Network editor

Overview

The network editor lets you see and edit the nodes in a folder in the scene hierarchy, and the connections between the nodes.

Usually you will have a network editor alongside a linked parameter editor , which lets you view and edit the parameters of the nodes you select in the network editor. You can also press P in the network editor to split it into a network view and a mini parameter editor.

Navigating the view

To...	Do this
Pan and zoom the network view	Drag to pan Use to zoom in and out. Or, hold Space and drag . Press O to show a small representation of the current network in the corner. Clicking or dragging on this overview pans the view around the network.
Automatically Pan and zoom to show all nodes	Right click the network background and choose Viewing controls > Home all, or press H.
Automatically Pan and zoom to show the selected nodes	Right click the network background and choose Viewing controls > Home selected, or press G.
Display detailed information on a node	Press on a node.
Select nodes by name	Click the Search icon, or press / to start quick find. To only find nodes of a certain type, choose a type from the pop-up menu. Type the name of the node and press Enter. You can use wildcards. If there is no node with the exact name you type, Houdini selects all nodes whose names contain the text.
Go inside a network node to see its contents	Double-click the node. Select the node and press I or Enter. on an object in the viewer and choose Edit Geometry, Edit subnetwork, etc.
Go up a level to the network above the current one	Click the level to go to in the path above the network view. Press U in the network editor.
Select the previous (input) or next (output) node in the current chain	Press Page Up or Page Down.
Go the previous or next network location viewed (like a browser)	Use the arrow buttons in the network editor toolbar. Press Alt + or Alt + .

How to create nodes in the network editor

Many actions in Houdini automatically create nodes in the network, such as using the tools on the shelf. You can also manually add nodes in the network editor. For example, if you are viewing a geometry network you can create a Sphere node in the network editor pane, and a sphere will appear in the 3D view.

See the tab menu for information on using the tab menu to create nodes.

Work with nodes

To...	Do this
Select a node	Click the node in the network view.
Select multiple nodes	Hold Shift and click nodes to add them to the selection. Press in empty space and drag a marquee around the nodes you want to select.
Move a node in the layout area	Drag the node.
Rename a node	Click the name next to the node to make it editable.
Set the color of a node	This lets you annotate different areas, functions, and/or classes of nodes in the view using color. For example, in a shading network you could color all the lighting model VOPs yellow to make them easier to find at a glance. Press C to show the color palette. Select the node(s) you want to color. Click a color in the palette. Press C again to hide the palette.
Delete nodes	Select the nodes and press Del.
Cut, copy, and paste one or more nodes	Select the nodes. Right-click a selected node and choose Cut, Copy, or Paste.
Show a context menu of commands related to a node	Press on the node. You can also press on the name of a node in the path shown at the top of a pane to get that node’s context menu.
Change the display of the network	In the network editor, press D or right-click the background and choose Viewing controls > Display options. Settings you can change include: How wires are drawn. Show a grid behind the nodes. Show a background image behind the nodes. This is often useful for characters where you can positions the character’s nodes over a schematic image of the character. Show extra links between nodes representing dependencies.

Tip	When you delete a node, the network tries to regenerate itself by connecting all the descendants of the deleted node to the node’s first parent. You can Delete a node and then insert a new node using on the parent’s output to quickly replace a node.

Tip	You can copy and paste between multiple running instances of Houdini. To make each running copy of Houdini have a separate clipboard, use the SESI_COPY_SUFFIX environment variable.

Connect (wire) nodes together

Connections between nodes have different meanings in different network types.

In Object (OBJ) networks, the connections control parenting (connect child objects to the outputs of parent objects).
In particle (POPs), geometry (SOPs), and compositing (COPs) networks, connections control the flow of geometry data through the network.
In render output (ROPs) networks, connections control dependencies between render passes (connect later renders to the outputs of prerequisite renders).

Tip	You can hold Space and pan and zoom the network view while connecting, if you need to bring the node you're connecting to into view.

To...

Do this

Connect existing nodes

Click the input/output connector of a node.
Click another node.

Note	You can also press Alt + on a node to start the connection. This is useful in large networks where you may be zoomed out, making the nodes appear smaller.

Click in empty space to cancel the connection.

Manage connections

Right-click a wire to see a menu of operations you can perform on the wire.

Delete a connection

Click the connection’s input, then click in empty space in the layout area.
or
Right-click the wire and choose Disconnect.

Branch an already-connected output to another input

Click an output that is already connected to something.
Click the additional input.

Create a new node and connect it to an input

Press on the input box of a node.
A menu pops up showing the nodes that can be connected here. Choose an node.
Click to create a node and connect it automatically.

Create a new node and insert it in the output stream

Press on the output box of a node.
A menu pops up showing the nodes that can be connected here. Choose an node.
Click to create a node and connect it automatically.

Create a new node and branch it off the output stream

Press on the output box of a node.
A menu pops up showing the nodes that can be connected here. Choose an node to create a node and branch it automatically.

Connect multiple node outputs at once to a single input

This is useful when you want to connect a bunch of nodes to a Merge node.

Select the nodes whose outputs you want to wire.
Click the output of one of the selected nodes and then click the input of the destination node.
Houdini wires the outputs of all the selected nodes to the input of the destination node.

Get help on an input

Press on the input box of a node. Houdini shows a short description of what kind of input is expected.

How to reorder inputs

Many nodes let you connect multiple nodes to their input, and the order of the inputs is sometimes significant (for example, Render nodes (ROPs) render their inputs in order).

To reorder the inputs, use the controls on the input list in the parameter editor.

Select the node in the network pane.
In the parameter editor, scroll down to the bottom of the panel. For nodes with multiple inputs there will be a list of inputs.
Drag and drop the inputs in the list or click the up-arrow icons to reorder them.

Network view

The layout area shows a graphical representation of the nodes in the network and connections between them.

Node

Nodes are represented in the layout area by a box, with an icon showing the node type, flags, and inputs and outputs.

If the node is drawn with red stripes, it means the node did not cook properly. If the tile is drawn with a yellow and black border, it means the node only partially cooked, and may be producing bad data. In either case, press MMB on the tile to see the error or warning message.

Inputs	Connect to the outputs of other nodes. Many SOPs take geometry as input from “upstream” nodes, modify it, and output the modified geometry, to be used as input to other nodes.
Outputs	Connect to the inputs of other nodes. Many SOPs take geometry as input from “upstream” nodes, modify it, and output the modified geometry, to be used as input to other nodes.
Flags	Most node tiles have areas on the tile that you can click to turn certain flags on or off for that node. Some flags toggle on and off when you click the flag area, but some flags can only be on for one node at a time.

Press RMB on the icon to show a context menu for the node.

Node context menu

Toggle `flag name`	Use these menu items to switch the display flag, render flag, template flag, or lock flag on and off. They are the same as clicking the colored flags on the tile.
Take Toggle `flag name`	Add the state of the flag to the current take.
Preview Window	(SOPs only) Opens a new viewport in which you can view and edit the SOP’s geometry. This lets you see the effects of work in other panes, or higher up in the chain, on that sop.
Spreadsheet	(SOPs only) Opens a geometry spreadsheet for the SOP in a new floating window.

Save Geometry

(SOPs only) Lets you save the SOP’s geometry as a .geo, .bgeo (Houdini geometry), .poly, .bpoly (action geometry), .rib (RenderMan), .dxf, .iv (Inventor), or .obj file (see file formats supported by Houdini ).

You can read the geometry back into Houdini use the File SOP.

Parameters	Opens a parameter editor for the node in a new floating window.
Edit Comment	Opens a window where you can view and edit the node’s comment field. This is the same as clicking the comment icon in the parameter editor.

Viewport Sync	(POPs only) Matches the viewport display to the current frame of the simulation. This ensures the display is accurate, but takes more time. For example, say you have changed Editors and the Playbar is set later than the last time the pop was cooked. With Viewport Sync enabled, all the in-between frames will be cooked and the updated simulation shown. If Viewport Sync is off, no frames are cooked, so you must remember that what is shown in the viewport may not be up-to-date. The Reset button in the Playbar will turn yellow to indicate that the viewport is out of date.
Change Type	This lets you change the node to a different node type, while optionally keeping its name, parameters, and network contents.
Export Type	Saves the custom data associated with the node type of the selected node to a `.optype` file. This is the information shared by all nodes of the node type, not information for this specific node. This includes the creation script, the dialog script, the implementation (`.so`, `.vex`, or `.vfl`), the help card, the icon, and the label. You can edit the node type values by choosing Type Properties.
Create Digital Asset	(Subnets only) Creates a new scripted subnet from the node and saves it to a `.optype` file. You can then load the `.optype` file using File > Install OP Type. You can script this using optyperead, optypewrite, optypeinstall, and optypeuninstall.
Create New VOP Type	Embeds the node, for example, a shader node (SHOP), in a new VOP type and saves it to disk. You can then load the new VOP into a VOP network.
Edit Sub-Network	Jump down into the contents of a sub-network.
Edit POP Network	(POP SOP only) Opens a editor for a POP SOP’s POP network .
Hide/Expose Inputs/Outputs	Hide or show the chains of nodes preceding/following the node. A collapsed chain is represented by a dashed connection line.
Select Inputs/Outputs	Selects all nodes that connect into/out of the node.
Channels	Lets you scope channels on the node and create channel groups. This is the same as right-clicking a node in the channels list beside the channel editor.
Reference Copy	Creates a copy of this node that gets all its values from the original.
Delete	Deletes the node. The network does its best to reconnect the remaining nodes.
Type Properties	Opens the Operator Type Properties window.

List view

List view shows all nodes in the current network and their flags and attributes in a tabular format. Right click in the network editor and choose View nodes as a list.

Ordering

Right click in list view and open the List Order submenu to set the order in which nodes appear in the list.

User Defined	Lets you manually set the order of nodes by dragging them.
Alphabetical	Sorts the nodes by name.
Operator Type	Sorts the nodes by type.
Hierarchical	In Object networks (OBJ), this sorts the nodes into a tree according to parent-child relationships. Child objects are indented under parent objects. If the indenting would push the node off the edge of the pane, `+number` is used instead to show how many additional levels the node would be indented.

How to encapsulate nodes inside subnetworks

Subnetworks let you encapsulate multiple nodes inside a single node, streamlining your network visually and conceptually.

The subnet node can have up to four inputs, which are passed down to nodes in the subnet.

You can use subnetworks in geometry (SOP), particle (POP), and compositing (COP) networks.

To...	Do this
Create an empty subnet	Use the tab menu to create a `Subnetwork` node.
Create a subnetwork from existing nodes	Select the nodes you want to go inside the new subnet. Right click one of the nodes and choose Collapse Selected into Subnet.
Connect the subnet’s inputs to node inputs inside the subnet	Go inside the subnet. The layout view shows four pseudo-nodes which represent the subnet’s inputs. Connect the output of one of the pseudo-nodes to the input of one of the real nodes in the subnet.
Choose the subnet’s output	Set the display flag of the node you want to provide the subnet’s output.
Label the subnet’s inputs	View the subnet’s parameters in the parameter editor and click the Subnet tab. Edit the Subnet Input # text fields to control the message that pops up when the user presses on the inputs.

Group nodes visually with network boxes

To...	Do this
Create a network box	Select the nodes you want inside the new box. Press on the network background and choose Viewing Controls > Create Network Box from Selected.
Add or remove a node to/from a network box	Drag the node onto or out of the box.
Move or resize a network box	To move a network box, drag the bar across the top. To resize it, drag from one of the corners.
Remove a network box	Click the X button at the top right of the box.
Collapse a box to a stamp	Click the left button at the top right of the box.
Assign colors to network boxes	Press C in the network editor to show the palette. Select the network box you want to color. Click a color in the color palette.

Troubleshooting

Node icon gets yellow and back outline or red stripes

A yellow-and-black outline means the node only partially cooked, possibly because of bad input data.

Red stripes over the icon mean the node had an error.

Press MMB on the node’s icon to see the warning or error message.