The introduction of (_ScreenParams.y / _ScreenParams.x) to the 'metersPerPixel' variables has stabilised the shift under changes in aspect ratio and screen pixel dimensions. (again due to Access Denied issues from Unity Forums) These offets/translations all assume a fixed Z position in world space for the object (and to make life simpler, let's confirm that the camera has no rotation applied).Ĭontinued in next post. However, I need to be able to control the screen space offset in terms of world meters along the X and Y axis.

This will basically allow rendering with a particular perspective to a different point on the screen. However, after I have collapsed it (to clip space or whatever) I want to 'move/translate' the object on screen by X or Y number of meters in the x/y axis as shown below:

See cubes1.jpg (sorry it's not inline here, Unity Forums is giving me loads of Access Denied messages curating this post)īasically I want to use this camera's MV matrix to render, for example, the bottom-left cube with this perspective. Let's take the following cubes rendered by a camera: I have some sketches to help try and illustrate. This can essentially be thought of as collapsing the object to a camera facing plane using an initial perspective camera and then moving that in world space so that it can be rendered as a flat object by a second (or the same) camera. What I am trying to achieve is basically moving a perspective 'rasterized' object in screen space by 'n' meters in world space, then rendering it.

Athome camera forum