High Level Shading Language (HLSL) is a shader programming language for GPU. Pixel Composer has HLSL supports through the HLSL node. This page will only cover basics of HLSL. For more information, check out the further reading section.

HLSL Node

You can find the HLSL node under the Misc. category. HLSL node takes HLSL fragment code and custom arguments. The code will compile automatically, however you can also compile it manually by clicking the Compile button in the inspector.

Syntax

HLSL runs over every pixel. Process and returns the output pixel color. Let's take a look at the default HLSL fragment code. float4 surfaceColor = gm_BaseTextureObject.Sample(gm_BaseTexture, input.uv);
output.color = surfaceColor; This code does 2 things.

The first line float4 surfaceColor = gm_BaseTextureObject.Sample(gm_BaseTexture, input.uv); Get the color of the current pixel from the input surface and stores it in the surfaceColor variable.

The seconds line output.color = surfaceColor; apply the color to the output.

Variable

Variable in HLSL are strongly typed (type need to be defined when create). Variable types you can use are:

Type	Description
float	Floating point number
int	Integer
bool	Boolean
float2	2D vector
float3	3D vector
float4	4D vector
float3x3	3x3 matrix
float4x4	4x4 matrix

Operators

Operators in HLSL are similar to other programming languages. Here are some of the operators you can use:

+	Addition
-	Subtraction
*	Multiplication
/	Division
%	Modulus
++	Increment
--	Decrement
==	Equal
!=	Not equal
>	Greater than
<	Less than
>=	Greater than or equal
<=	Less than or equal
!	Logical NOT
&&	Logical AND
||	Logical OR

Swizzling

Swizzling is a feature in HLSL that allows you to access the components of a vector. For example, you can access the x and y components of a float3 vector by using .xy suffixes. Here are some examples:

.x	Get a single value from vector.
.xz	Create new float2 consisting of the first and third elements.
.xxx	Create new float3 with all the values equals to the first element.
.rgb	Create new float3 with the first three elements.

Functions Reference

abs(x)

abs

acos(x)

acos

asfloat(x)

asfloat

asin(x)

asin

asint(x)

asint

atan(x)

atan

atan2(y, x)

atan2

ceil(x)

ceil

clamp(x, min, max)

clamp

clip(x)

clip

cos(x)

cos

cosh(x)

cosh

cross(x, y)

cross

ddx(x)

ddx

ddy(x)

ddy

degrees(x)

degrees

determinant(x)

determinant

distance(x, y)

distance

dot(x, y)

dot

exp(x)

exp

exp2(x)

exp2

faceforward(n, i, ng)

faceforward

floor(x)

floor

fma(a, b, c)

fma

fmod(x, y)

fmod

frac(x)

frac

frexp(x, exp)

frexp

fwidth(x)

fwidth

isfinite(x)

isfinite

isinf(x)

isinf

isnan(x)

isnan

ldexp(x, exp)

ldexp

length(x)

length

lerp(x, y, s)

lerp

lit(n_dot_l, n_dot_h, m)

lit

log(x)

log

log10(x)

log10

log2(x)

log2

max(x, y)

max

min(x, y)

min

modf(x, out ip)

modf

mul(x, y)

mul

noise(x)

noise

normalize(x)

normalize

pow(x, y)

pow

radians(x)

radians

rcp(x)

rcp

reflect(i, n)

reflect

refract(i, n, ?)

refract

round(x)

round

rsqrt(x)

rsqrt

saturate(x)

saturate

sign(x)

sign

sin(x)

sin

sincos(x, out s, out c)

sincos

sinh(x)

sinh

smoothstep(min, max, x)

smoothstep

sqrt(x)

sqrt

step(y, x)

step

tan(x)

tan

tanh(x)

tanh

transpose(x)

transpose

trunc(x)

trunc

Further Readings