2D Camera in XNA

The XNA2dCamera class provides a clean, simple interface to camera functionality in 2-D games.
To use it, call ViewTransformationMatrix() and pass the resulting matrix to a SpriteBatch.Begin() call.

The full contents of the XNA2dCamera class are on the last page.

Fields and Properties

The Position property accesses the camera’s game-world position in Vector2 format. The Position property isn’t dependent on the game engine (unless you’re using polar coordinates), so it can be used with isometric/2.5-D games.

The Offset property accesses the offset between the camera’s actual in-world position and the position used for drawing. The Offset property also provides a way to change the viewing position without mucking around with the camera’s actual position.
Example:

The ScreenPosition property returns a Vector2 representing the camera’s position in screen/viewport coordinates. In the base XNA2dCamera class, the camera is always located at center of the current screen/viewport.

NOTE: “Screen coordinates”, as used above, actually refers to viewport coordinates. Viewports represent drawable areas of the screen. By default, the viewport is the same size as the screen, but you can change the size and top-left position of the viewport. If the viewport is not located at (0,0), all coordinates passed to XNA drawing functions are first offset by the viewport’s top-left position. Drawing is restricted to the viewport’s area; anything outside of the viewport’s area is clipped. I’ll post a class encapsulating viewports in August.

The VisibleArea property returns the camera’s visible area in world coordinates, centered at the camera’s current position. VisibleArea is used to cull objects from a scene graph before data is passed to the GPU for rendering. By default, the VisibleArea is set to the size of the current viewport.

NOTE: RectangleF is simply a float-based Rectangle. Position-wise, floats are more precise than integers. Plus, most XNA math objects, like the Vector2 or Matrix, are float-based, so using the RectangleF class eliminates the need for numerous unnecessary casts.

View Transformation

The ViewTransformationMatrix represents the heart of the XNA2dCamera class. It returns a transformation matrix based on the camera’s offset-position, rotation, and zoom. The transformation matrix is passed to a SpriteBatch’s Begin() method, and is forwarded on to all SpriteBatch Draw() calls made until the SpriteBatch’s End() method is called.

First, we create two Vector3s representing the camera’s offset-position and it’s screen position. Then we create a translation matrix (matrixRotOrigin) to translate coordinates to an origin based on the camera’s offset-position because scaling and rotation transformation is origin-based. We multiply this matrix by scale and rotation matrices.

Note: while the order of transformation matrices is usually important, for scaling and rotation, it doesn’t really matter what order you use: you’ll get the same result either way.

Finally, we multiply our current transformation matrix by a translation matrix that translates positions by the camera’s screen position. A translation matrix must be concatenated in the last step of producing a transformation matrix. Concatenating a translation matrix to a transformation matrix before a rotation or scale matrix is concatenated will radically change the results of applying the rotation or scale matrix.