)
)
)
)
)
)
| Dietrich Zawischa | Contact | Deutsche Version |
The following considerations were triggered by a line of light on a metallic hemisphere in a fountain. The question was: is the curve (ideally) a circle, i.e. an ellipse in perspective?
It is not a circle. This becomes clear if you change the lighting and observation directions – which is not so easy to do with a fountain, but is quite simple in a simulation. Instead of a compact sphere, a group of shiny meridians is observed. Experimentally, this could be realised using a shiny wire ring that is quickly rotated by one diameter. The calculation is not particularly complicated; the rays reflected from a line form the same angle with the line as the incident ray. Compared to a solid sphere, a rotating ring has the advantage that you can also obtain highlights that would be shaded or obscured by a sphere.
Stereoscopic image pairs were created in order to see the gloss curves in three dimensions. The images can be viewed cross-eyed; the left partial image is calculated for the right eye and vice versa.
But there is a difficulty: the light lines change with the viewing angle. This can make stereoscopic viewing difficult or even impossible if the light paths differ too much. For this reason, image pairs were created in which the left partial image was calculated correctly, but the same light path was calculated for the right partial image as for the left. These are the pairs shown slightly larger below. The small images next to them show the correct calculation of both partial images. Click on the images to enlarge them, select a somewhat larger distance for viewing and then squint until the images coincide.