Deakins
Small Canvas 2D Camera
Small and simple 2D viewport/camera management for Canvas. Named after the legendary English cinematographer Roger Deakins known for Blade Runner, Fargo, No Country for Old Men and The Shawshank Redemption.
Install
$ npm install deakins
This module is delivered as:
- ES Module:
dist/index.mjs
- UMD:
dist/deakins.umd.js
Usage
import { Deakins } from 'deakins';
const canvas = document.createElement(`canvas`);
const context = this.canvas.getContext(`2d`);
const camera = new Deakins(context);
const player = new Player();
function loop() {
camera.begin();
// Look at point in space,
camera.lookAt(player.position);
// Zoom
camera.zoomTo(500);
camera.end();
requestAnimationFrame(loop);
}
loop();
Moving the camera
camera.lookAt([x, y]);
camera.zoomTo(z);
API
Deakins(context, [options])
Initializes a new Deakins
camera instance.
context
Type: CanvasRenderingContext2D
options.fieldOfView
type: number
default: 1000
This value is used to imitate a FOV camera effect as you zoom.
options.flipAspectRatio
type: boolean
default: false
By default, everything scales based on the width of the canvas. When true
, the aspect ratio is defined by the height instead.
options.margin
type: LookAtMargins
default: {top: 0, right: 0, bottom: 0, left: 0}
Margins for all sides defined in screen space.
This is used if the lazy
option in lookAt
is true
. If true
, the camera only follows the look-at point when the point is inside the screen space margin.
camera.lookAt(point, [lazy])
Move the centre of the viewport to the location specified by the point.
Call this in the RAF loop to follow a player character for example.
camera.moveTo([11, 8]);
point
type: [number, number]
default: [0, 0]
lazy
type: boolean
default: false
Enable lazy look-at. If true, the camera won't move before the look at point is inside the margin defined in options.margin
. If false the look-at point stays centered. Below is an illustration of a lazy camera.
camera.zoomTo(zoom)
Zoom to the specified distance from the 2D plane.
camera.zoomTo(500);
zoom
type: number
default: 1000
camera.begin()
On each render pass call .begin()
before drawing to set the right transforms.
Appropriate transformations will be applied to the context, and world coordinates can be used directly with all canvas context calls.
camera.begin();
// Draw stuff here
camera.end();
camera.end()
Call .end()
when you're done drawing to reset the context.
camera.worldToScreen(point)
point
type: [number, number]
Transform a point
from world coordinates into screen coordinates - useful for placing DOM elements over the scene.
camera.screenToWorld()
returns: [number, number]
Transform and return the world space coordinate of the current look at point.
Useful if you want to project clicks and other screen space coordinates into 2D world coordinates.
camera.resize()
Call this in your resize handler to make sure the viewport is updated.
Credit
TypeScript conversion based on camera by Rob Ashton with minor tweaks. Added support for more options, lazy camera movements and slightly modified API.
License
MIT ยฉ Terkel Gjervig