First off, I'm new at WebGL. I am trying to applying multiple fragment shaders(here, 2 shaders) on a single image to be rendered. I read in different articles and other stack overflow questions that we should use framebuffers (ping pong method) for this purpose but couldn't find any sample code snippets anywhere. What I understood is first I create two programs each with a different fragment shader. Then use a framebuffer object where I can use my first program (first shader) on the original image and output it to that fbo texture. Then use this output texture as the input in the second program so that both shaders are retained. This output is finally rendered on the canvas.
I tried doing the same thing , but my canvas is completely black. I am not getting any errors in the console everything seems fine but not result.
I am struck with for hours . Could anyone help me check it?
Below is the code I wrote
const canvas = document.querySelector("canvas")
const gl = canvas.getContext("webgl");
//create two programs using a createprogram function written in my code.
const programA = createProgram(gl, vertexShader, fragmentShaderA); // program using #shader1
const programB = createProgram(gl, vertexShader, fragmentShaderB);
const texFbPair = createTextureAndFramebuffer(gl); //function defined below
setAttributes(programA);
setAttributes(programB);
function setAttributes(program) {
const positionLocation = gl.getAttribLocation(program, 'position');
const positionBuffer = gl.createBuffer();
gl.bindBuffer(gl.ARRAY_BUFFER, positionBuffer);
gl.bufferData(gl.ARRAY_BUFFER, new Float32Array([
-1, -1, -1, 1, 1, -1,
1, 1, 1, -1, -1, 1,
]), gl.STATIC_DRAW);
gl.enableVertexAttribArray(positionLocation);
gl.vertexAttribPointer(positionLocation, 2, gl.FLOAT, false, 0, 0);
const texCoordLocation = gl.getAttribLocation(program, "a_texCoord");
const texCoordBuffer = gl.createBuffer();
gl.bindBuffer(gl.ARRAY_BUFFER, texCoordBuffer);
gl.bufferData(gl.ARRAY_BUFFER, new Float32Array([
0.0, 1.0,
0.0, 0.0,
1.0, 1.0,
1.0, 0.0,
1.0, 1.0,
0.0, 0.0]), gl.STATIC_DRAW);
gl.enableVertexAttribArray(texCoordLocation);
gl.vertexAttribPointer(texCoordLocation, 2, gl.FLOAT, false, 0, 0);
}
const texture = gl.createTexture();
texture.image = new Image();
texture.image.onload = function () {
handleLoadedTexture(gl, texture);
};
texture.image.crossOrigin = '';
texture.image.src = 'skogafoss_waterfall_iceland.jpg';
function handleLoadedTexture(gl, texture, callback) {
gl.bindTexture(gl.TEXTURE_2D, texture);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_S, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_T, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MAG_FILTER, gl.NEAREST);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MIN_FILTER, gl.NEAREST);
gl.texImage2D(gl.TEXTURE_2D, 0, gl.RGBA, gl.RGBA, gl.UNSIGNED_BYTE, texture.image);
}
gl.useProgram(programA);
gl.bindFramebuffer(gl.FRAMEBUFFER, texFbPair.fb);
gl.bindTexture(gl.TEXTURE_2D, texture);
gl.clearColor(0, 0, 1, 1);
gl.viewport(0, 0, gl.canvas.width, gl.canvas.height);
gl.drawArrays(gl.TRIANGLES, 0, 6);
gl.useProgram(programB);
gl.bindFramebuffer(gl.FRAMEBUFFER, null);
gl.bindTexture(gl.TEXTURE_2D, texFbPair.tex);
gl.clearColor(0, 0, 0, 1);
gl.viewport(0, 0, gl.canvas.width, gl.canvas.height);
gl.drawArrays(gl.TRIANGLES, 0, 6)
function createTextureAndFramebuffer(gl) {
const tex = gl.createTexture();
gl.bindTexture(gl.TEXTURE_2D, tex);
gl.texImage2D(gl.TEXTURE_2D, 0, gl.RGBA, canvas.width, canvas.height, 0, gl.RGBA, gl.UNSIGNED_BYTE, null);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MIN_FILTER, gl.LINEAR);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_S, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_T, gl.CLAMP_TO_EDGE);
const fb = gl.createFramebuffer();
gl.bindFramebuffer(gl.FRAMEBUFFER, fb);
gl.framebufferTexture2D( gl.FRAMEBUFFER, gl.COLOR_ATTACHMENT0, gl.TEXTURE_2D, tex, 0);
return { tex: tex, fb: fb };
}
It looks like from your code you have a mis-understanding of how attributes work. Attributes are global state in WebGL1 so these lines
setAttributes(programA);
setAttributes(programB);
Won't work. The second call to setAttributes will just change the global attributes to the second call's settings.
See this and this
The next issue is the code does not wait for the image to load so it creates an image, sets a callback for when it finishes loading, it then draws 2 things. Then later, the image finishes loading and is copied to the texture but no drawing happens after that.
The code also never allocates the actual texture in createTextureAndFramebuffer
To do that you to call gl.texImage2D
Here is some working code.
const vertexShader = `
attribute vec4 position;
attribute vec2 a_texCoord;
varying vec2 v_texCoord;
void main() {
gl_Position = position;
v_texCoord = a_texCoord;
}
`;
const fragmentShaderA = `
precision highp float;
varying vec2 v_texCoord;
uniform sampler2D tex;
void main() {
gl_FragColor = texture2D(tex, v_texCoord);
}
`;
const fragmentShaderB = `
precision highp float;
varying vec2 v_texCoord;
uniform sampler2D tex;
void main() {
gl_FragColor = texture2D(tex, v_texCoord);
}
`;
const canvas = document.querySelector("canvas")
const gl = canvas.getContext("webgl");
//create two programs using a createprogram function written in my code.
const programA = createProgram(gl, vertexShader, fragmentShaderA); // program using #shader1
const programB = createProgram(gl, vertexShader, fragmentShaderB);
const texFbPair = createTextureAndFramebuffer(gl); //function defined below
function setAttributes(program) {
const positionLocation = gl.getAttribLocation(program, 'position');
const positionBuffer = gl.createBuffer();
gl.bindBuffer(gl.ARRAY_BUFFER, positionBuffer);
gl.bufferData(gl.ARRAY_BUFFER, new Float32Array([
-1, -1, -1, 1, 1, -1,
1, 1, 1, -1, -1, 1,
]), gl.STATIC_DRAW);
gl.enableVertexAttribArray(positionLocation);
gl.vertexAttribPointer(positionLocation, 2, gl.FLOAT, false, 0, 0);
const texCoordLocation = gl.getAttribLocation(program, "a_texCoord");
const texCoordBuffer = gl.createBuffer();
gl.bindBuffer(gl.ARRAY_BUFFER, texCoordBuffer);
gl.bufferData(gl.ARRAY_BUFFER, new Float32Array([
0.0, 1.0,
0.0, 0.0,
1.0, 1.0,
1.0, 0.0,
1.0, 1.0,
0.0, 0.0]), gl.STATIC_DRAW);
gl.enableVertexAttribArray(texCoordLocation);
gl.vertexAttribPointer(texCoordLocation, 2, gl.FLOAT, false, 0, 0);
}
const texture = gl.createTexture();
texture.image = new Image();
texture.image.onload = function () {
handleLoadedTexture(gl, texture);
};
texture.image.crossOrigin = '';
texture.image.src = 'https://i.imgur.com/ZKMnXce.png';
function handleLoadedTexture(gl, texture, callback) {
gl.bindTexture(gl.TEXTURE_2D, texture);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_S, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_T, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MAG_FILTER, gl.NEAREST);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MIN_FILTER, gl.NEAREST);
gl.texImage2D(gl.TEXTURE_2D, 0, gl.RGBA, gl.RGBA, gl.UNSIGNED_BYTE, texture.image);
setAttributes(programA);
gl.useProgram(programA);
gl.bindFramebuffer(gl.FRAMEBUFFER, texFbPair.fb);
gl.bindTexture(gl.TEXTURE_2D, texture);
gl.clearColor(0, 0, 1, 1);
gl.viewport(0, 0, gl.canvas.width, gl.canvas.height);
gl.drawArrays(gl.TRIANGLES, 0, 6);
setAttributes(programB);
gl.useProgram(programB);
gl.bindFramebuffer(gl.FRAMEBUFFER, null);
gl.bindTexture(gl.TEXTURE_2D, texFbPair.tex);
gl.clearColor(0, 0, 0, 1);
gl.viewport(0, 0, gl.canvas.width, gl.canvas.height);
gl.drawArrays(gl.TRIANGLES, 0, 6)}
function createTextureAndFramebuffer(gl) {
const tex = gl.createTexture();
gl.bindTexture(gl.TEXTURE_2D, tex);
gl.texImage2D(
gl.TEXTURE_2D,
0, // mip level
gl.RGBA, // internal format
gl.canvas.width, // width
gl.canvas.height, // height
0, // border
gl.RGBA, // format
gl.UNSIGNED_BYTE, // type
null); // data
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MIN_FILTER, gl.LINEAR);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_S, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_T, gl.CLAMP_TO_EDGE);
const fb = gl.createFramebuffer();
gl.bindFramebuffer(gl.FRAMEBUFFER, fb);
gl.framebufferTexture2D( gl.FRAMEBUFFER, gl.COLOR_ATTACHMENT0, gl.TEXTURE_2D, tex, 0);
return { tex: tex, fb: fb };
}
function createProgram(gl, vs, fs) {
return twgl.createProgram(gl, [vs, fs]);
}
<script src="https://twgljs.org/dist/4.x/twgl.min.js"></script>
<canvas></canvas>
Other things I noticed.
The uniforms for the samplers are never looked up (of course I don't know what your actual shaders look like, I used placeholders). As such it works because uniforms default to 0 so the programs will reference the texture bound to texture unit 0 which is the default.
I also noticed the first draw via the framebuffer sets the viewport to the size of the canvas. That's correct if the size of the attachments in the framebuffer are the size of the canvas (which I made them be when I added the call to texImage2D) but it would probably be more appropriate to record a width and height for that texture so if you change its size the code won't fail.
Finally while I moved the calls to setAttributes to the correct places it's not common to also create and fill out buffers when rendering. It's more common to create buffers at init time and set attributes at render time but I didn't want to change more code.
You might find these tutorials helpful.
We're working in Webgl 1 and attempting to render to a cubemap using a stencil. Rendering to the cubemap on its own works fine. When we add a DEPTH_STENCIL renderbuffer it stops writing to the cubemap and issues no error.
This doesn't happen with a normal TEXTURE_2D instead of a TEXTURE_CUBE_MAP.
Depth/stencil/scissor tests are disabled.
The call to framebufferRenderbuffer is what breaks it.
Switching the renderbuffer to be either just a stencil or just a depth has the same effect.
Switching the renderbuffer to be a colour buffer makes it work again.
Here's a minimal-ish recreation. As you can see, we're getting a console output with the correct values for the first three calls and zeroes for the last call.
Why is this happening and what little thing are we missing to make renderbuffers work with cubemaps?
const canvas = document.createElement("canvas");
const gl = canvas.getContext("webgl");
console.log(TEST(false, false));
console.log(TEST(false, true));
console.log(TEST(true, false));
console.log(TEST(true, true));
function TEST(useCubemap, useBuffer) {
const size = 512;
const textureType = useCubemap ? gl.TEXTURE_CUBE_MAP : gl.TEXTURE_2D;
// SETUP THE PROGRAM
{
const program = gl.createProgram();
const vertShader = gl.createShader(gl.VERTEX_SHADER);
const fragShader = gl.createShader(gl.FRAGMENT_SHADER);
gl.shaderSource(vertShader, `
attribute vec2 a_position;
void main() {
gl_Position = vec4(a_position, 0.2, 1.0);
}
`);
gl.compileShader(vertShader);
gl.attachShader(program, vertShader);
gl.shaderSource(fragShader, `
void main() {
gl_FragColor = vec4(0.1, 0.2, 0.3, 0.4);
}
`);
gl.compileShader(fragShader);
gl.attachShader(program, fragShader);
gl.linkProgram(program);
gl.useProgram(program);
}
// SETUP THE QUAD
{
const posBuffer = gl.createBuffer();
gl.bindBuffer(gl.ARRAY_BUFFER, posBuffer);
gl.enableVertexAttribArray(0);
gl.vertexAttribPointer(0, 2, gl.FLOAT, false, 0, 0);
gl.bufferData(gl.ARRAY_BUFFER, new Float32Array([-1, +1, -1, -1, +1, +1, +1, -1]), gl.STATIC_DRAW);
}
// SETUP THE FRAMEBUFFER
{
const fb = gl.createFramebuffer();
const targetTexture = gl.createTexture();
gl.bindFramebuffer(gl.FRAMEBUFFER, fb);
gl.bindTexture(textureType, targetTexture);
gl.texParameteri(textureType, gl.TEXTURE_MIN_FILTER, gl.NEAREST);
gl.texParameteri(textureType, gl.TEXTURE_MAG_FILTER, gl.NEAREST);
gl.texParameteri(textureType, gl.TEXTURE_WRAP_S, gl.CLAMP_TO_EDGE);
gl.texParameteri(textureType, gl.TEXTURE_WRAP_T, gl.CLAMP_TO_EDGE);
// SWITCH TEXTURE TYPE
if (textureType === gl.TEXTURE_2D) {
gl.texImage2D(gl.TEXTURE_2D, 0, gl.RGBA, size, size, 0, gl.RGBA, gl.UNSIGNED_BYTE, null);
gl.framebufferTexture2D(gl.FRAMEBUFFER, gl.COLOR_ATTACHMENT0, gl.TEXTURE_2D, targetTexture, 0);
} else {
for (let i = 0; i < 6; i++) gl.texImage2D(gl.TEXTURE_CUBE_MAP_POSITIVE_X + i, 0, gl.RGBA, size, size, 0, gl.RGBA, gl.UNSIGNED_BYTE, null);
gl.framebufferTexture2D(gl.FRAMEBUFFER, gl.COLOR_ATTACHMENT0, gl.TEXTURE_CUBE_MAP_POSITIVE_X, targetTexture, 0);
}
}
// SETUP THE RENDER BUFFER
{
const rb = gl.createRenderbuffer();
gl.bindRenderbuffer(gl.RENDERBUFFER, rb);
gl.renderbufferStorage(gl.RENDERBUFFER, gl.DEPTH_STENCIL, size, size);
// TAKING THIS OUT MAKES IT WORK
if (useBuffer) gl.framebufferRenderbuffer(gl.FRAMEBUFFER, gl.DEPTH_STENCIL_ATTACHMENT, gl.RENDERBUFFER, rb);
}
// DISABLE THE OBVIOUS CULPRITS
gl.disable(gl.DEPTH_TEST);
gl.disable(gl.STENCIL_TEST);
gl.disable(gl.SCISSOR_TEST);
// DO A RENDERYFUCK
gl.viewport(0, 0, size, size);
gl.drawArrays(gl.TRIANGLE_STRIP, 0, 4);
// GET THE OUTFUCK
const pixels = new Uint8Array(4);
gl.readPixels(0, 0, 1, 1, gl.RGBA, gl.UNSIGNED_BYTE, pixels);
return pixels;
}
It works for me. I get the same values for all 4 calls in the code you posted. What OS/GPU/Driver are you using? Can you pastebin your about:gpu contents if you're using Chrome?
It sounds like a bug in your drivers. It's possible it's a also a bug in the WebGL spec.
The OpenGL ES spec does not require any combinations of framebuffer attachments to work (zero, zilch, nada). The WebGL spec require 3 combinations to work. From the spec, section 6.8:
The following combinations of framebuffer object attachments, when all of the attachments are framebuffer attachment complete, non-zero, and have the same width and height, must result in the framebuffer being framebuffer complete:
COLOR_ATTACHMENT0 = RGBA/UNSIGNED_BYTE texture
COLOR_ATTACHMENT0 = RGBA/UNSIGNED_BYTE texture + DEPTH_ATTACHMENT = DEPTH_COMPONENT16 renderbuffer
COLOR_ATTACHMENT0 = RGBA/UNSIGNED_BYTE texture + DEPTH_STENCIL_ATTACHMENT = DEPTH_STENCIL renderbuffer
But looking at the WebGL Conformance tests only TEXTURE_2D is tested.
So, first off that suggests your driver/gpu doesn't support that combination with cubemaps. Test by calling gl.checkFramebufferStatus. If it doesn't return gl.FRAMEBUFFER_COMPLETE your setup doesn't support rendering to a cubemap with a depth stencil attachment.
const canvas = document.createElement("canvas");
const gl = canvas.getContext("webgl");
TEST("TEXTURE_2D", "DEPTH_COMPONENT16");
TEST("TEXTURE_2D", "DEPTH_STENCIL");
TEST("TEXTURE_CUBE_MAP", "DEPTH_COMPONENT16");
TEST("TEXTURE_CUBE_MAP", "DEPTH_STENCIL");
function TEST(target, depthBufferFormat) {
const size = 16;
const textureType = gl[target];
// SETUP THE FRAMEBUFFER
{
const fb = gl.createFramebuffer();
const targetTexture = gl.createTexture();
gl.bindFramebuffer(gl.FRAMEBUFFER, fb);
gl.bindTexture(textureType, targetTexture);
gl.texParameteri(textureType, gl.TEXTURE_MIN_FILTER, gl.NEAREST);
gl.texParameteri(textureType, gl.TEXTURE_MAG_FILTER, gl.NEAREST);
gl.texParameteri(textureType, gl.TEXTURE_WRAP_S, gl.CLAMP_TO_EDGE);
gl.texParameteri(textureType, gl.TEXTURE_WRAP_T, gl.CLAMP_TO_EDGE);
// SWITCH TEXTURE TYPE
if (textureType === gl.TEXTURE_2D) {
gl.texImage2D(gl.TEXTURE_2D, 0, gl.RGBA, size, size, 0, gl.RGBA, gl.UNSIGNED_BYTE, null);
gl.framebufferTexture2D(gl.FRAMEBUFFER, gl.COLOR_ATTACHMENT0, gl.TEXTURE_2D, targetTexture, 0);
} else {
for (let i = 0; i < 6; i++) gl.texImage2D(gl.TEXTURE_CUBE_MAP_POSITIVE_X + i, 0, gl.RGBA, size, size, 0, gl.RGBA, gl.UNSIGNED_BYTE, null);
gl.framebufferTexture2D(gl.FRAMEBUFFER, gl.COLOR_ATTACHMENT0, gl.TEXTURE_CUBE_MAP_POSITIVE_X, targetTexture, 0);
}
}
// SETUP THE RENDER BUFFER
{
const rb = gl.createRenderbuffer();
const format = gl[depthBufferFormat];
gl.bindRenderbuffer(gl.RENDERBUFFER, rb);
gl.renderbufferStorage(gl.RENDERBUFFER, format, size, size);
// TAKING THIS OUT MAKES IT WORK
const attachmentPoint = depthBufferFormat === "DEPTH_COMPONENT16"
? gl.DEPTH_ATTACHMENT
: gl.DEPTH_STENCIL_ATTACHMENT;
gl.framebufferRenderbuffer(gl.FRAMEBUFFER, attachmentPoint, gl.RENDERBUFFER, rb);
}
const success = gl.checkFramebufferStatus(gl.FRAMEBUFFER) === gl.FRAMEBUFFER_COMPLETE;
console.log(target, depthBufferFormat, success ? "PASS" : "**FAIL**");
}
Do you need the stencil or just the depth buffer? Does this sample run for you? It's using a DEPTH_COMPONENT16 attachment.
I am trying to draw image to webgl canvas from a 2d canvas.
If I use:
gl.texImage2D(gl.TEXTURE_2D, 0, gl.RGBA, gl.RGBA, gl.UNSIGNED_BYTE, image);
, it works and renders the image successfully, but if I use :
gl.texImage2D(gl.TEXTURE_2D, 0, gl.RGBA, c.width, c.height, 0, gl.RGBA, gl.UNSIGNED_BYTE, dataTypedArray);
, it just shows a black screen.
Here's my Code :
Vertex Shader
attribute vec2 a_position;
uniform vec2 u_resolution;
uniform mat3 u_matrix;
varying vec2 v_texCoord;
void main() {
gl_Position = vec4(u_matrix * vec3(a_position, 1), 1);
v_texCoord = a_position;
}
Fragment Shader
precision mediump float;
// our texture
uniform sampler2D u_image;
// the texCoords passed in from the vertex shader.
varying vec2 v_texCoord;
void main() {
gl_FragColor = texture2D(u_image, v_texCoord);
}
Javascript
window.onload = main;
var buffer = null;
function main() {
var image = new Image();
image.src = "images/GL.jpg"
image.onload = function() {
render(image);
}
}
function render(image) {
var c = document.getElementById("c");
c.width = window.innerWidth*0.90;
c.height = window.innerHeight*0.90;
var context = c.getContext('2d');
context.drawImage(image, 0, 0);
var imageData = context.getImageData(0,0,image.width,image.height);
buffer = imageData.data.buffer; // ArrayBuffer
var canvas = document.getElementById("canvas");
canvas.width = window.innerWidth*0.90;
canvas.height = window.innerHeight*0.90;
//Get A WebGL context
var gl = getWebGLContext(canvas);
if (!gl) {
return;
}
// setup GLSL program
var program = createProgramFromScripts(gl, ["2d-vertex-shader", "2d-fragment-shader"]);
gl.useProgram(program);
// look up where the vertex data needs to go.
var positionLocation = gl.getAttribLocation(program, "a_position");
// look up uniform locations
var u_imageLoc = gl.getUniformLocation(program, "u_image");
var u_matrixLoc = gl.getUniformLocation(program, "u_matrix");
// provide texture coordinates for the rectangle.
var positionBuffer = gl.createBuffer();
gl.bindBuffer(gl.ARRAY_BUFFER, positionBuffer);
gl.bufferData(gl.ARRAY_BUFFER, new Float32Array([
0.0, 0.0,
1.0, 0.0,
0.0, 1.0,
0.0, 1.0,
1.0, 0.0,
1.0, 1.0]), gl.STATIC_DRAW);
gl.enableVertexAttribArray(positionLocation);
gl.vertexAttribPointer(positionLocation, 2, gl.FLOAT, false, 0, 0);
var texture = gl.createTexture();
gl.bindTexture(gl.TEXTURE_2D, texture);
// Set the parameters so we can render any size image.
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_S, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_T, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MIN_FILTER, gl.NEAREST);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MAG_FILTER, gl.NEAREST);
// Upload the image into the texture.
var dataTypedArray = new Uint8Array(buffer);
//gl.texImage2D(gl.TEXTURE_2D, 0, gl.RGBA, gl.RGBA, gl.UNSIGNED_BYTE, image);
textureFromPixelArray(gl, buffer, gl.RGBA, canvas.width, canvas.height);
var dstX = 20;
var dstY = 30;
var dstWidth = canvas.width;
var dstHeight = canvas.height;
// convert dst pixel coords to clipspace coords
var clipX = dstX / gl.canvas.width * 2 - 1;
var clipY = dstY / gl.canvas.height * -2 + 1;
var clipWidth = dstWidth / gl.canvas.width * 2;
var clipHeight = dstHeight / gl.canvas.height * -2;
// build a matrix that will stretch our
// unit quad to our desired size and location
gl.uniformMatrix3fv(u_matrixLoc, false, [
clipWidth, 0, 0,
0, clipHeight, 0,
clipX, clipY, 1,
]);
// Draw the rectangle.
gl.drawArrays(gl.TRIANGLES, 0, 6);
}
function textureFromPixelArray(gl, dataArray, type, width, height) {
var dataTypedArray = new Uint8Array(dataArray); // Don't need to do this if the data is already in a typed array
var texture = gl.createTexture();
gl.bindTexture(gl.TEXTURE_2D, texture);
gl.texImage2D(gl.TEXTURE_2D, 0, type, width, height, 0, type, gl.UNSIGNED_BYTE, dataTypedArray);
// Other texture setup here, like filter modes and mipmap generation
console.log(dataTypedArray);
return texture;
}
So first off, you can pass a canvas directly to gl.texImage2D. There's no good reason to first call ctx.getImageData and get the data out. Just call
gl.texImage2D(gl.TEXTURE_2D, 0, gl.RGBA, gl.RGBA, gl.UNSIGNED_BYTE, someCanvas);
Second, looking at your code you first create a texture, then set filtering. You then call textureFromPixelArray, WHICH CREATES A NEW TEXTURE, that texture does not have filtering set so if it's not a power-of-2 then it won't render. Just a guess but did you check the JavaScript console? I'm just guessing it probably printed a warning about your texture not being renderable.
On top of that, even though textureFromPixelArray creates a new texture the code ignores the return value.
To make the code work as is I think you want to change it to this
// not needed -- var texture = gl.createTexture();
// not needed -- gl.bindTexture(gl.TEXTURE_2D, texture);
// moved from below
// Upload the image into the texture.
var dataTypedArray = new Uint8Array(buffer);
//gl.texImage2D(gl.TEXTURE_2D, 0, gl.RGBA, gl.RGBA, gl.UNSIGNED_BYTE, image);
var texture = textureFromPixelArray(gl, buffer, gl.RGBA, canvas.width, canvas.height);
// Set the parameters so we can render any size image.
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_S, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_T, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MIN_FILTER, gl.NEAREST);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MAG_FILTER, gl.NEAREST);
Currently, I'm using 2D canvas context to draw an image generated (from pixel to pixel, but refreshed as a whole buffer in once after a generated frame) from JavaScript at about a 25fps rate. The generated image is always one byte (integer / typed array) per pixel and a fixed palette is used to generate RGB final result. Scaling is also needed to adopt to the size of the canvas (ie: going to fullscreen) and/or at user request (zoom in/out buttons).
The 2D context of canvas is OK for this purpose, however I'm curious if WebGL can provide better result and/or better performance. Please note: I don't want to put pixels via webGL, I want to put pixels into my buffer (which is basically Uint8Array), and use that buffer (in once) to refresh the context. I don't know too much about WebGL, but using the needed generated image as some kind of texture would work somehow for example? Then I would need to refresh the texture at about 25fps rate, I guess.
It would be really fantastic, if WebGL support the colour space conversion somehow. With 2D context, I need to convert 1 byte / pixel buffer into RGBA for the imagedata in JavaScript for every pixel ... Scaling (for 2D context) is done now by altering the height/width style of the canvas, so browsers scales the image then. However I guess it can be slower than what WebGL can do with hw support, and also (I hope) WebGL can give greater flexibility to control the scaling, eg with the 2D context, browsers will do antialiasing even if I don't want to do (eg: integer zooming factor), and maybe that's a reason it can be quite slow sometimes.
I've already tried to learn several WebGL tutorials but all of them starts with objects, shapes, 3D cubes, etc, I don't need any - classical - object to render only what 2D context can do as well - in the hope that WebGL can be a faster solution for the very same task! Of course if there is no win here with WebGL at all, I would continue to use 2D context.
To be clear: this is some kind of computer hardware emulator done in JavaScript, and its output (what would be seen on a PAL TV connected to it) is rendered via a canvas context. The machine has fixed palette with 256 elements, internally it only needs one byte for a pixel to define its colour.
You can use a texture as your palette and a different texture as your image. You then get a value from the image texture and use it too look up a color from the palette texture.
The palette texture is 256x1 RGBA pixels. Your image texture is any size you want but just a single channel ALPHA texture. You can then look up a value from the image
float index = texture2D(u_image, v_texcoord).a * 255.0;
And use that value to look up a color in the palette
gl_FragColor = texture2D(u_palette, vec2((index + 0.5) / 256.0, 0.5));
Your shaders might be something like this
Vertex Shader
attribute vec4 a_position;
varying vec2 v_texcoord;
void main() {
gl_Position = a_position;
// assuming a unit quad for position we
// can just use that for texcoords. Flip Y though so we get the top at 0
v_texcoord = a_position.xy * vec2(0.5, -0.5) + 0.5;
}
Fragment shader
precision mediump float;
varying vec2 v_texcoord;
uniform sampler2D u_image;
uniform sampler2D u_palette;
void main() {
float index = texture2D(u_image, v_texcoord).a * 255.0;
gl_FragColor = texture2D(u_palette, vec2((index + 0.5) / 256.0, 0.5));
}
Then you just need a palette texture.
// Setup a palette.
var palette = new Uint8Array(256 * 4);
// I'm lazy so just setting 4 colors in palette
function setPalette(index, r, g, b, a) {
palette[index * 4 + 0] = r;
palette[index * 4 + 1] = g;
palette[index * 4 + 2] = b;
palette[index * 4 + 3] = a;
}
setPalette(1, 255, 0, 0, 255); // red
setPalette(2, 0, 255, 0, 255); // green
setPalette(3, 0, 0, 255, 255); // blue
// upload palette
...
gl.texImage2D(gl.TEXTURE_2D, 0, gl.RGBA, 256, 1, 0, gl.RGBA,
gl.UNSIGNED_BYTE, palette);
And your image. It's an alpha only image so just 1 channel.
// Make image. Just going to make something 8x8
var image = new Uint8Array([
0,0,1,1,1,1,0,0,
0,1,0,0,0,0,1,0,
1,0,0,0,0,0,0,1,
1,0,2,0,0,2,0,1,
1,0,0,0,0,0,0,1,
1,0,3,3,3,3,0,1,
0,1,0,0,0,0,1,0,
0,0,1,1,1,1,0,0,
]);
// upload image
....
gl.texImage2D(gl.TEXTURE_2D, 0, gl.ALPHA, 8, 8, 0, gl.ALPHA,
gl.UNSIGNED_BYTE, image);
You also need to make sure both textures are using gl.NEAREST for filtering since one represents indices and the other a palette and filtering between values in those cases makes no sense.
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MIN_FILTER, gl.NEAREST);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MAG_FILTER, gl.NEAREST);
Here's a working example:
var canvas = document.getElementById("c");
var gl = canvas.getContext("webgl");
// Note: createProgramFromScripts will call bindAttribLocation
// based on the index of the attibute names we pass to it.
var program = twgl.createProgramFromScripts(
gl,
["vshader", "fshader"],
["a_position", "a_textureIndex"]);
gl.useProgram(program);
var imageLoc = gl.getUniformLocation(program, "u_image");
var paletteLoc = gl.getUniformLocation(program, "u_palette");
// tell it to use texture units 0 and 1 for the image and palette
gl.uniform1i(imageLoc, 0);
gl.uniform1i(paletteLoc, 1);
// Setup a unit quad
var positions = [
1, 1,
-1, 1,
-1, -1,
1, 1,
-1, -1,
1, -1,
];
var vertBuffer = gl.createBuffer();
gl.bindBuffer(gl.ARRAY_BUFFER, vertBuffer);
gl.bufferData(gl.ARRAY_BUFFER, new Float32Array(positions), gl.STATIC_DRAW);
gl.enableVertexAttribArray(0);
gl.vertexAttribPointer(0, 2, gl.FLOAT, false, 0, 0);
// Setup a palette.
var palette = new Uint8Array(256 * 4);
// I'm lazy so just setting 4 colors in palette
function setPalette(index, r, g, b, a) {
palette[index * 4 + 0] = r;
palette[index * 4 + 1] = g;
palette[index * 4 + 2] = b;
palette[index * 4 + 3] = a;
}
setPalette(1, 255, 0, 0, 255); // red
setPalette(2, 0, 255, 0, 255); // green
setPalette(3, 0, 0, 255, 255); // blue
// make palette texture and upload palette
gl.activeTexture(gl.TEXTURE1);
var paletteTex = gl.createTexture();
gl.bindTexture(gl.TEXTURE_2D, paletteTex);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_S, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_T, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MIN_FILTER, gl.NEAREST);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MAG_FILTER, gl.NEAREST);
gl.texImage2D(gl.TEXTURE_2D, 0, gl.RGBA, 256, 1, 0, gl.RGBA, gl.UNSIGNED_BYTE, palette);
// Make image. Just going to make something 8x8
var image = new Uint8Array([
0,0,1,1,1,1,0,0,
0,1,0,0,0,0,1,0,
1,0,0,0,0,0,0,1,
1,0,2,0,0,2,0,1,
1,0,0,0,0,0,0,1,
1,0,3,3,3,3,0,1,
0,1,0,0,0,0,1,0,
0,0,1,1,1,1,0,0,
]);
// make image textures and upload image
gl.activeTexture(gl.TEXTURE0);
var imageTex = gl.createTexture();
gl.bindTexture(gl.TEXTURE_2D, imageTex);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_S, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_T, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MIN_FILTER, gl.NEAREST);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MAG_FILTER, gl.NEAREST);
gl.texImage2D(gl.TEXTURE_2D, 0, gl.ALPHA, 8, 8, 0, gl.ALPHA, gl.UNSIGNED_BYTE, image);
gl.drawArrays(gl.TRIANGLES, 0, positions.length / 2);
canvas { border: 1px solid black; }
<script src="https://twgljs.org/dist/twgl.min.js"></script>
<script id="vshader" type="whatever">
attribute vec4 a_position;
varying vec2 v_texcoord;
void main() {
gl_Position = a_position;
// assuming a unit quad for position we
// can just use that for texcoords. Flip Y though so we get the top at 0
v_texcoord = a_position.xy * vec2(0.5, -0.5) + 0.5;
}
</script>
<script id="fshader" type="whatever">
precision mediump float;
varying vec2 v_texcoord;
uniform sampler2D u_image;
uniform sampler2D u_palette;
void main() {
float index = texture2D(u_image, v_texcoord).a * 255.0;
gl_FragColor = texture2D(u_palette, vec2((index + 0.5) / 256.0, 0.5));
}
</script>
<canvas id="c" width="256" height="256"></canvas>
To animate just update the image and then re-upload it into the texture
gl.texImage2D(gl.TEXTURE_2D, 0, gl.ALPHA, 8, 8, 0, gl.ALPHA,
gl.UNSIGNED_BYTE, image);
Example:
var canvas = document.getElementById("c");
var gl = canvas.getContext("webgl");
// Note: createProgramFromScripts will call bindAttribLocation
// based on the index of the attibute names we pass to it.
var program = twgl.createProgramFromScripts(
gl,
["vshader", "fshader"],
["a_position", "a_textureIndex"]);
gl.useProgram(program);
var imageLoc = gl.getUniformLocation(program, "u_image");
var paletteLoc = gl.getUniformLocation(program, "u_palette");
// tell it to use texture units 0 and 1 for the image and palette
gl.uniform1i(imageLoc, 0);
gl.uniform1i(paletteLoc, 1);
// Setup a unit quad
var positions = [
1, 1,
-1, 1,
-1, -1,
1, 1,
-1, -1,
1, -1,
];
var vertBuffer = gl.createBuffer();
gl.bindBuffer(gl.ARRAY_BUFFER, vertBuffer);
gl.bufferData(gl.ARRAY_BUFFER, new Float32Array(positions), gl.STATIC_DRAW);
gl.enableVertexAttribArray(0);
gl.vertexAttribPointer(0, 2, gl.FLOAT, false, 0, 0);
// Setup a palette.
var palette = new Uint8Array(256 * 4);
// I'm lazy so just setting 4 colors in palette
function setPalette(index, r, g, b, a) {
palette[index * 4 + 0] = r;
palette[index * 4 + 1] = g;
palette[index * 4 + 2] = b;
palette[index * 4 + 3] = a;
}
setPalette(1, 255, 0, 0, 255); // red
setPalette(2, 0, 255, 0, 255); // green
setPalette(3, 0, 0, 255, 255); // blue
// make palette texture and upload palette
gl.activeTexture(gl.TEXTURE1);
var paletteTex = gl.createTexture();
gl.bindTexture(gl.TEXTURE_2D, paletteTex);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_S, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_T, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MIN_FILTER, gl.NEAREST);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MAG_FILTER, gl.NEAREST);
gl.texImage2D(gl.TEXTURE_2D, 0, gl.RGBA, 256, 1, 0, gl.RGBA, gl.UNSIGNED_BYTE, palette);
// Make image. Just going to make something 8x8
var width = 8;
var height = 8;
var image = new Uint8Array([
0,0,1,1,1,1,0,0,
0,1,0,0,0,0,1,0,
1,0,0,0,0,0,0,1,
1,0,2,0,0,2,0,1,
1,0,0,0,0,0,0,1,
1,0,3,3,3,3,0,1,
0,1,0,0,0,0,1,0,
0,0,1,1,1,1,0,0,
]);
// make image textures and upload image
gl.activeTexture(gl.TEXTURE0);
var imageTex = gl.createTexture();
gl.bindTexture(gl.TEXTURE_2D, imageTex);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_S, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_T, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MIN_FILTER, gl.NEAREST);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MAG_FILTER, gl.NEAREST);
gl.texImage2D(gl.TEXTURE_2D, 0, gl.ALPHA, width, height, 0, gl.ALPHA, gl.UNSIGNED_BYTE, image);
var frameCounter = 0;
function render() {
++frameCounter;
// skip 3 of 4 frames so the animation is not too fast
if ((frameCounter & 3) == 0) {
// rotate the image left
for (var y = 0; y < height; ++y) {
var temp = image[y * width];
for (var x = 0; x < width - 1; ++x) {
image[y * width + x] = image[y * width + x + 1];
}
image[y * width + width - 1] = temp;
}
// re-upload image
gl.activeTexture(gl.TEXTURE0);
gl.texImage2D(gl.TEXTURE_2D, 0, gl.ALPHA, width, height, 0, gl.ALPHA,
gl.UNSIGNED_BYTE, image);
gl.drawArrays(gl.TRIANGLES, 0, positions.length / 2);
}
requestAnimationFrame(render);
}
render();
canvas { border: 1px solid black; }
<script src="https://twgljs.org/dist/twgl.min.js"></script>
<script id="vshader" type="whatever">
attribute vec4 a_position;
varying vec2 v_texcoord;
void main() {
gl_Position = a_position;
// assuming a unit quad for position we
// can just use that for texcoords. Flip Y though so we get the top at 0
v_texcoord = a_position.xy * vec2(0.5, -0.5) + 0.5;
}
</script>
<script id="fshader" type="whatever">
precision mediump float;
varying vec2 v_texcoord;
uniform sampler2D u_image;
uniform sampler2D u_palette;
void main() {
float index = texture2D(u_image, v_texcoord).a * 255.0;
gl_FragColor = texture2D(u_palette, vec2((index + 0.5) / 256.0, 0.5));
}
</script>
<canvas id="c" width="256" height="256"></canvas>
Of course that assumes your goal is to do the animation on the CPU by manipulating pixels. Otherwise you can use any normal webgl techniques to manipulate texture coordinates or whatever.
You can also update the palette similarly for palette animation. Just modify the palette and re-upload it
gl.texImage2D(gl.TEXTURE_2D, 0, gl.RGBA, 256, 1, 0, gl.RGBA,
gl.UNSIGNED_BYTE, palette);
Example:
var canvas = document.getElementById("c");
var gl = canvas.getContext("webgl");
// Note: createProgramFromScripts will call bindAttribLocation
// based on the index of the attibute names we pass to it.
var program = twgl.createProgramFromScripts(
gl,
["vshader", "fshader"],
["a_position", "a_textureIndex"]);
gl.useProgram(program);
var imageLoc = gl.getUniformLocation(program, "u_image");
var paletteLoc = gl.getUniformLocation(program, "u_palette");
// tell it to use texture units 0 and 1 for the image and palette
gl.uniform1i(imageLoc, 0);
gl.uniform1i(paletteLoc, 1);
// Setup a unit quad
var positions = [
1, 1,
-1, 1,
-1, -1,
1, 1,
-1, -1,
1, -1,
];
var vertBuffer = gl.createBuffer();
gl.bindBuffer(gl.ARRAY_BUFFER, vertBuffer);
gl.bufferData(gl.ARRAY_BUFFER, new Float32Array(positions), gl.STATIC_DRAW);
gl.enableVertexAttribArray(0);
gl.vertexAttribPointer(0, 2, gl.FLOAT, false, 0, 0);
// Setup a palette.
var palette = new Uint8Array(256 * 4);
// I'm lazy so just setting 4 colors in palette
function setPalette(index, r, g, b, a) {
palette[index * 4 + 0] = r;
palette[index * 4 + 1] = g;
palette[index * 4 + 2] = b;
palette[index * 4 + 3] = a;
}
setPalette(1, 255, 0, 0, 255); // red
setPalette(2, 0, 255, 0, 255); // green
setPalette(3, 0, 0, 255, 255); // blue
// make palette texture and upload palette
gl.activeTexture(gl.TEXTURE1);
var paletteTex = gl.createTexture();
gl.bindTexture(gl.TEXTURE_2D, paletteTex);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_S, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_T, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MIN_FILTER, gl.NEAREST);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MAG_FILTER, gl.NEAREST);
gl.texImage2D(gl.TEXTURE_2D, 0, gl.RGBA, 256, 1, 0, gl.RGBA, gl.UNSIGNED_BYTE, palette);
// Make image. Just going to make something 8x8
var width = 8;
var height = 8;
var image = new Uint8Array([
0,0,1,1,1,1,0,0,
0,1,0,0,0,0,1,0,
1,0,0,0,0,0,0,1,
1,0,2,0,0,2,0,1,
1,0,0,0,0,0,0,1,
1,0,3,3,3,3,0,1,
0,1,0,0,0,0,1,0,
0,0,1,1,1,1,0,0,
]);
// make image textures and upload image
gl.activeTexture(gl.TEXTURE0);
var imageTex = gl.createTexture();
gl.bindTexture(gl.TEXTURE_2D, imageTex);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_S, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_WRAP_T, gl.CLAMP_TO_EDGE);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MIN_FILTER, gl.NEAREST);
gl.texParameteri(gl.TEXTURE_2D, gl.TEXTURE_MAG_FILTER, gl.NEAREST);
gl.texImage2D(gl.TEXTURE_2D, 0, gl.ALPHA, width, height, 0, gl.ALPHA, gl.UNSIGNED_BYTE, image);
var frameCounter = 0;
function render() {
++frameCounter;
// skip 3 of 4 frames so the animation is not too fast
if ((frameCounter & 3) == 0) {
// rotate the 3 palette colors
var tempR = palette[4 + 0];
var tempG = palette[4 + 1];
var tempB = palette[4 + 2];
var tempA = palette[4 + 3];
setPalette(1, palette[2 * 4 + 0], palette[2 * 4 + 1], palette[2 * 4 + 2], palette[2 * 4 + 3]);
setPalette(2, palette[3 * 4 + 0], palette[3 * 4 + 1], palette[3 * 4 + 2], palette[3 * 4 + 3]);
setPalette(3, tempR, tempG, tempB, tempA);
// re-upload palette
gl.activeTexture(gl.TEXTURE1);
gl.texImage2D(gl.TEXTURE_2D, 0, gl.RGBA, 256, 1, 0, gl.RGBA,
gl.UNSIGNED_BYTE, palette);
gl.drawArrays(gl.TRIANGLES, 0, positions.length / 2);
}
requestAnimationFrame(render);
}
render();
canvas { border: 1px solid black; }
<script src="https://twgljs.org/dist/twgl.min.js"></script>
<script id="vshader" type="whatever">
attribute vec4 a_position;
varying vec2 v_texcoord;
void main() {
gl_Position = a_position;
// assuming a unit quad for position we
// can just use that for texcoords. Flip Y though so we get the top at 0
v_texcoord = a_position.xy * vec2(0.5, -0.5) + 0.5;
}
</script>
<script id="fshader" type="whatever">
precision mediump float;
varying vec2 v_texcoord;
uniform sampler2D u_image;
uniform sampler2D u_palette;
void main() {
float index = texture2D(u_image, v_texcoord).a * 255.0;
gl_FragColor = texture2D(u_palette, vec2((index + 0.5) / 256.0, 0.5));
}
</script>
<canvas id="c" width="256" height="256"></canvas>
Slightly related is this tile shader example
http://blog.tojicode.com/2012/07/sprite-tile-maps-on-gpu.html
Presumably you're building up a javascript array that's around 512 x 512 (PAL size)...
A WebGL fragment shader could definitely do your palette conversion pretty nicely. The recipe would go something like this:
Set up WebGL with a "geometry" of just two triangles that span your viewport. (GL is all triangles.) This is the biggest bother, if you're not already GL fluent. But it's not that bad. Spend some quality time with http://learningwebgl.com/blog/?page_id=1217 . But it will be ~100 lines of stuff. Price of admission.
Build your in-memory frame buffer 4 times bigger. (I think textures always have to be RGBA?) And populate every fourth byte, the R component, with your pixel values. Use new Float32Array to allocate it. You can use values 0-255, or divide it down to 0.0 to 1.0. We'll pass this to webgl as a texture. This one changes every frame.
Build a second texture that's 256 x 1 pixels, which is your palette lookup table. This one never changes (unless the palette can be modified?).
In your fragment shader, use your emulated frame buffer texture as a lookup into your palette. The first pixel in the palette is accessed at location (0.5/256.0, 0.5), middle of the pixel.
On each frame, resubmit the emulated frame buffer texture and redraw. Pushing pixels to the GPU is expensive... but a PAL-sized image is pretty small by modern standards.
Bonus step: You could enhance the fragment shader to imitate scanlines, interlace video, or other cute emulation artifacts (phosphor dots?) on modern high-resolution displays, all at no cost to your javascript!
This is just a sketch. But it will work. WebGL is a pretty low-level API, and quite flexible, but well worth the effort (if you like that kind of thing, which I do. :-) ).
Again, http://learningwebgl.com/blog/?page_id=1217 is well-recommended for overall WebGL guidance.