Skip to content

Examples - Move Draw Indirect to TSL #32222

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Nov 9, 2025
Merged

Examples - Move Draw Indirect to TSL #32222

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
119 changes: 46 additions & 73 deletions examples/webgpu_struct_drawindirect.html
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -34,7 +34,7 @@
<script type="module">

import * as THREE from 'three/webgpu';
import { struct, storage, wgslFn, instanceIndex, time, varyingProperty, attribute } from 'three/tsl';
import { struct, storage, sin, cross, normalize, abs, mix, Fn, vec4, max, pow, time, varyingProperty, attribute, uint, atomicStore } from 'three/tsl';

import { OrbitControls } from 'three/addons/controls/OrbitControls.js';

Expand Down Expand Up @@ -138,110 +138,83 @@
offset: 'uint'
}, 'DrawBuffer' );

const writeDrawBuffer = wgslFn( `
fn compute(
index: u32,
drawBuffer: ptr<storage, DrawBuffer, read_write>,
instances: f32,
time: f32,
) -> void {
const drawStorage = storage( drawBuffer, drawBufferStruct, drawBuffer.count );

let instanceCount = max( instances * pow( sin( time * 0.5 ) + 1, 4.0 ), 100 );
computeDrawBuffer = Fn( () => {

atomicStore( &drawBuffer.instanceCount, u32( instanceCount ) );
}
` );
const halfTime = sin( time.mul( 0.5 ) );

computeDrawBuffer = writeDrawBuffer( {
drawBuffer: storage( drawBuffer, drawBufferStruct, drawBuffer.count ),
instances: instances,
index: instanceIndex,
time: time
} ).compute( instances ); // not necessary in this case but normally one wants to run through all instances
const instanceCount = max( ( pow( halfTime.add( 1 ), 4.0 ) ).mul( instances ), 100 ).toVar( 'instanceCount' );
atomicStore( drawStorage.get( 'instanceCount' ), instanceCount );

} )().compute( instances );

const initDrawBuffer = wgslFn( `
fn compute(
drawBuffer: ptr< storage, DrawBuffer, read_write >,
) -> void {
computeInitDrawBuffer = Fn( () => {

drawBuffer.vertexCount = 3u;
atomicStore(&drawBuffer.instanceCount, 0u);
drawBuffer.firstVertex = 0u;
drawBuffer.firstInstance = 0u;
drawBuffer.offset = 0u;
}
` );
const drawInfo = drawStorage;

drawInfo.get( 'vertexCount' ).assign( 3 );
atomicStore( drawInfo.get( 'instanceCount' ), uint( 0 ) );
drawInfo.get( 'firstVertex' ).assign( 0 );
drawInfo.get( 'firstInstance' ).assign( 0 );
drawInfo.get( 'offset' ).assign( 0 );

computeInitDrawBuffer = initDrawBuffer( {
drawBuffer: storage( drawBuffer, drawBufferStruct, drawBuffer.count ),
} ).compute( 1 );
} )().compute( 1 );


const vPosition = varyingProperty( 'vec3', 'vPosition' );
const vColor = varyingProperty( 'vec4', 'vColor' );
const vPosition = varyingProperty( 'vec3', 'vPosition' );
const vColor = varyingProperty( 'vec4', 'vColor' );

const positionShaderParams = {
const positionShaderParams = {
position: attribute( 'position' ),
offset: attribute( 'offset' ),
color: attribute( 'color' ),
orientationStart: attribute( 'orientationStart' ),
orientationEnd: attribute( 'orientationEnd' ),
time: time
};


const positionFn = Fn( () => {

const positionShader = wgslFn( `
fn main_vertex(
position: vec3<f32>,
offset: vec3<f32>,
color: vec4<f32>,
orientationStart: vec4<f32>,
orientationEnd: vec4<f32>,
time: f32
) -> vec4<f32> {
const { position, offset, color, orientationStart, orientationEnd } = positionShaderParams;

var vPosition = offset * max( abs( sin( time * 0.5 ) * 2.0 + 1.0 ), 0.5 ) + position;
var orientation = normalize( mix( orientationStart, orientationEnd, sin( time * 0.5 ) ) );
var vcV = cross( orientation.xyz, vPosition );
vPosition = vcV * ( 2.0 * orientation.w ) + ( cross( orientation.xyz, vcV ) * 2.0 + vPosition );
const halfTime = sin( time.mul( 0.5 ) );

var vColor = color;
// Convert slowed sign range of (-1 to 1) to range of (1 -> 0 / 0.5 -> 3)
const oscilationRange = max( abs( halfTime.mul( 2.0 ).add( 1.0 ) ), 0.5 );

var outPosition = vec4f(vPosition, 1);
const sphereOscilation = offset.mul( oscilationRange ).add( position ).toVar();

varyings.vPosition = vPosition;
varyings.vColor = vColor;
const orientation = normalize( mix( orientationStart, orientationEnd, halfTime ) );
const vcV = cross( orientation.xyz, sphereOscilation );
const crossvcV = cross( orientation.xyz, vcV );

return outPosition;
}
`, [ vPosition, vColor ] );
vPosition.assign( vcV.mul( orientation.w.mul( 2.0 ) ).add( crossvcV.mul( 2.0 ).add( sphereOscilation ) ) );
vColor.assign( color );

const fragmentShaderParams = {
time: time,
vPosition: vPosition,
vColor: vColor
};
return vPosition;

} )();

const fragmentFn = Fn( () => {

const color = vec4( vColor ).toVar();

const fragmentShader = wgslFn( `
fn main_fragment(
time: f32,
vPosition: vec3<f32>,
vColor: vec4<f32>
) -> vec4<f32> {
color.r.addAssign( sin( vPosition.x.mul( 10.0 ).add( time ) ).mul( 0.5 ) );

var color = vec4f( vColor );
color.r += sin( vPosition.x * 10.0 + time ) * 0.5;
return color;

return color;
}
` );
} )();

const material = new THREE.MeshBasicNodeMaterial( {
side: THREE.DoubleSide,
forceSinglePass: true,
transparent: true
} );

material.positionNode = positionShader( positionShaderParams );
material.fragmentNode = fragmentShader( fragmentShaderParams );
material.positionNode = positionFn;
material.fragmentNode = fragmentFn;

const mesh = new THREE.Mesh( geometry, material );
scene.add( mesh );
Expand Down