3d-camera-core

3d-camera-core

A common interface for 3D cameras. This module is a caching layer for maintaining coordinate system transformations and computing camera properties from a set of generating matrices. This module is intended to be used as a common interface and should not be required directly.

Notes on coordinates

By convention, we will define 4 different 3 dimensional projective homogeneous coordinate systems:

• Data coordinates: The coordinates used by models and 3D data
• World coordinates: A common coordinate system for all objects within the scene
• Camera coordinates: The coordinate system in the world where the camera is at the center
• Clip coordinates: The device clip coordinate system

These coordinates are related by a set of three homogeneous 4x4 matrices:

• Model matrix: Maps data coordinates to world coordinates
• View matrix: Maps world coordinates to camera coordinates
• Projection matrix: Maps view coordinates to device clip coordinates

The goal of this module is to maintain the relationships between these coordinate systems as matrices and to define a standard interface for renderable objects which need to consume camera information. Implementors should take this module and hook up whatever methods they want to compute the model/view/projection matrices, while users can then treat the resulting camera interface as a black box handling the various coordinate system conversions.

User side

For most users of this module, you only need to worry about the stuff in this section.

User example

//You should call some other module to create a camera controller
var myCamera = createCameraType()

//Once you have a camera, then you can access the coordinate conversions directly
var dataToClip = myCamera.data.toClip

//You can also access the origin of the camera in any coordinate system too
var eyePosition = myCamera.world.origin

User API

The overall goal of this module is to keep track of conversions between a number of different coordinate systems. Because multiplying and recalculating these conversions is expensive, this module caches this data for future use. After a camera object has been created, no further memory allocations are performed.

Coordinate system conversions

The most basic function of this module is to provide a convenient syntax for getting whatever camera transformations you need.

coords.toClip

A 4x4 matrix representing the conversion from coords into clip coordinates.

coords.toCamera

A 4x4 matrix representing the conversion from coords into camera coordinates.

coords.toWorld

A 4x4 matrix representing the conversion from coords into world coordinates

coords.toData

A 4x4 matrix representing the conversion from coords into data coordinates.

Origin

coords.origin

The position of the camera in the coordinate system.

For implementors

A camera implementation should provide one or more "controllers" for each of the model, view and projection matrices. Each controller is an object with two methods; one which tests if the controller has changed and one which reads out the state of the matrix for the controller.

Implementation example

var createCamera = require('3d-camera-core')

//A simple implementation of a camera controller
function simpleController() {
  var data = [1,0,0,0,
              0,1,0,0,
              0,0,1,0,
              0,0,0,1]
  var isDirty = false
  return {
    dirty: function() {
      return isDirty
    },
    get: function(m) {
      isDirty = false
      for(var i=0; i<16; ++i) {
        m[i] = data[i]
      }
    },
    set: function(m) {
      isDirty = true
      for(var i=0; i<16; ++i) {
        data[i] = m[i]
      }
    }
  }
}

//Create a set of controllers for the camera object
var controllers = {
  model: simpleController(),
  view: simpleController(),
  projection: simpleController()
}

//Return camera
var camera = createCamera(controllers)

Implementor API

Constructor

var camera = createCamera(controllers)

This creates a new camera object with the given controllers. controllers is an object with the following properties:

• controllers.model a controller for the model matrix
• controllers.view a controller for view matrix
• controllers.projection a controller for the projection matrix

Returns A new camera object

Controller interface

Each controller is an object which provides two methods:

controller.dirty()

This method should test if the state of the controller has changed since the last time controller.get() was called. If it has, then the matrix value will be recomputed.

Returns true if the camera matrix has changed, otherwise false

controller.get(matrix)

This retrieves the state of the controller's matrix. The result should be written into matrix

Methods

camera.setController(matrix, controller)

Replaces the controller on the camera for matrix with controller.

• matrix is the name of the matrix, which is either model, view or projection
• controller is the new controller for the matrix

Legal

(c) 2015 Mikola Lysenko. MIT License