I have some code at https://jsfiddle.net/72mnd2yt/1/ that doesn't display the sprite I'm trying to draw. I tried to follow the code over at https://stemkoski.github.io/Three.js/Sprite-Text-Labels.html and read it line by line, but I'm not sure where I went wrong. would someone mind taking a look at it?
Here is some relevant code:
// picture
var getPicture = function(message){
var canvas = document.createElement('canvas');
canvas.width = "100%";
canvas.height = "100%";
var context = canvas.getContext('2d');
context.font = "10px";
context.fillText(message, 0, 10);
var picture = canvas.toDataURL();
// return picture;
return canvas;
};
// let there be light and so forth...
var getScene = function(){
var scene = new THREE.Scene();
var camera = new THREE.PerspectiveCamera(
70,
$('body').width(),
$('body').height(),
1,
1000
);
var light = new THREE.PointLight(0xeeeeee);
scene.add(camera);
scene.add(light);
var renderer = new THREE.WebGLRenderer();
renderer.setClearColor(0xefefef);
renderer.setSize($('body').width(), $('body').height());
camera.position.z = -10;
camera.lookAt(new THREE.Vector3(0, 0, 0));
return [scene, renderer, camera];
};
// now for the meat
var getLabel = function(message){
var texture = new THREE.Texture(getPicture(message));
var spriteMaterial = new THREE.SpriteMaterial(
{map: texture }
);
var sprite = new THREE.Sprite(spriteMaterial);
//sprite.scale.set(100, 50, 1.0);
return sprite
};
var setup = function(){
var scene;
var renderer;
var camera;
[scene, renderer, camera] = getScene();
$('body').append(renderer.domElement);
var label = getLabel("Hello, World!");
scene.add(label);
var animate = function(){
requestAnimationFrame(animate);
renderer.render(scene, camera);
};
animate();
};
setup();
A few points:
1) canvas.width = "100%" should be canvas.width = "100" (canvas sizes are assumed to be px). Same with canvas.height.
2) $('body').height() is 0, so the renderer canvas is not visible (you can check this out in dev tools, it's in the element tree but squashed to 0px high). I know nothing about jQuery, so not sure why this is, but I would recommend using window.innerHeight and window.innerWidth instead anyways. So renderer.setSize(window.innerWidth, window.innerHeight). You'll also want to make this change in the camera initialization.
3) Speaking of the camera initialization, you are passing in width and height as separate arguments, when there should only be an aspect ratio argument. See the docs. So
var camera = new THREE.PerspectiveCamera(70, window.innerWidth, window.innerHeight, 1, 1000)
becomes
var camera = new THREE.PerspectiveCamera(70, window.innerWidth/window.innerHeight, 1, 1000)
4) Because textures are assumed to be static, you need to add something to this part:
var texture = new THREE.Texture(getPicture(message))
texture.needsUpdate = true // this signals that the texture has changed
That's a one-time flag, so you need to set it every time the canvas changes if you want a dynamic texture. You don't have to make a new THREE.Texture each time, just add texture.needsUpdate in the render loop (or in an event that only fires when you want the texture to change, if you're going for efficiency). See the docs, under .needsUpdate.
At this point, it should work. Here are some further things to consider:
5) Instead of using Texture you could use CanvasTexture, which sets .needsUpdate for you. The fiddle you posted is using Three.js r71, which doesn't have it, but newer versions do. That would look like this:
var texture = new THREE.CanvasTexture(getPicture(message));
// no needsUpdate necessary
6) It looks like you were on this path already based on the commented out return picture, but you can use either canvas element, or a data url generated from the canvas for a texture. If getPicture now returns a data url, try this:
var texture = new THREE.TextureLoader().load(getPicture(message))
You can also indirectly use a data url with Texture:
var img = document.createElement('img')
var img = new Image()
img.src = getPicture(message)
var texture = new THREE.Texture(img);
texture.needsUpdate = true
If not, just stick with Texture or CanvasTexture, both will take a canvas element. Texture can indirectly take a url by passing in an image element whose src is set to the data url.
Fiddle with the outlined fixes:
https://jsfiddle.net/jbjw/x0uL1kbh/
Related
I have a GLTF version 1.0 model that I am importing into Three.js using LegacyGLTFLoader.js. When I do so, everything looks good, except that the model does not receive shadows. I am guessing that this is because the imported model's material is THREE.RawShaderMaterial, which does not support receiving shadows (I think). How can I fix this so that my imported model can receive shadows?
Here is sample code:
// Construct scene.
var scene = new THREE.Scene();
// Get window dimensions.
var width = window.innerWidth;
var height = window.innerHeight;
// Construct camera.
var camera = new THREE.PerspectiveCamera(75, width/height);
camera.position.set(20, 20, 20);
camera.lookAt(scene.position);
// Construct renderer.
var renderer = new THREE.WebGLRenderer();
renderer.setSize(width, height);
renderer.shadowMapEnabled = true;
document.body.appendChild(renderer.domElement);
// Construct cube.
var cubeGeometry = new THREE.BoxGeometry(10, 1, 10);
var cubeMaterial = new THREE.MeshPhongMaterial({color: 0x00ff00});
var cube = new THREE.Mesh(cubeGeometry, cubeMaterial);
cube.castShadow = true;
cube.translateY(15);
scene.add(cube);
// Construct floor.
var floorGeometry = new THREE.BoxGeometry(20, 1, 20);
var floorMaterial = new THREE.MeshPhongMaterial({color: 0x00ffff});
var floor = new THREE.Mesh(floorGeometry, floorMaterial);
floor.receiveShadow = true;
scene.add(floor);
// Construct light.
var light = new THREE.DirectionalLight(0xffffff);
light.position.set(0, 20, 0);
light.castShadow = true;
scene.add(light);
// Construct light helper.
var lightHelper = new THREE.DirectionalLightHelper(light);
scene.add(lightHelper);
// Construct orbit controls.
new THREE.OrbitControls(camera, renderer.domElement);
// Construct GLTF loader.
var loader = new THREE.LegacyGLTFLoader();
// Load GLTF model.
loader.load(
"https://dl.dropboxusercontent.com/s/5piiujui3sdiaj3/1.glb",
function(event) {
var model = event.scene.children[0];
var mesh = model.children[0];
mesh.receiveShadow = true;
scene.add(model);
},
null,
function(event) {
alert("Loading model failed.");
}
);
// Animates the scene.
var animate = function () {
requestAnimationFrame(animate);
renderer.render(scene, camera);
};
// Animate the scene.
animate();
Here are my resources:
https://dl.dropboxusercontent.com/s/y2r8bsrppv0oqp4/three.js
https://dl.dropboxusercontent.com/s/5wh92lnsxz2ge1e/LegacyGLTFLoader.js
https://dl.dropboxusercontent.com/s/1jygy1eavetnp0d/OrbitControls.js
https://dl.dropboxusercontent.com/s/5piiujui3sdiaj3/1.glb
Here is a JSFiddle:
https://jsfiddle.net/rmilbert/8tqc3yx4/26/
One way to fix the problem is to replace the instance of RawShaderMaterial with MeshStandardMaterial. To get the intended effect, you have to apply the existing texture to the new material like so:
var newMaterial = new THREE.MeshStandardMaterial( { roughness: 1, metalness: 0 } );
newMaterial.map = child.material.uniforms.u_tex.value;
You also have to compute normal data for the respective geometry so lighting can be computed correctly. If you need no shadows, the unlint MeshBasicMaterial is actually the better choice.
Updated fiddle: https://jsfiddle.net/e67hbj1q/2/
/examples/js/postprocessing/OutlinePass.js from THREE.js r102 does not appear to work with skinned meshes. Specifically, the rendered outline always stays in the mesh's rest position.
Is there some way to get this working (that is, to update the outline to reflect the current pose of an animated mesh)? OutlinePass does not appear to be documented (mod the comments in the code itself).
Is there some other accepted method of outlining animated meshes? I'm in the process of migrating some code from r7x, where I ended up accomplishing this by manually creating a copy of the mesh and applying a shader material that scales along the normals. I can do that again, but if there's a simpler/better supported method to accomplish the same effect I'd rather use it instead of reproducing a method that breaks every new major release.
A simple jsfiddle illustrating the issue:
https://jsfiddle.net/L69pe5q2/3/
This is the code from the jsfiddle. The mesh I use is the SimpleSkinning.gltf example from the three.js distribution. In the jsfiddle I load it from a dataURI so it doesn't complain about XSS loading, and I've edited the base64-encoded data out (and replaced it with [FOO]) in the code below, purely for readability.
The OutlinePass is created and added to the composer in initComposer().
var camera, light, renderer, composer, mixer, loader, clock;
var scene, mesh, outlinePass;
var height = 480,
width = 640;
var clearColor = '#666666';
load();
function load() {
loader = new THREE.GLTFLoader();
clock = new THREE.Clock();
scene = new THREE.Scene();
loader.load('data:text/plain;base64,[FOO]', function(obj) {
scene.add(obj.scene);
mixer = new THREE.AnimationMixer(obj.scene);
var clip = THREE.AnimationClip.findByName(obj.animations,
'Take 01');
var a = mixer.clipAction(clip);
a.reset();
a.play();
mesh = obj.scene;
mesh.position.set(-7, 2.5, -7);
init();
animate();
});
}
function init() {
initCamera();
initScene();
initRenderer();
initComposer();
outlinePass.selectedObjects = [mesh];
}
function initCamera() {
camera = new THREE.PerspectiveCamera(30, width / height, 1, 10000);
camera.position.set(7, 0, 7);
camera.lookAt(0, 0, 0);
}
function initScene() {
light = new THREE.AmbientLight(0xffffff)
scene.add(light);
}
function initRenderer() {
renderer = new THREE.WebGLRenderer({
width: width,
height: height,
antialias: false,
});
renderer.setSize(width, height);
renderer.setClearColor(clearColor);
document.body.appendChild(renderer.domElement);
}
function initComposer() {
var renderPass, copyPass;
composer = new THREE.EffectComposer(renderer);
renderPass = new THREE.RenderPass(scene, camera);
composer.addPass(renderPass);
outlinePass = new THREE.OutlinePass(new THREE.Vector2(width, height),
scene, camera);
composer.addPass(outlinePass);
outlinePass.edgeStrength = 10;
outlinePass.edgeThickness = 4;
outlinePass.visibleEdgeColor.set('#ff0000');
copyPass = new THREE.ShaderPass(THREE.CopyShader);
copyPass.renderToScreen = true;
composer.addPass(copyPass);
}
function animate() {
var delta = clock.getDelta();
requestAnimationFrame(animate);
update(delta);
render(delta);
}
function update(delta) {
if (mixer) mixer.update(delta);
}
function render(delta) {
composer.render();
}
according to Mugen87 in Jan 2019 he said:
With this small patch, it's now possible to use the outline pass with animated meshes. The only thing users have to do at app level is to set morphTargets or skinning to true for OutlinePass.depthMaterial and OutlinePass.prepareMaskMaterial. That's of course still a manual effort but at least the more complicated shader enhancement is already done.
take this example:
https://jsfiddle.net/2ybks7rd/
reference link on github
I'm trying to have text sprites in the 3d scene with constant size (regardless of camera distance) using a PerspectiveCamera. In order to get non-sprites to have constant size, I make them children of a special "scaled" object which adjusts its scale as the camera distance to origin changes (see the code below). This works well to keep a general object roughly the same visual size, but when I add a sprite to the scaled object, the sprite seems to ignore its parent's scale (so it gets smaller and bigger as you zoom out and in).
Interestingly, when we switch to an orthographic camera (uncomment the appropriate line below), this special scaled object doesn't seem to affect children anymore (i.e., children don't stay a constant size). However, since we have an orthographic camera, sprites no longer scale as the camera distance changes (so they maintain a constant size), but this is independent of the scaled object.
I notice a few other similar questions and answers, including adjust the scale of the sprites themselves (it seems much easier to add all my sprites to a single scaling object), use an orthographic camera overlay to draw sprites (see also this) (but I want my sprites to be inside the 3d perspective scene).
So, my questions are: why do sprites not use scale according to their parent's scale when using a PerspectiveCamera? Also, why does my scaled object not work with the orthographic camera? Are these bugs or features of the cameras?
Thanks!
http://jsfiddle.net/LLbcs/8/
var camera, scene, renderer, geometry, material, mesh, text, controls;
init();
animate();
function init() {
scene = new THREE.Scene();
camera = new THREE.PerspectiveCamera(50, window.innerWidth / window.innerHeight, 1, 10000); var scenescale=1;
//camera = new THREE.OrthographicCamera( -7,7,7,-7, 1, 20 );
camera.position.z = 10;
scene.add(camera);
scaled=new THREE.Object3D();
scene.add(scaled);
var textmaterial = new THREE.SpriteMaterial( {color: 'red', useScreenCoordinates: true, map: texttexture("hi")});
text = new THREE.Sprite( textmaterial );
text.position.set( 1, 1, 0);
scaled.add(text);
var geometry = new THREE.BoxGeometry( 1, 1,1 );
var material = new THREE.MeshLambertMaterial( { color: 0xffffff } );
mesh = new THREE.Mesh( geometry, material );
mesh.position.set(0,3,0);
scaled.add(mesh);
var light = new THREE.PointLight('green');
light.position.set(10,15,10);
camera.add(light);
light = new THREE.PointLight(0x333333);
light.position.set(-10,-15,-8);
camera.add(light);
renderer = new THREE.WebGLRenderer();
controls = new THREE.OrbitControls( camera, renderer.domElement );
renderer.setSize(window.innerWidth, window.innerHeight);
document.body.appendChild(renderer.domElement);
}
function animate() {
requestAnimationFrame(animate);
var scale = camera.position.length()/10;
scaled.scale.set(scale,scale,scale);
render();
}
function render() {
renderer.render(scene, camera);
}
function texttexture(string) {
var fontFace = "Arial"
var size = "50";
var color = "white"
var squareTexture = true;
var canvas = document.createElement("canvas");
var context = canvas.getContext("2d");
canvas.height = size;
var font = "Normal " + size + "px " + fontFace;
context.font = font;
var metrics = context.measureText(string);
var textWidth = metrics.width;
canvas.width = textWidth;
if (squareTexture) {
canvas.height = canvas.width;
}
var aspect = canvas.width / canvas.height;
context.textAlign = "center";
context.textBaseline = "middle";
context.fillStyle = color;
// Must set the font again for the fillText call
context.font = font;
context.fillText(string, canvas.width / 2, canvas.height / 2);
var t = new THREE.Texture(canvas);
t.needsUpdate = true;
return t;
}
If you want text to appear over a 3D scene and you don't care if it is static, why not try layering a div over the scene instead?
This will allow you to save graphics bandwidth and memory, improving performance of your scene and give you much better flexibility over what you display. It's also much easier to do and to maintain.
I was playing with webGL and ThreeJS, then I've got the following issue:
Textures with large images gets pixelated when seen from distance.
Check the example: http://jsfiddle.net/4qTR3/1/
Below is the code:
var scene = new THREE.Scene();
var camera = new THREE.PerspectiveCamera(40, window.innerWidth / window.innerHeight, 10, 7000);
var light = new THREE.PointLight(0xffffff);
light.position.set(0, 150, 100);
scene.add(light);
var light2 = new THREE.AmbientLight(0x444444);
scene.add(light2);
var renderer = new THREE.WebGLRenderer({
antialias: true
});
renderer.setSize(window.innerWidth, window.innerHeight);
document.body.appendChild(renderer.domElement);
var geometry = new THREE.PlaneGeometry(500, 500, 10, 10);
//I use different textures in my project
var texture = new THREE.ImageUtils.loadTexture(TEST_IMAGE);
var textureBack = new THREE.ImageUtils.loadTexture(TEST_IMAGE);
textureBack.anisotropy = renderer.getMaxAnisotropy();
texture.anisotropy = renderer.getMaxAnisotropy();
//Filters
texture.magFilter = THREE.NearestFilter;
texture.minFilter = THREE.LinearMipMapLinearFilter;
textureBack.magFilter = THREE.NearestFilter;
textureBack.minFilter = THREE.LinearMipMapLinearFilter;
var materials = [
new THREE.MeshLambertMaterial({
transparent: true,
map: texture,
side: THREE.FrontSide
}),
new THREE.MeshLambertMaterial({
transparent: true,
map: textureBack,
side: THREE.BackSide
})];
for (var i = 0, len = geometry.faces.length; i < len; i++) {
var face = geometry.faces[i].clone();
face.materialIndex = 1;
geometry.faces.push(face);
geometry.faceVertexUvs[0].push(geometry.faceVertexUvs[0][i].slice(0));
}
planeObject = new THREE.Mesh(geometry, new THREE.MeshFaceMaterial(materials));
planeObject.overdraw = true;
planeObject.position.z = -5000;
scene.add(planeObject);
camera.position.z = 1000;
(function render() {
requestAnimationFrame(render);
planeObject.rotation.y += 0.02;
renderer.render(scene, camera);
})();
If the image of the texture has got text in it, the text becomes very pixelated with poor quality.
How can I fix this?
In order to not get pixelated you need to use mips but WebGL can't generate mips for non-power-of-2 textures. Your texture is 800x533, neither of those is a power of 2.
a couple of options
1) Scale the picture offline to powers of 2 like 512x512 or 1024x512
2) Scale the picture at runtime before making a texture.
Load the image yourself, once loaded make a canvas that is power-of-2. call drawImage(img, 0, 0, canvas.width, canvas.height) to scale the image into the canvas. Then load the canvas into a texture.
You also probably want to change your mag filtering from NearestFilter to LinearFilter.
Note: (1) is the better option. (2) takes time on the user's machine, uses more memory, and you have no guarantee what the quality of the scaling will be.
Example here.
// set the scene size
var WIDTH = 1650,
HEIGHT = 700;
// set some camera attributes
var VIEW_ANGLE = 100,
ASPECT = WIDTH / HEIGHT,
NEAR = 0.1,
FAR = 10000;
// get the DOM element to attach to
// - assume we've got jQuery to hand
var $container = $('#container');
// create a WebGL renderer, camera
// and a scene
var renderer = new THREE.WebGLRenderer();
var camera = new THREE.PerspectiveCamera( VIEW_ANGLE,
ASPECT,
NEAR,
FAR );
//camera.lookAt(new THREE.Vector3( 0, 0, 0 ));
var scene = new THREE.Scene();
// the camera starts at 0,0,0 so pull it back
camera.position.x = 200;
camera.position.y = 200;
camera.position.z = 300;
// start the renderer
renderer.setSize(WIDTH, HEIGHT);
// attach the render-supplied DOM element
$container.append(renderer.domElement);
// create the sphere's material
var sphereMaterial = new THREE.MeshLambertMaterial(
{
color: 0xCC0000
});
// set up the sphere vars
var radius = 60, segments = 20, rings = 20;
// create a new mesh with sphere geometry -
// we will cover the sphereMaterial next!
var sphere = new THREE.Mesh(
new THREE.SphereGeometry(radius, segments, rings),
img);
// add the sphere to the scene
scene.add(sphere);
// and the camera
scene.add(camera);
// create a point light
var pointLight = new THREE.PointLight( 0xFFFFFF );
// set its position
pointLight.position.x = 50;
pointLight.position.y = 100;
pointLight.position.z = 180;
// add to the scene
scene.add(pointLight);
// add a base plane
var planeGeo = new THREE.PlaneGeometry(500, 500,8, 8);
var planeMat = new THREE.MeshLambertMaterial({color: 0x666699});
var plane = new THREE.Mesh(planeGeo, planeMat);
plane.position.x = 160;
plane.position.y = 0;
plane.position.z = 20;
//rotate it to correct position
plane.rotation.x = -Math.PI/2;
scene.add(plane);
// add 3D img
var img = new THREE.MeshBasicMaterial({
map:THREE.ImageUtils.loadTexture('cube.png')
});
img.map.needsUpdate = true;
// draw!
renderer.render(scene, camera);
I've put the var img as a material to the sphere but everytime I render it aoutomaticly changes color...
How do I do so it just will have the image as I want... not all the colors?
I whould possibly put the img on a plane.
You need to use the callback of the loadTexture function. If you refer to the source of THREE.ImageUtils you will see that the loadTexture function is defined as
loadTexture: function ( url, mapping, onLoad, onError ) {
var image = new Image();
var texture = new THREE.Texture( image, mapping );
var loader = new THREE.ImageLoader();
loader.addEventListener( 'load', function ( event ) {
texture.image = event.content;
texture.needsUpdate = true;
if ( onLoad ) onLoad( texture );
} );
loader.addEventListener( 'error', function ( event ) {
if ( onError ) onError( event.message );
} );
loader.crossOrigin = this.crossOrigin;
loader.load( url, image );
return texture;
},
Here, only the url parameter is required. The texture mapping can be passed in as null. The last two parameters are the callbacks. You can pass in a function for each one which will be called at the respective load and error events. The problem you are experiencing is most likely caused by the fact that three.js is attempting to apply your material before your texture is properly loaded. In order to remedy this, you should only do this inside a function passed as the onLoad callback (which will only be called after the image has been loaded up). Also, you should always create your material before you apply it to an object.
So you should change your code from:
// set up the sphere vars
var radius = 60, segments = 20, rings = 20;
// create a new mesh with sphere geometry -
// we will cover the sphereMaterial next!
var sphere = new THREE.Mesh(
new THREE.SphereGeometry(radius, segments, rings),
img);
// add the sphere to the scene
scene.add(sphere);
to something like:
var sphereGeo, sphereTex, sphereMat, sphereMesh;
var radius = 60, segments = 20, rings = 20;
sphereGeo = new THREE.SphereGeometry(radius, segments, rings);
sphereTex = THREE.ImageUtils.loadTexture('cube.png', null, function () {
sphereMat = new THREE.MeshBasicMaterial({map: sphereTex});
sphereMesh = new THREE.Mesh(sphereGeo, sphereMat);
scene.add(sphereMesh);
});