Common properties of nodes -------------------------- The nodes will always have outputs that will be connected to the inputs of other nodes. They may also (and will probably) have parameters that can be modified in the user interface, and inputs to receive data from other nodes. .. image:: images/node_example.png :align: center Outputs ^^^^^^^ Nodes can have several outputs that represent the images they provide. Each node output can be connected to several node inputs. Each node output can be either **greyscale**, **color** or **RGBA**. Greyscale ports are drawn in grey, color in blue and RGBA in semi-transparent green. Next to the output, an (open or closed) eye icon indicates that a preview is available. Clicking on it will show the preview at the bottom of the node. When a node has several outputs, only one of them can be previewed at a time. Inputs ^^^^^^ Inputs are images the node will transform (if any). An input is always connected to at most an output. Inputs generally have a default image that is used when it is not connected. Parameters ^^^^^^^^^^ Parameters are used to configure nodes. The following types are supported: * **float** parameters are used whenever a real number is needed. Float parameters have bounds, but it is possible to ignore them (at your own risk). To modify a float parameter, click in the text field and enter a new value, or click and drag left or right to decrease or decrease its value. When the lower or upper bound is reach, the value wil stick to it, but dragging again from there makes it possible to go beyond the limit. * **size** parameters are power of two values for the image size. They are used when actually storing data in a texture, or when performing resolution dependent calculations such as convolutions. They are shown as drop down list boxes. * **enumerated** parameters are used when nodes have different named options and shown as drop down list boxes. * **boolean** parameters are shown as check boxes. * **color** parameters are shown as Color selectors. * **gradient** parameters are edited in dedicated editor widgets. Cursors can be added by double-clicking in the lower part of the widget, dragged left and right using the left mouse button, and removed using the right mouse button. The drop down list box can be used to select one of the 4 interpolation options. Double clicking in the upper part of the gradient editor will open a larger version of the widget to move cursors mode precisely. Finally, it is possible to drag and drop from the upper part of a gradient editor widget to another to duplicate the gradient parameter, or from any **Colorize** node variant in the library to a gradient editor widget. .. image:: images/gradient_editor.png :align: center Randomness ^^^^^^^^^^ Nodes that provide random patterns have an implicit **seed** parameter. It is not possible to edit it directly, but moving the node in the graph will change its value. It is possible to freeze the seed value using the small dice button in the node's title bar in the graph. .. image:: images/random_node.png :align: center Subgraphs also have their seed and transmit it to their children unless not configured to do so, or the children's seeds are frozen. Modifying nodes ^^^^^^^^^^^^^^^ Most nodes in Material Maker can be modified, but they first have to be made editable. To do this, select a node, and use the **Tools -> Make the selected nodes editable** menu item or the **Control+W** keyboard shortcut. .. image:: images/editable_node.png :align: center When made editable, 3 buttons are shown at the bottom of the node: * A pencil-shaped button to edit the node (the precise behavior of this button depends on the node type) * A folder-like button to load an existing node template * A floppy disk button to save the node as a template To be reusable directly, templates must be saved in the **generators** directory in the install dir (or the **addon/material_maker/nodes** directory when using Material Maker as a Godot addon). All nodes templates saved in this directory are shown in the **Tools -> Create** menu. It is absolutely not necessary to save newly created nodes as templates, but this makes them a lot easier to access and results in smaller material files (only references to the templates are saved and not the whole node description). Consequently, modifying a template without ensuring compatibility with the old version (i.e. removing or renaming parameters, removing or swapping inputs or outputs).