We have got the camera panning by calculating the resultant of two vectors(previous vector before panning and changed vector after panning) and compared its magnitude with a value to restrict its position in camera view. But when we zoom in and zoom out the magnitude of the vector that geometry can cover side by side changes and hence the hardcoded value is of no help to us anymore.