Three.js Shading Language

Three.js Shading Language

Introduction

Why TSL?

Creating shaders has always been an advanced step for most developers, many game creators have never created GLSL code from scratch. The shader graph solution adopted today by the industry has allowed developers more focused on dynamics to create the necessary graphic effects to meet the demands of their projects.

The aim of the project is to create an easy-to-use, even if for this we need to create complexity behind, this happened initially with Renderer and now with the TSL.

Other benefits that TSL brings besides simplifying shading creation is keeping the renderer agnostic, while all the complexity of a material can be imported into different modules and use tree shaking without breaking during the process.

Example

A detail map makes things look more real in games. It adds tiny details like cracks or bumps to surfaces, like walls. In this example we will scale uv to improve details when seen up close and multiply with a base texture.

Old

This is how we would achieve that using .onBeforeCompile():

With TSL the code would look like this:

import { texture, uv } from 'three/tsl'

const detail = texture( detailMap, uv().mul( 10 ) );

const material = new MeshStandardNodeMaterial();
material.colorNode = texture( colorMap ).mul( detail );

TSL is also capable of encoding code into different outputs such as WGSL/GLSLWebGPU/WebGL, in addition to optimizing the shader graph automatically and through codes that can be inserted within each Node. This allows the developer to focus on productivity and leave the graphical management part to the Node System.

Another important feature of a graph shader is that we will no longer need to care about the sequence in which components are created, because the Node System will only declare and include it once.

Let’s say that you import positionWorld into your code, even if another component uses it, the calculations performed to obtain position world will only be performed once, as is the case with any other renderer component such as: normalWorld, modelPosition, etc.

Architeture

All TSL component is created from a Node. The Node allows it to communicate with any other, value conversions can be automatic or manual, a Node can receive the output value expected by the parent Node and modify its own output snippet.

Since they are all components are extended from the Node class, it is possible to modulate them using tree shaking. In the shader construction process, the Node will have important information such as geometry, material, renderer as well as the backend, which can influence the type and value of output.

The build process is based on three pillars: setup, analyze and generate.

setup Use TSL to create a completely customized code for the Node output. The Node can use many others within itself, have countless inputs, but there will always be a single output.
analyze This proccess will check the nodes that were created in order to create useful information for generate the snippet, such as the need to create or not a cache/variable for optimizing a node.
generate An output of string will be sent to each node independently, the node will also be able to create code in the flow, supporting multiple lines.

Node also have a native update process invoked by the update() function, these events be called by frame, render call and object draw.

It is also possible to serialize or deserialize a Node using serialize() and deserialize() functions.

Constants and explicit conversions

Input functions can be used to create contants and do explicit conversions.

Conversions are also performed automatically if the output and input are of different types.

Name Returns a constant or convertion of type:
float( node | number ) float
int( node | number ) int
uint( node | number ) uint
bool( node | value ) boolean
color( node | hex | r,g,b ) color
vec2( node | Vector2 | x,y ) vec2
vec3( node | Vector3 | x,y,z ) vec3
vec4( node | Vector4 | x,y,z,w ) vec4
mat2( node | Matrix2 | a,b,c,d ) mat2
mat3( node | Matrix3 | a,b,c,d,e,f,g,h,i ) mat3
mat4( node | Matrix4 | a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p ) mat4
Advanced
ivec2( node | x,y ) ivec2
ivec3( node | x,y,z ) ivec3
ivec4( node | x,y,z,w ) ivec4
uvec2( node | x,y ) uvec2
uvec3( node | x,y,z ) uvec3
uvec4( node | x,y,z,w ) uvec4
bvec2( node | x,y ) bvec2
bvec3( node | x,y,z ) bvec3
bvec4( node | x,y,z,w ) bvec4
imat2( node | Matrix2 | a,b,c,d ) imat2
imat3( node | Matrix3 | a,b,c,d,e,f,g,h,i) imat3
imat4( node | Matrix4 | a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p ) imat4
umat2( node | Matrix2 | a,b,c,d ) umat2
umat3( node | Matrix3 | a,b,c,d,e,f,g,h,i ) umat3
umat4( node | Matrix4 | a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p ) umat4
bmat2( node | Matrix2 | a,b,c,d ) bmat2
bmat3( node | Matrix3 | a,b,c,d,e,f,g,h,i ) bmat3
bmat4( node | Matrix4 | a,b,c,d,e,f,g,h,i,j,k,l,m,n,o,p ) bmat4

Example:

import { color, vec2, positionWorld } from 'three/tsl';

// constant
material.colorNode = color( 0x0066ff );

// conversion
material.colorNode = vec2( positionWorld ); // result positionWorld.xy

Conversions

It is also possible to perform conversions using the method chain:

Name Returns a constant or conversion of type:
.toFloat() float
.toInt() int
.toUint() uint
.toBool() boolean
.toColor() color
.toVec2() vec2
.toVec3() vec3
.toVec4() vec4
.toMat2() mat2
.toMat3() mat3
.toMat4() mat4
Advanced
.toIvec2() ivec2
.toIvec3() ivec3
.toIvec4() ivec4
.toUvec2() uvec2
.toUvec3() uvec3
.toUvec4() uvec4
.toBvec2() bvec2
.toBvec3() bvec3
.toBvec4() bvec4
.toImat2() imat2
.toImat3() imat3
.toImat4() imat4
.toUmat2() umat2
.toUmat3() umat3
.toUmat4() umat4
.toBmat2() bmat2
.toBmat3() bmat3
.toBmat4() bmat4

Example:

import { positionWorld } from 'three/tsl';

// conversion
material.colorNode = positionWorld.toVec2(); // result positionWorld.xy

Method chaining

Method chaining will only be including operators, converters, math and some core functions. These functions, however, can be used on any node.

Example:

// it will invert the texture color
material.colorNode = texture( map ).rgb.oneMinus();

Swizzle

Swizzling is the technique that allows you to access, reorder, or duplicate the components of a vector using a specific notation within TSL. This is done by combining the identifiers:

const original = vec3( 1.0, 2.0, 3.0 ); // (x, y, z)
const swizzled = original.zyx; // swizzled = (3.0, 2.0, 1.0)

It’s possible use xyzw, rgba or stpq.

Operators

Name Description
.add( node | value, ... ) Return the addition of two or more value.
.sub( node | value ) Return the subraction of two or more value.
.mul( node | value ) Return the multiplication of two or more value.
.div( node | value ) Return the division of two or more value.
.assign( node | value ) Assign one or more value to a and return the same.
.remainder( node | value ) Computes the remainder of dividing the first node by the second.
.equal( node | value ) Checks if two nodes are equal.
.notEqual( node | value ) Checks if two nodes are not equal.
.lessThan( node | value ) Checks if the first node is less than the second.
.greaterThan( node | value ) Checks if the first node is greater than the second.
.lessThanEqual( node | value ) Checks if the first node is less than or equal to the second.
.greaterThanEqual( node | value ) Checks if the first node is greater than or equal to the second.
.and( node | value ) Performs logical AND on two nodes.
.or( node | value ) Performs logical OR on two nodes.
.not( node | value ) Performs logical NOT on a node.
.xor( node | value ) Performs logical XOR on two nodes.
.bitAnd( node | value ) Performs bitwise AND on two nodes.
.bitNot( node | value ) Performs bitwise NOT on a node.
.bitOr( node | value ) Performs bitwise OR on two nodes.
.bitXor( node | value ) Performs bitwise XOR on two nodes.
.shiftLeft( node | value ) Shifts a node to the left.
.shiftRight( node | value ) Shifts a node to the right.

const a = float( 1 );
const b = float( 2 );

const result = a.add( b ); // output: 3

Functions

tslFn( function )

It is possible to use classic JS functions or a tslFn() interface. The main difference is that tslFn() creates a controllable environment, allowing the use of stack where you can use assign and conditional, while the classic function only allows inline approaches.

Example:

{

return timer.add( 0.75 ).mul( Math.PI * 2 ).sin().mul( 0.5 ).add( 0.5 );

} );

const value = oscSine( { timer: value } );”>

const oscSine = tslFn( ( { timer = timerGlobal } ) => {

	return timer.add( 0.75 ).mul( Math.PI * 2 ).sin().mul( 0.5 ).add( 0.5 );

} );

const value = oscSine( { timer: value } );

If you want to use an export function compatible with tree shaking, remember to use /*@__PURE__*/

Name Description
PI The value of π (pi).
PI2 The value of 2π (two pi).
EPSION A small value used to handle floating-point precision errors.
INFINITY Represent infinity.
abs( x ) Return the absolute value of the parameter.
acos( x ) Return the arccosine of the parameter.
all( x ) Return true if all components of x are true.
any( x ) Return true if any component of x is true.
asin( x ) Return the arcsine of the parameter.
atan( x ) Return the arc-tangent of the parameters.
atan2( y, x ) Return the arc-tangent of the quotient of its arguments.
bitcast( x, y ) Reinterpret the bits of a value as a different type.
cbrt( x ) Return the cube root of the parameter.
ceil( x ) Find the nearest integer that is greater than or equal to the parameter.
clamp( x, min, max ) Constrain a value to lie between two further values.
cos( x ) Return the cosine of the parameter.
cross( x, y ) Calculate the cross product of two vectors.
dFdx( p ) Return the partial derivative of an argument with respect to x.
dFdy( p ) Return the partial derivative of an argument with respect to y.
degrees( radians ) Convert a quantity in radians to degrees.
difference( x, y ) Calculate the absolute difference between two values.
distance( x, y ) Calculate the distance between two points.
dot( x, y ) Calculate the dot product of two vectors.
equals( x, y ) Return true if x equals y.
exp( x ) Return the natural exponentiation of the parameter.
exp2( x ) Return 2 raised to the power of the parameter.
faceforward( N, I, Nref ) Return a vector pointing in the same direction as another.
floor( x ) Find the nearest integer less than or equal to the parameter.
fract( x ) Compute the fractional part of the argument.
fwidth( x ) Return the sum of the absolute derivatives in x and y.
inverseSqrt( x ) Return the inverse of the square root of the parameter.
invert( x ) Invert an alpha parameter ( 1. – x ).
length( x ) Calculate the length of a vector.
lengthSq( x ) Calculate the squared length of a vector.
log( x ) Return the natural logarithm of the parameter.
log2( x ) Return the base 2 logarithm of the parameter.
max( x, y ) Return the greater of two values.
min( x, y ) Return the lesser of two values.
mix( x, y, a ) Linearly interpolate between two values.
negate( x ) Negate the value of the parameter ( -x ).
normalize( x ) Calculate the unit vector in the same direction as the original vector.
oneMinus( x ) Return 1 minus the parameter.
pow( x, y ) Return the value of the first parameter raised to the power of the second.
pow2( x ) Return the square of the parameter.
pow3( x ) Return the cube of the parameter.
pow4( x ) Return the fourth power of the parameter.
radians( degrees ) Convert a quantity in degrees to radians.
reciprocal( x ) Return the reciprocal of the parameter (1/x).
reflect( I, N ) Calculate the reflection direction for an incident vector.
refract( I, N, eta ) Calculate the refraction direction for an incident vector.
round( x ) Round the parameter to the nearest integer.
saturate( x ) Constrain a value between 0 and 1.
sign( x ) Extract the sign of the parameter.
sin( x ) Return the sine of the parameter.
smoothstep( e0, e1, x ) Perform Hermite interpolation between two values.
sqrt( x ) Return the square root of the parameter.
step( edge, x ) Generate a step function by comparing two values.
tan( x ) Return the tangent of the parameter.
transformDirection( dir, matrix ) Transform the direction of a vector by a matrix and then normalize the result.
trunc( x ) Truncate the parameter, removing the fractional part.

const value = float( -1 );

// It's possible use `value.abs()` too.
const positiveValue = abs( value ); // output: 1

Inputs

Attributes

Name Description Type
attribute( name, type = null, default = null ) Getting geometry attribute using name and type. any
uv( index = 0 ) UV attribute named uv + index. vec2
vertexColor( index = 0 ) Vertex color node for the specified index. color

Position

Name Description Type
positionGeometry Position attribute of geometry. vec3
positionLocal Local variable for position. vec3
positionWorld World position. vec3
positionWorldDirection Normalized world direction. vec3
positionView View position. vec3
positionViewDirection Normalized view direction. vec3

positionLocal represents the position after modifications made by skinning, morpher, etc.

Normal

Name Description Type
normalGeometry Normal attribute of geometry. vec3
normalLocal Local variable for normal. vec3
normalView Normalized view normal. vec3
normalWorld Normalized world normal. vec3
transformedNormalView Transformed normal in view space. vec3
transformedNormalWorld Normalized transformed normal in world space. vec3
transformedClearcoatNormalView Transformed clearcoat normal in view space. vec3

transformed* represents the normal after modifications made by skinning, morpher, etc.

Tangent

Name Description Type
tangentGeometry Tangent attribute of geometry. vec4
tangentLocal Local variable for tangent. vec3
tangentView Normalized view tangent. vec3
tangentWorld Normalized world tangent. vec3
transformedTangentView Transformed tangent in view space. vec3
transformedTangentWorld Normalized transformed tangent in world space. vec3

Bitangent

Name Description Type
bitangentGeometry Normalized bitangent in geometry space. vec3
bitangentLocal Normalized bitangent in local space. vec3
bitangentView Normalized bitangent in view space. vec3
bitangentWorld Normalized bitangent in world space. vec3
transformedBitangentView Normalized transformed bitangent in view space. vec3
transformedBitangentWorld Normalized transformed bitangent in world space. vec3

Camera

Name Description Type
cameraNear Near plane distance of the camera. float
cameraFar Far plane distance of the camera. float
cameraLogDepth Logarithmic depth value for the camera. float
cameraProjectionMatrix Projection matrix of the camera. mat4
cameraProjectionMatrixInverse Inverse projection matrix of the camera. mat4
cameraViewMatrix View matrix of the camera. mat4
cameraWorldMatrix World matrix of the camera. mat4
cameraNormalMatrix Normal matrix of the camera. mat3
cameraPosition World position of the camera. vec3

Texture

Name Description Type
texture( texture, uv = uv(), level = null ) Retrieves texels from a texture. vec4
cubeTexture( texture, uvw = reflectVector, level = null ) Retrieves texels from a cube texture. vec4
triplanarTexture( textureX, textureY = null, textureZ = null, scale = float( 1 ), position = positionLocal, normal = normalLocal ) Computes texture using triplanar mapping based on provided parameters. vec4

Model

Name Description Type
modelDirection Direction of the model. vec3
modelViewMatrix View matrix of the model. mat4
modelNormalMatrix Normal matrix of the model. mat4
modelWorldMatrix World matrix of the model. mat4
modelPosition Position of the model. vec3
modelScale Scale of the model. vec3
modelViewPosition View position of the model. vec3
modelWorldMatrixInverse Inverse world matrix of the model. mat4

Utils

Reflect

Name Description Type
reflectView Computes reflection direction in view space. vec3
reflectVector Transforms the reflection direction to world space. vec3

UV

Name Description Type
matcapUV UV coordinates for matcap material computation. vec2
rotateUV( uv, rotation, centerNode = vec2( 0.5 ) ) Rotates UV coordinates around a center point. vec2
spritesheetUV( count, uv = uv(), frame = float( 0 ) ) Computes UV coordinates for a sprite sheet based on the number of frames, UV coordinates, and frame index. vec2
equirectUV( direction = positionWorldDirection ) Computes UV coordinates for equirectangular mapping based on the direction vector. vec2

Remap

Variable Description Type
remap Remaps a value from one range to another. any
remapClamp Remaps a value from one range to another, with clamping. any

Random

Variable Description Type
hash( seed ) Generates a hash value in the range [ 0, 1 ] from the given seed. float
range( min, max ) Generates a range attribute of values between min and max. any

Oscillators

Variable Description Type
oscSine( timer = timerGlobal ) Generates a sine wave oscillation based on a timer. float
oscSquare( timer = timerGlobal ) Generates a square wave oscillation based on a timer. float
oscTriangle( timer = timerGlobal ) Generates a triangle wave oscillation based on a timer. float
oscSawtooth( timer = timerGlobal ) Generates a sawtooth wave oscillation based on a timer. float

Packing

Variable Description Type
directionToColor( value ) Converts direction vector to color. color
colorToDirection( value ) Converts color to direction vector. vec3

Functions

.toVar( name = null )

To create a variable from a node use .toVar().

The first parameter is used to add a name to it, otherwise the node system will name it automatically, it can be useful in debugging or access using wgslFn.

const uvScaled = uv().mul( 10 ).toVar();

material.colorNode = texture( map, uvScaled );

varying( node, name = null )

Let’s suppose you want to optimize some calculation in the vertex stage but are using it in a slot like material.colorNode.

For example:

// multiplication will be executed in vertex stage
const normalView = varying( modelNormalMatrix.mul( normalLocal ) );

// normalize will be executed in fragment stage
// because .colorNode is fragment stage slot as default
material.colorNode = normalView.normalize();

The first parameter of varying modelNormalMatrix.mul( normalLocal ) will be executed in vertex stage, and the return from varying() will be a varying as we are used in WGSL/GLSL, this can optimize extra calculations in the fragment stage. The second parameter allows you to add a custom name to varying.

If varying() is added only to .positionNode, it will only return a simple variable and varying will not be created.

Transitioning common GLSL properties to TSL

GLSL TSL Type
position positionGeometry vec3
transformed positionLocal vec3
transformedNormal normalLocal vec3
vWorldPosition positionWorld vec3
vColor vertexColor() vec3
vUv | uv uv() vec2
vNormal normalView vec3
viewMatrix cameraViewMatrix mat4
modelMatrix modelWorldMatrix mat4
modelViewMatrix modelViewMatrix mat4
projectionMatrix cameraProjectionMatrix mat4
diffuseColor material.colorNode vec4
gl_FragColor material.fragmentNode vec4

Read More

Leave a Reply

Your email address will not be published. Required fields are marked *