Core Image Filter (Built-in Patch) in Programming, Program
This patch implements an image processing filter written using a subset of OpenGL Shading Language and the Core Image extensions (refer to the OpenGL and Core Image documentations for a description of the language).
In default mode, users define a unique image kernel, and the inputs of this patch are automatically created according to the prototype of the kernel function: “float”, “vec2”, “vec3” and “vec4” parameters are turned into number inputs, “__color” parameters into color inputs and “sampler” parameters into image inputs. Be aware that changing the name of a kernel function parameter will break the connections to the related input.
This patch always has a single output image, which represents the result produced by the kernel.
In that mode, by default the Domain Of Definition of the resulting image is defined as the union of the input images used by the kernel. If there are no input images or that behavior has been disabled in the patch’s settings, the resulting image will have infinite dimensions. Because Quartz Composer doesn’t know how to render such images, you will need to convert it to finite dimensions before rendering it – typically using the Image Crop patch.
Note that this patch might not produce any image if any of its image inputs is undefined.
If the “Edit Filter Function” setting is active, users can use more than one kernel and define the filter function that calls them using JavaScript. They can also define precisely the DOD of the produced image. In this mode, the inputs of the patch are automatically created from the prototype of the JavaScript “main” function.
|
Input Attributes
|
Output Attributes
|