Related
I been stuck on getting the waves to look just like I want. I'm trying to figure out how to get the base of the wave to be the color I need it. I can do my desired color but it blocks the background. I can not see anything behind it because I was using like a reflection. Maybe someone can figure it out cause I'm having difficulties getting it to work... I plan on making the wave drop and rise. Here is a link to the code pen: HERE
Here is where I have the vertical reflection:
var x = $.cx - $.length / 2 + $.length / $.count * i,
y = height + $.simplex.noise2D($.xoff, $.yoff) * amp + sway;
$.ctx[i === 0 ? 'moveTo' : 'lineTo'](x, y);
}
$.ctx.lineTo($.w, $.h); // -$.h - Vertically reflection
$.ctx.lineTo(0, $.h); // -$.h - Vertically reflection
$.ctx.closePath();
$.ctx.fillStyle = color;
if (comp) {
$.ctx.globalCompositeOperation = comp;
}
$.ctx.fill();
My desired look for the waves is below:
Here is what I got with a successful transparent top, just not the right coloring:
Your problem is that the screen blending of the three colors generates a solid white color, so all the bottom of your canvas becomes white.
Here I simplified a lot the situation, with just 3 rectangles. Your bottom of canvas is my central white square:
const c2 = canvas.cloneNode();
const ctx = canvas.getContext("2d");
ctx.globalCompositeOperation = 'screen';
ctx.fillStyle = '#fb0000';
ctx.fillRect(0,0,50,50);
ctx.fillStyle = "#00ff8e";
ctx.fillRect(12,12,50,50);
ctx.fillStyle = "#6F33FF";
ctx.fillRect(25,25,50,50);
body {
background: #CCC;
}
<canvas id="canvas"></canvas>
So what we need, is a way to make this central square transparent so that we can draw our background behind.
To do this, we will need to draw our shapes at least two times:
once in normal compositing mode, so that we get the full overlap.
once again as source-in compositing mode, so that we get only where all our shapes do overlap.
const ctx = canvas.getContext("2d");
function drawShapes(mode) {
ctx.globalCompositeOperation = mode;
ctx.fillStyle = '#fb0000';
ctx.fillRect(0,0,50,50);
ctx.fillStyle = "#00ff8e";
ctx.fillRect(12,12,50,50);
ctx.fillStyle = "#6F33FF";
ctx.fillRect(25,25,50,50);
}
drawShapes('screen');
drawShapes('source-in');
body {
background: #CCC;
}
<canvas id="canvas"></canvas>
Now we have our overlapping area, we will be able to use it as a cutting shape in a third operation. But to do it, we will need a second, off-screen canvas to perform the compositing of the two states:
const c2 = canvas.cloneNode();
const ctx = canvas.getContext("2d");
const ctx2 = c2.getContext("2d");
function drawShapes(ctx, comp) {
ctx.globalCompositeOperation = comp;
ctx.fillStyle = '#fb0000';
ctx.fillRect(0, 0, 50, 50);
ctx.fillStyle = "#00ff8e";
ctx.fillRect(12, 12, 50, 50);
ctx.fillStyle = "#6F33FF";
ctx.fillRect(25, 25, 50, 50);
}
// first draw our screen, with unwanted white square
drawShapes(ctx, 'screen');
// draw it on the offscreen canvas
ctx2.drawImage(ctx.canvas, 0, 0)
// draw the shapes once again on the offscreen canvas to get the cutting shape
drawShapes(ctx2, 'source-in');
// cut the visible canvas
ctx.globalCompositeOperation = 'destination-out'
ctx.drawImage(ctx2.canvas, 0, 0);
body {
background: #CCC
}
<canvas id="canvas"></canvas>
And voilà, our white square is now transparent, we can draw whatever we want behind our scene using the destination-over composite operation.
I have two canvas elements and need them to be resized on buttons click.
<div class="sDetails"><div>
<div id="canvasDiv" style="width: 310px;"><canvas id="canvasGraph"></canvas></div></div>
<div class="kDetails"><div><div>
<div id="canvasDiv" style="width: 310px; height: 240px;"><canvas id="canvasGraph"></canvas></div></div>
and the script:
var sketch;var sketch_sl;var onPaint;var canvas=null;var ctx=null;var tmp_ctx=null;
function drawCanvas(div) {
canvas = document.querySelector(div + " #canvasGraph");
ctx = canvas.getContext('2d');
sketch = document.querySelector(div + " #canvasDiv");
sketch_sl = getComputedStyle(sketch);
canvas.width = parseInt(sketch_style.getPropertyValue('width'));
canvas.height = parseInt(sketch_style.getPropertyValue('height'));
tmp_canvas = document.createElement('canvas');
tmp_ctx = tmp_canvas.getContext('2d');
tmp_canvas.id = 'tmp_canvas';
tmp_canvas.width = canvas.width;
tmp_canvas.height = canvas.height;
sketch.appendChild(tmp_canvas);
the redraw function:
// here I must redraw my lines resized 2 times ( *cScale ) where cScale=2 or =1
function drawScales(ctx, canvas)
ctx.strokeStyle = 'green';
ctx.fillStyle = 'green';
ctx.beginPath();
ctx.moveTo(5, 0);
ctx.lineTo(0, canvas.height);
scaleStep = 24*cScale;
for some reason it works really bad, old positions stay.
Is there a way to completely delete the whole canvas and append it or redraw it completely?
I tried canvas.width=canvas.width, tried ctx.clearRect(0, 0, canvas.width, canvas.height);tmp_ctx.clearRect(0, 0, canvas.width, canvas.height);, tried $(".sDetails #canvasGraph")[0].reset();
logically, drawCanvas(".sDetails");drawLines(ctx, canvas); should redraw it from scratch but it will not.
Resize the canvas element's width & height and use context.scale to redraw the original drawings at their newly scaled size.
Resizing the canvas element will automatically clear all drawings off the canvas.
Resizing will also automatically reset all context properties back to their default values.
Using context.scale is useful because then the canvas will automatically rescale the original drawings to fit on the newly sized canvas.
Important: Canvas will not automatically redraw the original drawings...you must re-issue the original drawing commands.
Illustration with 2 canvases at same size (their sizes are controlled by range controls)
Illustration with left canvas resized larger
Illustration with right canvas resized larger
Here's example code and a Demo. This demo uses range elements to control the resizing, but you can also do the resizing+redrawing inside window.onresize
var canvas1=document.getElementById("canvas1");
var ctx1=canvas1.getContext("2d");
var canvas2=document.getElementById("canvas2");
var ctx2=canvas2.getContext("2d");
var originalWidth=canvas1.width;
var originalHeight=canvas1.height;
var scale1=1;
var scale2=1;
$myslider1=$('#myslider1');
$myslider1.attr({min:50,max:200}).val(100);
$myslider1.on('input change',function(){
var scale=parseInt($(this).val())/100;
scale1=scale;
redraw(ctx1,scale);
});
$myslider2=$('#myslider2');
$myslider2.attr({min:50,max:200}).val(100);
$myslider2.on('input change',function(){
var scale=parseInt($(this).val())/100;
scale2=scale;
redraw(ctx2,scale);
});
draw(ctx1);
draw(ctx2);
function redraw(ctx,scale){
// Resizing the canvas will clear all drawings off the canvas
// Resizing will also automatically clear the context
// of all its current values and set default context values
ctx.canvas.width=originalWidth*scale;
ctx.canvas.height=originalHeight*scale;
// context.scale will scale the original drawings to fit on
// the newly resized canvas
ctx.scale(scale,scale);
draw(ctx);
// always clean up! Reverse the scale
ctx.scale(-scale,-scale);
}
function draw(ctx){
// note: context.scale causes canvas to do all the rescaling
// math for us, so we can always just draw using the
// original sizes and x,y coordinates
ctx.beginPath();
ctx.moveTo(150,50);
ctx.lineTo(250,150);
ctx.lineTo(50,150);
ctx.closePath();
ctx.stroke();
ctx.fillStyle='skyblue';
ctx.beginPath();
ctx.arc(150,50,20,0,Math.PI*2);
ctx.closePath();
ctx.fill();
ctx.stroke();
ctx.beginPath();
ctx.arc(250,150,20,0,Math.PI*2);
ctx.closePath();
ctx.fill();
ctx.stroke();
ctx.beginPath();;
ctx.arc(50,150,20,0,Math.PI*2);
ctx.fill();
ctx.stroke();
}
$("#canvas1, #canvas2").mousemove(function(e){handleMouseMove(e);});
var $mouse=$('#mouse');
function handleMouseMove(e){
// tell the browser we're handling this event
e.preventDefault();
e.stopPropagation();
var bb=e.target.getBoundingClientRect();
mouseX=parseInt(e.clientX-bb.left);
mouseY=parseInt(e.clientY-bb.top);
if(e.target.id=='canvas1'){
$mouse.text('Mouse1: '+mouseX/scale1+' / '+mouseY/scale1+' (scale:'+scale1+')');
}else{
$mouse.text('Mouse2: '+mouseX/scale2+' / '+mouseY/scale2+' (scale:'+scale2+')');
}
}
body{ background-color: ivory; }
canvas{border:1px solid red;}
<script src="https://ajax.googleapis.com/ajax/libs/jquery/1.9.1/jquery.min.js"></script>
<div>Resize left canvas</div>
<input id=myslider1 type=range><br>
<div>Resize right canvas</div>
<input id=myslider2 type=range><br>
<h4 id=mouse>Mouse coordinates:</h4>
<canvas id="canvas1" width=300 height=300></canvas>
<canvas id="canvas2" width=300 height=300></canvas>
If you need scale-independent positions you could use normalized values ([0, 1]) instead and use the size of canvas as the scale factor. This way you can scale and store values without too much concern about the actual target size.
You would also be able to use the mouse positions almost as is and normalize by just dividing them on canvas size.
For example:
When rendering, a point of (1,1) will always draw in lower-right corner as you would do (1 * canvas.width, 1 * canvas.height).
When you store a point you would use the mouse position and divide it on the canvas dimension, for example, if I click in the lower right corner of a canvas of size 400x200, the points would be 400/400 = 1, 200/200 = 1.
Note that width and height would be exclusive (ie. width-1 etc.), but for sake of simplicity...
Example
In this example you can start with any size of the canvas, draw points which are normalized, change size of canvas and have the points redrawn proportionally relative to the original position.
var rng = document.querySelector("input"),
c = document.querySelector("canvas"),
ctx = c.getContext("2d"),
points = [];
// change canvas size and redraw all points
rng.onchange = function() {
c.width = +this.value;
render();
};
// add a new normalized point to array
c.onclick = function(e) {
var r = this.getBoundingClientRect(), // to adjust mouse position
x = e.clientX - r.left,
y = e.clientY - r.top;
points.push({
x: x / c.width, // normalize value to range [0, 1]
y: y / c.height
}); // store point
render(); // redraw (for demo)
};
function render() {
ctx.clearRect(0, 0, c.width, c.height); // clear canvas
ctx.beginPath(); // clear path
for(var i = 0, p; p = points[i]; i++) { // draw points as fixed-size circles
var x = p.x * c.width, // normalized to absolute values
y = p.y * c.height;
ctx.moveTo(x + 5, y);
ctx.arc(x, y, 5, 0, 6.28);
ctx.closePath();
}
ctx.stroke();
}
canvas {background:#ddd}
<h3>Click on canvas to add points, then resize</h3>
<label>Width: <input type="range" min=50 max=600 value=300></label><br>
<canvas></canvas>
I decided to use a scale variable to resize my scales. I resize the canvas canvas.width *= 2; and then I redraw my scales.
var scaleStep;
and use add it into the code: ctx.lineTo(12*24*cScale+12, canvas.height-24); where the scaling needs to be done.
The scaleStep is 2 when maximizing the canvas and 1 when returning to the original size.
I found images that depict what is my problem:
User will able to choose four points on canvas to crop the part of image and than stretch it.
How to do that in HTML5? drawImage function (as I know) works only with rectangles (takes x, y, width and height values) so I can't use irregular shape. The solution have to work in every modern browser, so I don't want things based on webgl or something.
EDIT:
More info: this will be app for editing pictures. I want to let user cut some part of bigger picture and edit that. It will be similar to Paint, so canvas is required to edit pixels.
The effect you're going for is "perspective warping".
Canvas's 2D context cannot do this "out-of-the-box" because it can't turn a rectangle into a trapezoid. Canvas 2D can only do affine transforms which can only form parallelograms.
As user #Canvas says, Canvas 3D (webgl) can do the transforms you're going for.
I did this a while back. It uses Canvas 2d and it redraws an image using 1 pixel wide vertical slices which are stretched to "fake" a perspective warp. You are welcome to use it as a starting point for your project.
Example code and a Demo: http://jsfiddle.net/m1erickson/y4kst2pk/
<!doctype html>
<html>
<head>
<link rel="stylesheet" type="text/css" media="all" href="css/reset.css" /> <!-- reset css -->
<script type="text/javascript" src="http://code.jquery.com/jquery.min.js"></script>
<style>
body{ background-color: ivory; }
#canvas{border:1px solid red;}
</style>
<script>
$(function(){
var canvas=document.getElementById("canvas");
var ctx=canvas.getContext("2d");
var $canvas=$("#canvas");
var canvasOffset=$canvas.offset();
var offsetX=canvasOffset.left;
var offsetY=canvasOffset.top;
var scrollX=$canvas.scrollLeft();
var scrollY=$canvas.scrollTop();
//
var isDown=false;
var PI2=Math.PI*2;
var selectedGuide=-1;
var guides=[];
//
var marginLeft=50;
var marginTop=50;
var iw,ih,cw,ch;
var img=new Image();
img.onload=start;
img.src='https://dl.dropboxusercontent.com/u/139992952/stack1/buildings1.jpg';
function start(){
iw=img.width;
ih=img.height;
canvas.width=iw+100;
canvas.height=ih+100;
cw=canvas.width;
ch=canvas.height;
ctx.strokeStyle="blue";
ctx.fillStyle="blue";
guides.push({x:0,y:0,r:10});
guides.push({x:0,y:ih,r:10});
guides.push({x:iw,y:0,r:10});
guides.push({x:iw,y:ih,r:10});
//
$("#canvas").mousedown(function(e){handleMouseDown(e);});
$("#canvas").mousemove(function(e){handleMouseMove(e);});
$("#canvas").mouseup(function(e){handleMouseUp(e);});
$("#canvas").mouseout(function(e){handleMouseOut(e);});
drawAll();
}
function drawAll(){
ctx.clearRect(0,0,cw,ch);
drawGuides();
drawImage();
}
function drawGuides(){
for(var i=0;i<guides.length;i++){
var guide=guides[i];
ctx.beginPath();
ctx.arc(guide.x+marginLeft,guide.y+marginTop,guide.r,0,PI2);
ctx.closePath();
ctx.fill();
}
}
function drawImage(){
// TODO use guides
var x1=guides[0].x;
var y1=guides[0].y;
var x2=guides[2].x;
var y2=guides[2].y;
var x3=guides[1].x;
var y3=guides[1].y;
var x4=guides[3].x;
var y4=guides[3].y;
// calc line equations slope & b (m,b)
var m1=Math.tan( Math.atan2((y2-y1),(x2-x1)) );
var b1=y2-m1*x2;
var m2=Math.tan( Math.atan2((y4-y3),(x4-x3)) );
var b2=y4-m2*x4;
// draw vertical slices
for(var X=0;X<iw;X++){
var yTop=m1*X+b1;
var yBottom=m2*X+b2;
ctx.drawImage( img,X,0,1,ih,
X+marginLeft,yTop+marginTop,1,yBottom-yTop );
}
// outline
ctx.save();
ctx.translate(marginLeft,marginTop);
ctx.beginPath();
ctx.moveTo(x1,y1);
ctx.lineTo(x2,y2);
ctx.lineTo(x4,y4);
ctx.lineTo(x3,y3);
ctx.closePath();
ctx.strokeStyle="black";
ctx.stroke();
ctx.restore();
}
function handleMouseDown(e){
e.preventDefault();
var mouseX=parseInt(e.clientX-offsetX);
var mouseY=parseInt(e.clientY-offsetY);
// Put your mousedown stuff here
selectedGuide=-1;
for(var i=0;i<guides.length;i++){
var guide=guides[i];
var dx=mouseX-(guide.x+marginLeft);
var dy=mouseY-(guide.y+marginTop);
if(dx*dx+dy*dy<=guide.r*guide.r){
selectedGuide=i;
break;
}
}
isDown=(selectedGuide>=0);
}
function handleMouseUp(e){
e.preventDefault();
isDown=false;
}
function handleMouseOut(e){
e.preventDefault();
isDown=false;
}
function handleMouseMove(e){
if(!isDown){return;}
e.preventDefault();
var x=parseInt(e.clientX-offsetX)-marginLeft;
var y=parseInt(e.clientY-offsetY)-marginTop;
var guide=guides[selectedGuide];
guides[selectedGuide].y=y;
if(selectedGuide==0 && y>guides[1].y){guide.y=guides[1].y;}
if(selectedGuide==1 && y<guides[0].y){guide.y=guides[0].y;}
if(selectedGuide==2 && y>guides[3].y){guide.y=guides[3].y;}
if(selectedGuide==3 && y<guides[2].y){guide.y=guides[2].y;}
drawAll();
}
}); // end $(function(){});
</script>
</head>
<body>
<h4>Perspective Warp by vertically dragging left or right blue guides.</h4>
<canvas id="canvas" width=300 height=300></canvas>
</body>
</html>
So here's a killing trick : You can use the regular drawImage of the context2d to draw a texture inside a triangle.
The constraint is that the texture coordinates must be axis-aligned.
To draw a textured triangle you have to :
• Compute by yourself the transform required to draw the image.
• Clip to a triangle, since drawImage would draw a quad.
• drawImage with the right transform.
So the idea is to split your quad into two triangles, and render both.
But there's one more trick : when drawing the lower-right triangle, texture reading should start from the lower-right part of the texture, then move up and left. This can't be done in Firefox, which accepts only positive arguments to drawImage. So i compute the 'mirror' of the first point vs the two others, and i can draw in the regular direction again -the clipping will ensure only right part is drawn-.
fiddle is here :
http://jsfiddle.net/gamealchemist/zch3gdrx/
function rasterizeTriangle(v1, v2, v3, mirror) {
var fv1 = {
x: 0,
y: 0,
u: 0,
v: 0
};
fv1.x = v1.x;
fv1.y = v1.y;
fv1.u = v1.u;
fv1.v = v1.v;
ctx.save();
// Clip to draw only the triangle
ctx.beginPath();
ctx.moveTo(v1.x, v1.y);
ctx.lineTo(v2.x, v2.y);
ctx.lineTo(v3.x, v3.y);
ctx.clip();
// compute mirror point and flip texture coordinates for lower-right triangle
if (mirror) {
fv1.x = fv1.x + (v3.x - v1.x) + (v2.x - v1.x);
fv1.y = fv1.y + (v3.y - v1.y) + (v2.y - v1.y);
fv1.u = v3.u;
fv1.v = v2.v;
}
//
var angleX = Math.atan2(v2.y - fv1.y, v2.x - fv1.x);
var angleY = Math.atan2(v3.y - fv1.y, v3.x - fv1.x);
var scaleX = lengthP(fv1, v2);
var scaleY = lengthP(fv1, v3);
var cos = Math.cos,
sin = Math.sin;
// ----------------------------------------
// Transforms
// ----------------------------------------
// projection matrix (world relative to center => screen)
var transfMatrix = [];
transfMatrix[0] = cos(angleX) * scaleX;
transfMatrix[1] = sin(angleX) * scaleX;
transfMatrix[2] = cos(angleY) * scaleY;
transfMatrix[3] = sin(angleY) * scaleY;
transfMatrix[4] = fv1.x;
transfMatrix[5] = fv1.y;
ctx.setTransform.apply(ctx, transfMatrix);
// !! draw !!
ctx.drawImage(bunny, fv1.u, fv1.v, v2.u - fv1.u, v3.v - fv1.v,
0, 0, 1, 1);
//
ctx.restore();
};
Edit : i added the relevant comment of #szym , with his example picture :
This only sort of looks right. If there are any straight lines in the
original image you will see that each triangle is warped differently
(2 different affine transforms rather than a perspective transform).
You need to have a container with perspective and perspective-origin set
You need to use rotateY, skewY and change your heights and width on your image
There is probably a lot of math behind this - personally I just fiddle with it in my browser to make it look pretty close to what I need
So here is a fiddle:
http://jsfiddle.net/6egdevwe/1/
#container {
margin: 50px;
perspective: 166px; perspective-origin: 50% 0px; }
#testimage {
transform: rotateY(93.4deg) skewY(34deg);
width: 207px;
height: 195px; }
css3 transform -> rotation or rotationZ
http://www.w3schools.com/cssref/css3_pr_transform.asp
I am trying to draw lines with canvas and I am changing the coordinates with a for loop.
here is my canvas element:
<canvas id="c" width="300px" height="300px"></canvas>
and here is the js codes:
var c = document.getElementById('c');
ci = c.getContext('2d');
for(var a = 18; a < 300; a +=18){
fnc(a, ci);
}
function fnc(x, ci){
ci.strokeStyle = 'red';
ci.moveTo(0, x);
ci.lineTo(300, x); ci.lineWidth = 0.2; ci.stroke();
}
As you can see I am trying to draw these lines with 18px spaces between them. But the thickness of the lines and the color(or opacity, I am not sure) are changing from top to bottom.
Here is a fiddle : http://jsfiddle.net/J6zzD/1/
So what is wrong with that I can't find my mistake. Why are the color and the thicknesses are different?
UPDATE :
I just wrote these lines out of the function and now all the lines becomes faded but thicknesses are same. So strange :
ci.strokeStyle = 'red';
ci.lineWidth = 0.2; ci.stroke();
here is demo : http://jsfiddle.net/J6zzD/4/
That's again the eternal issue of forgetting to call beginPath.
Each time you call moveTo then lineTo, you create a new *sub*path, which adds to the current Path.
Then each time you call stroke(), the current path, so all the current subpaths get re-drawn, when the last added path is drawn for the first time.
Since opacities will add-up, top lines will reach 100% opacity (alpha=255) when the bottom line, drawn once, will have a 20% opacity (lineWidth=0.2) .
In your second fiddle, you stroke only once, so all lines have 20% opacity, which is correct for the 0.2 lineWidth.
So : use beginPath before drawing a new figure.
In this case you have two choices :
• draw line by line
OR
• draw once a path with all lines as subpath.
(see code below).
TIP : To get clean lines remember that pixels's center is at the (+0.5, +0.5) coordinates of each pixels, so
a 'trick' is to translate by 0.5, 0.5 on app start, then only use rounded coordinates and lineWidth.
1) draw line by line
http://jsfiddle.net/gamealchemist/J6zzD/6/
var c = document.getElementById('c');
var ctx = c.getContext('2d');
ctx.translate(0.5, 0.5);
ctx.lineWidth = 1;
for (var y = 18; y < 300; y += 18) {
strokeLine(ctx, y);
}
function strokeLine(ctx, y) {
ctx.beginPath();
ctx.strokeStyle = 'red';
ctx.moveTo(0, y);
ctx.lineTo(300, y);
ctx.stroke();
}
2) draw multiple subPath :
(you can have only one color for one stroke() )
http://jsfiddle.net/gamealchemist/J6zzD/7/
var c = document.getElementById('c');
var ctx = c.getContext('2d');
ctx.translate(0.5, 0.5);
ctx.lineWidth = 1;
ctx.strokeStyle = 'red';
ctx.beginPath();
for (var y = 18; y < 300; y += 18) {
addLineSubPath(ctx, y);
}
ctx.stroke();
function addLineSubPath(ctx, y) {
ctx.moveTo(0, y);
ctx.lineTo(300, y);
}
See: https://developer.mozilla.org/en-US/docs/Web/Guide/HTML/Canvas_tutorial/Applying_styles_and_colors#A_lineWidth_example
Because canvas coordinates do not directly reference pixels, special care must be taken to obtain crisp horizontal and vertical lines.
Basically, because you're trying to draw a line that's 0.2 pixels wide, the browser does some math to approximate a continuous number into discrete units and you get your "fading" lines.
So now we can fix up your code by changing context.lineWidth to 1 (I actually remove it because it defaults to 1) and shifting everything down by half a pixel.
var c = document.getElementById('c');
ci = c.getContext('2d');
for(var a = 18.5; a < 300.5; a +=18)
{
fnc(a, ci);
}
function fnc(x, ci)
{
ci.strokeStyle = 'red';
ci.moveTo(0, x);
ci.lineTo(300, x);
ci.stroke();
}
Demo
The final code that worked for me was:
<canvas id="bg-admin-canvas" width="500" height="500" style="margin:15px; background:#09F;"></canvas>
<script>
var postit = function(width,height,angle){
var canvas = document.getElementById("bg-admin-canvas");
var ctx = canvas.getContext("2d");
var radians = angle * Math.PI / 180;
var move = width*Math.sin(radians);
if(angle < 0 ){ ctx.translate(0,-move); }else{ ctx.translate(move,0); }
ctx.rotate(radians);
var gradient = ctx.createLinearGradient(0,height,width/2,height/2);
gradient.addColorStop(0.05,"rgba(0,0,0,0)");
gradient.addColorStop(0.5,"rgba(0,0,0,0.3)");
ctx.fillStyle = gradient;
ctx.fillRect(0,0,width,height);
ctx.beginPath();
ctx.moveTo(0,0);
ctx.lineTo(width, 0);
ctx.lineTo(width,height);
ctx.lineTo(width-width*.8,height-height*.02);
ctx.quadraticCurveTo(0+width*.02,height-height*.02,0+width*.02,(height - height*.2));
ctx.closePath();
var gradient = ctx.createLinearGradient(0,height,width/2,height/2);
gradient.addColorStop(0,'#f7f8b9');
gradient.addColorStop(1,'#feffcf');
ctx.fillStyle = gradient;
ctx.fill();
ctx.beginPath();
ctx.moveTo(width-width*.8,height-height*.02);
ctx.quadraticCurveTo(0+width*.02,height-height*.02,0+width*.02,(height - height*.2));
ctx.quadraticCurveTo(width*.05,height-height*.05,width*.1,height-height*.1);
ctx.quadraticCurveTo(width*.1,height-height*.07,width-width*.8,height-height*.02);
ctx.closePath();
ctx.fillStyle = '#ffffff';
ctx.fill();
var gradient = ctx.createLinearGradient(0,height,width*.1,height-height*.1);
gradient.addColorStop(0,"rgba(222,222,163,0.8)");
gradient.addColorStop(1,'#feffcf');
ctx.fillStyle = gradient;
ctx.fill();
}
postit(300, 300, 10);
</script>
Hi,
I made a quick and dirty "post-it" note with html5's canvas and some js.
I want to be able to rotate them anyway I want so I tried to use the translate. The example below I have a translate of 0,250 just so you could see the whole thing.
Ideally, I know if my canvas was 300,300 then I would
ctx.translate(150,150);
ctx.rotate(-30);
ctx.translate(-150,-150);
Of course since I'm rotating a square it gets cut off.
How would I rotate the square and move it on the canvas so the whole thing is showing but at the very top left edge of the canvas?
I added an image with my thinking of just getting the height of a triangle and moving it that much, but when translated, it doesn't seem to work just right.
I'll paste my whole function so you can look at it, but if you have any ideas, I would appreciate it. This isn't important, just messing around today.
var postit = function(width,height,angle){
var canvas = jQuery("#bg-admin-canvas").get(0);
var ctx = canvas.getContext("2d");
/*var area = (width*width*Math.sin(angle))/2;
var h = (area*2) / width + 30;
ctx.translate(0,h);
*/
//ctx.translate(150,150);
ctx.translate(0,250);
ctx.rotate(angle*Math.PI / 180);
//ctx.translate(-150,-150);
var gradient = ctx.createLinearGradient(0,height,width/2,height/2);
gradient.addColorStop(0.05,"rgba(0,0,0,0)");
gradient.addColorStop(0.5,"rgba(0,0,0,0.3)");
ctx.fillStyle = gradient;
ctx.fillRect(0,0,width,height);
ctx.beginPath();
ctx.moveTo(0,0);
ctx.lineTo(width, 0);
ctx.lineTo(width,height);
ctx.lineTo(width-width*.8,height-height*.02);
ctx.quadraticCurveTo(0+width*.02,height-height*.02,0+width*.02,(height - height*.2));
ctx.closePath();
var gradient = ctx.createLinearGradient(0,height,width/2,height/2);
gradient.addColorStop(0,'#f7f8b9');
gradient.addColorStop(1,'#feffcf');
ctx.fillStyle = gradient;
ctx.fill();
ctx.beginPath();
ctx.moveTo(width-width*.8,height-height*.02);
ctx.quadraticCurveTo(0+width*.02,height-height*.02,0+width*.02,(height - height*.2));
ctx.quadraticCurveTo(width*.05,height-height*.05,width*.1,height-height*.1);
ctx.quadraticCurveTo(width*.1,height-height*.07,width-width*.8,height-height*.02);
ctx.closePath();
ctx.fillStyle = '#ffffff';
ctx.fill();
var gradient = ctx.createLinearGradient(0,height,width*.1,height-height*.1);
gradient.addColorStop(0,"rgba(222,222,163,0.8)");
gradient.addColorStop(1,'#feffcf');
ctx.fillStyle = gradient;
ctx.fill();
}
postit(300, 300, -35);
MORE INFO
Phrog, I think you know what I'm trying to do. This image shows what I want to do:
Now, the only thing is, I want to be able to pass in any width and height and angle and make the adjustment on the fly.
As an example with the following code:
var canvas = document.getElementById("bg-admin-canvas");
var ctx = canvas.getContext("2d");
ctx.arc(0,0,3,0,360,true); ctx.fill();
ctx.translate(50, 50);
ctx.arc(0,0,3,0,360,true); ctx.fill();
ctx.translate(-25, -25);
ctx.arc(0,0,3,0,360,true); ctx.fill();
I get the following image:
Now, if I add a rotate in there like this:
var canvas = document.getElementById("bg-admin-canvas");
var ctx = canvas.getContext("2d");
ctx.arc(0,0,3,0,360,true); ctx.fill();
ctx.translate(50, 50);
ctx.arc(0,0,3,0,360,true); ctx.fill();
ctx.rotate(30*Math.PI/180);
ctx.translate(-25, -25);
ctx.arc(0,0,3,0,360,true); ctx.fill();
I now have a sloped coordinates as the result is:
As I found, this is because the coordinates are no longer horizontal and vertical.
So, with this rotated coordinate structure, I can't figure out how to move my square (which could be any size and rotated at any angle) back to the left and top (so it fits in as little space as possible)
Does that make sense?
In short:
Translate the context in the Y direction only to put the corner where it should be.
Rotate the context around this offset point.
Draw your object at 0,0.
Here is an interactive, working example, which you can see online here:
http://phrogz.net/tmp/canvas_rotate_square_in_corner.html
<!DOCTYPE HTML>
<html lang="en"><head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
<title>HTML5 Canvas Rotate Square in Corner</title>
<style type="text/css" media="screen">
body { background:#eee; margin:2em; text-align:center }
canvas { display:block; margin:auto; background:#fff; border:1px solid #ccc }
</style>
</head><body>
<canvas width="250" height="200"></canvas>
<script type="text/javascript" charset="utf-8">
var can = document.getElementsByTagName('canvas')[0];
var ctx = can.getContext('2d');
ctx.strokeStyle = '#600'; ctx.lineWidth = 2; ctx.lineJoin = 'round';
ctx.fillStyle = '#ff0'
document.body.onmousemove = function(evt){
var w=140, h=120;
var angle = evt ? (evt.pageX - can.offsetLeft)/100 : 0;
angle = Math.max(Math.min(Math.PI/2,angle),0);
ctx.clearRect(0,0,can.width,can.height); ctx.beginPath();
ctx.save();
ctx.translate(1,w*Math.sin(angle)+1);
ctx.rotate(-angle);
ctx.fillRect(0,0,w,h);
ctx.strokeRect(0,0,w,h);
ctx.restore();
};
document.body.onmousemove();
</script>
</body></html>
Analysis
In the above diagram, point A is the upper-left corner of our post-it note and point B is the upper-right corner. We have rotated the post-it note -a radians from the normal angle (clockwise rotations are positive, counter-clockwise are negative).
We can see that the point A stays on the y axis as the post-it rotates, so we only need to calculate how far down the y axis to move it. This distance is expressed in the diagram as BD. From trigonometry we know that
sin(a) = BD / AB
Rearranging this formula gives us
BD = AB * sin(a)
We know that AB is the width of our post-it note. A few details:
Because our angle will be expressed as a negative number, and the sin of a negative number yields a negative result, but because we want a positive result, we must either negate the result
BD = -AB * sin(-a)
or just 'cheat' and use a positive angle:
BD = AB * sin(a)
We need to remember to translate our context before we rotate it, so that we first move directly down the axis to establish our origin at the right spot.
Remember that rotations in HTML5 Canvas use radians (not degrees). If you want to rotate by 20 degrees, you need to convert that to radians by multiplying by Math.PI/180:
ctx.rotate( 20*Math.PI/180 );
This also applies to the arc command; you should be doing ctx.arc(x,y,r,0,Math.PI*2,false); for a full circle.
You should create you canvas element and then rotate it using CSS. It would keep your canvas intact and only rotate the element itself.
Here is some example css rules:
-webkit-transform: rotate(-30deg);
-moz-transform: rotate(-30deg);
Refer to http://snook.ca/archives/html_and_css/css-text-rotation