brightness. For example, reducing the lightness of a light green might give you a lime green whereas reducing the brightness of a light green might give you a darker shade of the same green. Values less than 1.0 decrease the lightness of colors in the image. Values greater than 1.0 increase the lightness of colors in the image. If setting a variable for this argument, use lightness_var instead. |
saturation | A multiplier to adjust the saturation of colors in the image. Values less than 1.0 decrease saturation and values greater than 1.0 increase the saturation. A value of 0.0 removes all color from the image. If setting a variable for this argument, use saturation_var instead. |
HSV
Use the hsv
block with its child arguments to adjust the hue, saturation, and value (HSV) of an image.
Notes:
HSL and HSV are similar. Where HSL's
lightness
attribute controls the vibrancy of a color, HSV'svalue
attribute darkens a color.Other transformations can also affect color, such as
grayscale
andmax_colors
.
All arguments optional.
Argument | Description |
---|---|
hue | The number of degrees to rotate colors around the color wheel. This is 0.0 by default. If setting a variable for this argument, use hue_var instead. |
saturation | A multiplier to adjust the saturation of colors in the image. Values less than 1.0 decrease saturation and values greater than 1.0 increase the saturation. A value of 0.0 removes all color from the image. If setting a variable for this argument, use saturation_var instead. |
value | A multiplier to adjust the lightness or darkness of the image's base color. Values less than 1.0 decrease the base colors in the image, making them appear darker. Values greater than 1.0 increase the base colors in the image, making them appear lighter. If setting a variable for this argument, use value_var instead. |
Max colors
Use the max_colors
block with its child argument colors
to set the maximum number of colors in an image’s palette. The fewer colors used, the smaller the file size.
If setting a variable for this argument, use colors_var
instead.
Mono hue
Use the mono_hue
block with its child argument hue
to set all hues in an image to a single hue represented by its number on the color wheel, 0
-360
.
Mono hue maintains the original color’s lightness and saturation but sets the hue to that of the specified value. This has the effect of making the image various shades of the specified hue.
If setting a variable for this argument, use hue_var
instead.
Opacity
The opacity
argument sets the level of transparency for an image represented as a value on a scale of 0
to 1
.
- An image with no transparency is opaque.
- Values below
1
decrease opacity. - A value of
0
makes an image completely transparent.
If setting a variable for this argument, use opacity_var
instead.
Remove color
Use the remove_color
block with its child arguments to remove a specified color from an image and replace it with transparent pixels.
All arguments optional.
Argument | Description |
---|---|
color | The hexadecimal CSS color value to remove. If setting a variable for this argument, use color_var . |
feather | Used to minimize any hard edges and to make the color removal more gradual in appearance. This option allows you to extend the color removal beyond the specified tolerance . The pixels in this extended tolerance become semi-transparent which creates a softer edge. The first time there’s a real-time request for an image, this option may result in a slow transformation time, but subsequent requests aren't impacted as the transformed image is served directly from the cache. If setting a variable for this argument, use feather_var instead. |
tolerance | how close the color needs to be to the selected color before it's changed to fully transparent. Set to 0.0 to remove only the exact color specified. If setting a variable for this argument, use tolerance_var instead. |
Unsharp mask
Use the unsharp_mask
block with its child arguments to emphasize edges and details in source images without distorting the colors.
Although this effect is often referred to as sharpening an image, it actually creates a blurred, inverted copy of the image known as an unsharp mask. Image and Video Manager combines the unsharp mask with the source image to create an image that is perceived by the human eye as clearer.
All arguments optional.
Argument | Description |
---|---|
gain | How much emphasis the filter applies to details. Higher values increase apparent sharpness of details. If setting a variable for this argument, use gain_var instead. |
sigma | The standard deviation of the Gaussian distribution used in the unsharp mask, measured in pixels. This is 1.0 by default. High values emphasize large details and low values emphasize small details. If setting a variable for this argument, use sigma_var instead. |
threshold | The minimum change required to include a detail in the filter. Higher values discard more changes. If setting a variable for this argument, use threshold_var instead. |
See also
For more information about the available image transformations, see Add image transformations and conditions.
Updated 8 months ago