Reflection across a plane
The unit normal and one point on the mirror uniquely define the virtual image position.
Optics 002 · Ray worlds, boundaries, and natural optics
A labeled mirror gallery constructs single-mirror, corridor, and kaleidoscope image orbits with exact planar reflection. Real and virtual objects share one geometry and scale; color, opacity, physical light paths, and backward construction lines explain what is real.
Physics tutorial
BackgroundA plane mirror image is a geometric reflection of the object across the mirror plane: . Repeating that operation creates image orbits for two mirrors; finite mirror panels then decide which backward construction rays can reach their apertures.
Why it mattersWhich virtual images can an observer actually see as mirrors move?
Start with the essentials
The unit normal and one point on the mirror uniquely define the virtual image position.
The image is upright and the same physical size as the object; its distance behind the plane equals the object distance in front.
This angular orbit count is exact only in its stated divisibility and placement cases; the apparatus labels it as ideal.
Typical misconceptionA virtual ghost sends rays outward from its displayed position.
Better mental modelThe dashed backward sight line is a construction. Real light reaches the observer only after reflection from a real mirror panel.
Move the observer laterally in the single-mirror scene and compare the object and image distances to the mirror plane.
What to observe: The image remains equally far behind the mirror and exactly the same height as the object even though the observer sight line changes.Change the corridor angle and follow successive image generations away from the real object.
What to observe: Smaller mirror angle increases ideal orbit order, but finite panels reject construction intersections outside their bounds.Use the kaleidoscope scene to test rotational closure and observer-aperture candidates separately.
What to observe: A closed angular pattern can contain more ideal images than the finite-aperture candidate count.