Extended Shader Nodes¶

1. General Eevee Utility Nodes¶

Render Info¶

Entry¶

Add > Input > Render Info

Outputs¶

Frag Coord: Screen-space coordinate (xy normalized to 0-1, z is depth)
Width: Render region width
Height: Render region height

Purpose¶

Provides the coordinate and pixel size of the current Eevee render window.

Scene Time¶

Entry¶

Add > Input > Scene Time

Input¶

Scale: Value used to scale frame numbers

Outputs¶

Frame: Current frame number
Seconds: Seconds corresponding to the current frame
Timeline: Scene time remapped to 0-1 from start frame to end frame
Scaled Frame: Current frame divided by Scale

Screen Derivative¶

Entry¶

Add > Utilities > Math > Screen Derivative

Feature¶

Gets differences between neighboring pixels in screen space:

DDX: Screen-space derivative in the X direction
DDY: Screen-space derivative in the Y direction
DDXY: Combination of DDX and DDY (DDX + DDY)

Portal In / Portal Out¶

Entry¶

Add > Layout > Portal In
Add > Layout > Portal Out

Feature Description¶

These are “portal” nodes used to organize node links.

Their workflow can be understood as:

Portal In: Store a named, typed value in the current node tree
Portal Out: Retrieve that value later by name within the same node tree

Other Notes¶

Creating a new Portal In automatically generates a unique name.
Portal Out has a magnifier button that can jump to the matching Portal In location.

Limits¶

Only recognized inside the same shader node tree.
Does not support crossing different node trees.
Does not automatically pass through node groups.
Inputs with the same name should keep only one source.

2. Eevee Object Material Nodes¶

Render Texture¶

Entry¶

Add > Texture > Render Texture

Purpose¶

Reads a Render Textures entry configured earlier in the scene.

Inputs / Outputs¶

Input: Vector
Outputs: Color, Alpha

Screenspace Info¶

Entry¶

Add > Input > Screenspace Info

Inputs / Outputs¶

Input: View Position (camera-space position)
Outputs: Scene Color (scene color), Scene Depth (scene depth)

Purpose¶

Gets the contents of the current render buffer color or depth.

Usage Notes¶

Raytracing must be enabled in render settings
Set material Render Method to Dithered
Enable Raytraced Transmission in material options
The default View Position input is position transformed into camera space, then with the z-axis inverted

World Environment¶

Entry¶

Add > Input > World Environment

Inputs / Outputs¶

Input: Direction (sampling direction)
Output: Color (environment color)

Purpose¶

Directly samples the Eevee world environment color without depending on whether any geometry exists behind the screen.

Notes¶

Reads the world lighting probe color; probe resolution can be adjusted in the world environment

If Direction is not connected, the current surface view direction is used by default
If Direction is connected, the world environment can be sampled in the specified direction

World To Tangent¶

Entry¶

Add > Utilities > Vector > World To Tangent

Inputs / Outputs¶

Input: Vector (world-space direction)
Output: Vector (tangent-space direction)

Purpose¶

Converts a world-space direction vector into the tangent space of the current surface.

Notes¶

A UV Map can be specified in the node panel, and the tangent of that UV is used as the transform basis.

Basis Transform¶

Entry¶

Add > Utilities > Vector > Basis Transform

Inputs / Outputs¶

Input: Vector (point, direction, or normal to transform)
Input: Origin (origin of the custom basis, used only in Point mode)
Input: X Axis, Y Axis, Z Axis (custom basis axes)
Output: Vector (transformed result)

Purpose¶

Uses origin + three basis axes inside material nodes to perform custom coordinate-system transforms. This is useful when there is no matrix input type available and you still need to process points, direction vectors, or normals.

Panel Options¶

Direction
- To Basis: Interpret the input from world / current coordinates into the custom basis
- From Basis: Convert the input from custom basis coordinates back to external coordinates
Type
- Point: Includes Origin translation
- Vector: Only transforms direction and length, without translation
- Normal: Transforms following normal rules and normalizes before output
Basis Input
- XYZ: Use all three input axes directly
- XY / XZ / YZ: Use only two axes; the third axis is generated automatically by cross product
Orthonormalize
- When enabled, input axes are orthogonalized and normalized. This is more suitable for tangent space or local orientation bases
- When disabled, input axis lengths are preserved, which can be used for basis transforms with scaling
Fallback
- Pass Through: Output the original input when the basis degenerates
- Zero: Output 0, 0, 0 when the basis degenerates

SDF Primitive¶

Entry¶

Add > Texture > SDF Primitive

Output¶

Distance

Purpose¶

Generates signed distance field (SDF) base shapes directly inside material nodes. It is useful for procedural masks, silhouettes, shape transitions, and as the starting point for boolean-style combinations.

Main Modes¶

3D shapes: Sphere, Box, Torus, Cone, Point Cone, Cylinder, Point Cylinder, Capsule / Line, Octahedron, Hex Prism, Hex Prism Incircle, Plane, Solid Angle, Pyramid, Disc, 3D Circle
2D shapes: Circle, Rectangle, Ellipse, Triangle, Pentagon, Hexagon, Isosceles Triangle, Trapezoid, Rhombus
Stylized 2D shapes: Star, Heart, Pie, Arc, Moon, Vesica, Cross, Rounded X, Horseshoe, Round Joint, Flat Joint
Curve / segment shapes: Line, Corner, Quadratic Bezier, Point Triangle, Quad, Parabola, Parabola Segment, Uneven Capsule

Input Notes¶

The fixed base input is Vector
Other sockets are shown and renamed dynamically per mode. Common parameters include Size, Radius, Angle, Roundness, Linewidth, Point to Point_003, and Value1 to Value4
The node panel provides Mode and Invert

Usage Notes¶

The output is a distance value, not a color
It is typically combined with nodes such as Math, ColorRamp, Map Range, and SDF Operator to turn the field into a mask or final shape
Invert flips the inside / outside relationship directly, which is useful for turning the same shape into a hole or shell

SDF Operator¶

Entry¶

Add > Converter > SDF Operator

Output¶

Distance

Purpose¶

Combines, trims, and reshapes one or two SDF distance fields so multiple primitives can be assembled into more complex results.

Main Operations¶

Single-input operations: Dilate, Onion, Annular, Mask, Flatten, Invert, Hermite Pulse
Two-input operations: Blend, Exclusion XOR, Divide, Pipe, Engrave, Groove, Tongue
Union family: Union, Smooth Union, Round Union, Columns Union, Stairs Union, Chamfer Union
Intersection family: Intersect, Smooth Intersect, Round Intersect, Columns Intersect, Stairs Intersect, Chamfer Intersect
Difference family: Difference, Smooth Difference, Round Difference, Columns Difference, Stairs Difference, Chamfer Difference

Input Notes¶

Base inputs include Distance, Distance_001, Value, Value_001, and Count
The visible socket names and counts change automatically with Operation
Mask exposes an extra Invert toggle

Usage Notes¶

A common workflow is to build several shapes with SDF Primitive, then combine them with SDF Operator through union, intersection, or difference
Smooth, Round, Chamfer, Stairs, and Columns are useful for more stylized boolean transitions
The node still outputs a distance value, so it usually needs a threshold, color remap, or alpha control afterward

SDF Vector Operator¶

Entry¶

Add > Utilities > Vector > SDF Vector Operator

Outputs¶

Vector
Position
Value

Purpose¶

Preprocesses the coordinate, UV, or vector domain used by SDF workflows. Instead of generating a distance field directly, it rewrites the sampling space before the data reaches SDF Primitive.

That makes it useful for workflows such as:

mirror or repeat the domain first
then generate the primitive in that modified space
then combine the resulting distance fields with SDF Operator

Main Operation Groups¶

Plane Reflect, Mirror, Polar, Repeat Infinite, Repeat Infinite Mirror, Repeat Finite, Octant
These modes handle reflection, symmetry, radial segmentation, and repeated spatial cells
Swizzle, Rotate, Spin, Extrude, Twist, Swirl, Pinch Inflate, Radial Shear, Bend
These modes reorder or deform the coordinate system itself
UV Rotate, UV Scale, UV Grid, UV Random Rotate, UV Random Flip, UV Tileset
These modes are focused on UV tiling, local UV transforms, and per-cell variation
Map -1-1, Map -0.5-0.5, Map 0-1
These modes quickly convert between normalized UV ranges and the centered ranges often used in SDF setups

Mode Reference¶

Plane Reflect
Reflects the domain using a custom plane normal and offset
Value can be used as a helper mask for which side of the plane is active
Mirror
Mirrors space across an axis-aligned plane controlled by the selected Axis
Spacing controls the reference interval
Position exposes an auxiliary mirrored / cell position
Polar
Rewrites planar space into repeated angular sectors around the origin
Useful for radial motifs, petals, emblems, and gear-like repetition
Repeat Infinite
Repeats the domain endlessly using Spacing
Repeat Infinite Mirror
Repeats endlessly while mirroring every second cell, which helps neighboring boundaries line up more naturally
Repeat Finite
Similar to infinite repeat, but constrained by Count
Octant
Folds the domain into an octant / symmetric wedge for quick symmetrical constructions
Swizzle
Reorders axis channels such as XYZ, XZY, or YZX
Rotate
Rotates the domain around the selected main axis
Spin
Applies an axis-based spin style coordinate offset
Useful for rotational motifs and axial distortion
Extrude
Turns a 2D distance domain into a thickness along the third axis
Value outputs an internal-distance helper value
Twist
Twists the domain along the chosen axis
Swirl
Creates a vortex-like distortion around a center
Center, Offset, Strength, and Radius control the affected region
Pinch Inflate
Compresses or inflates space around a central region
Radial Shear
Applies radial shear, useful for stronger rotational distortion patterns
Bend
Bends the domain along the selected axis
UV Rotate
Rotates UV coordinates around Center
UV Scale
Scales UV coordinates around the UV center
UV Grid
Splits 0-1 UV space into a regular grid
Vector outputs the local UV inside the current cell
Position outputs the cell coordinate for downstream indexing or randomization
UV Random Rotate
Uses the input Position to pick a 90-degree random rotation per cell
UV Random Flip
Uses the input Position to randomly flip or rotate each cell
UV Tileset
Remaps the current UV into a sub-tile inside a larger texture sheet
Index picks the tile, Padding controls margins, and Scale adjusts tile-space zoom
Map -1-1
Remaps 0-1 UV into -1 to 1
Map -0.5-0.5
Remaps 0-1 UV into -0.5 to 0.5
Map 0-1
Remaps a centered SDF-style range back into standard UV space

Bevel¶

Entry¶

Add > Input > Bevel

Inputs / Outputs¶

Input: Radius (bevel radius), Normal (surface normal hint)
Output: Normal (approximated beveled normal)

Panel Option¶

Samples (higher sample counts improve quality but cost more performance)

Purpose¶

Generates an approximate beveled normal in Eevee so hard edges can look smoother.

Notes¶

Cycles still uses the original true geometric bevel algorithm
Eevee here uses a same-object screen-space approximation
The result depends on the current view, depth buffer, and visible neighborhood, and is not equivalent to the true Bevel in Cycles

Curvature¶

Entry¶

Add > Input > Curvature

Inputs¶

Samples
Sample Radius
Thickness
Scale

Outputs¶

Scene Curvature: Curvature extracted from screen space
Scene Rim: Rim-light style edge output

Panel Option¶

Local: Ignore depth from other objects

Notes¶

A curvature node ported from Goo Engine that provides curvature and rim outputs.

Shader Info¶

Entry¶

Add > Input > Shader Info

Inputs¶

World Position: World-space position (defaults to the current position)
Normal: Surface normal (defaults to the current smooth normal)

Outputs¶

Diffuse Shading: Lambert lighting
Shadow: Shadow mask
Ambient Lighting: Indirect ambient light from the world environment and lighting probes
Half-Lambert Factor: Half-Lambert lighting term

Notes¶

Shadow
- Supports selectable shadow modes
- Built-in: Default mode, using Eevee's original shadow calculation
- Soft Filtered: Turns binary dithered shadows into smoother grayscale penumbra
The node panel includes a Lightgroup property
- Only lights with the same Lightgroup ID participate in this Shader Info node's direct lighting and shadow evaluation

The current implementation excludes the world sun from these outputs so HDRIs or “sun” contributions embedded in the world environment do not contaminate the direct result.

Light Info¶

Entry¶

Add > Input > Light Info

Feature Description¶

Reads information from a specified light.

Fixed Outputs¶

Color: Light color
Power: Light intensity
Type: Light type
- -1: No light specified
- 0: Point
- 1: Sun
- 2: Spot
- 3: Area

Outputs That Appear Depending on Light Type¶

Position: Light world position
Direction: Light direction
Radius: Light radius
Spot Size: Spot light size
Sun Angle: Sun angle

Notes¶

For per-light processing, use For Each Light inside NPR Tree instead.

Filter Object Info¶

Entry¶

Add > Input > Filter Object Info

Available only in the Filter domain.

Purpose¶

Reads the world-space transform and viewport display color of a chosen object, so filter materials can react to a controller object or scene helper.

This is useful for object-driven masks, directional gradients, moving focal effects, or passing a custom color control into a full-screen filter.

Node Setting¶

Object: Choose which object the node should read

Outputs¶

Location: Chosen object world-space location
Rotation: Chosen object world-space Euler rotation in radians
Scale: Chosen object world-space scale
Color: Chosen object viewport display color

Notes¶

This reads the explicitly selected object, not the object currently being filtered on screen
If no object is assigned, the node falls back to 0 for location / rotation / color and 1 for scale

Scene Color¶

Entry¶

Add > Input > Scene Color

Available only in the Filter domain.

Purpose¶

Reads the current Eevee scene buffer. The Source can be switched in the node panel:

Color: Reads the final rendered scene color
Depth: Reads linear depth
Normal: Reads rendered normals
Position: Reads world-space positions

Inputs / Outputs¶

Input: Vector
Outputs: Color, Alpha