3

I am trying to understand how specular highlights work. As I understand specular highlights are created by reflecting light-sources in a mirror-like fashion. Because faraway reflected objects are out of focus when focusing on a mirror surface, analogously I expect that specular highlights will be out of focus when focusing on an object surface. To test this out I did a "human eye" test. I looked at a close by wine bottle with my eyes and I can clearly focus either on the specular highlights or glass surface, but not both, so I surmise that understanding specular highlights as an imperfect mirror reflection seems to work.

Despite this I don't see blurry (out of focus) specular highlights when focusing on nearby objects. As a quick illustration see this image where the object camera distance seems to affect the blur of the specular highlights much more than the camera light-source distance, as evidenced by the greater blur of the specular highlights on the left apple then the right apple.

enter image description here

Why is this? Is there any way that a specular highlight is more "in focus" than the object surface it is reflecting off of (analogously to focusing on a light-source in the mirror)?


I am thinking of a couple explanations:

  1. this might be due to the convex surface of objects like apples or wine bottles, as I noticed that photographs of reflections in sunglasses have both the glasses and reflections mostly sharp, but this doesn't explain why the same effect is seen on water surfaces, which are approximately flat and as well as contradictory evidence from the "human eye" test.

  2. maybe this is an illusion due to the object surface structure being the main contributing factor to the apparent sharpness of specular highlights, but I am seriously unconvinced based on the "human eye" test and also observing the same effect on glass surfaces that should sufficiently smooth.

3

We see and photograph objects that are mostly illuminated by reflected light. Now most surfaces are irregular, thus the light they reflect is highly diffused. However, frequently, objects have shiny surfaces that are polished. Such surfaces reflect away a high percentage of the incident (about to hit) light. When by chance, a lustrous surface is situated so that the angle of incidence and the angle of reflection are closely equal, an image of the illuminant (light source) can be seen and photographed. This is the specular highlight you are asking about. A light source as seen in a mirror is revealed as a flawless duplicate, however, most surfaces in nature are not perfect, and we see a diffused reflection of the light source.

When we attempt to scale the distance to an object as seen in a mirror, we are frequently misled. We stand before a mirror 1 meter/yard distance and view our refection. The image we see gives us the illusion that we are situated 1 meter/yard behind the mirror. Do we focus on the mirror or on the image in the mirror? The specular highlight seen on the apple is an image of a light source that is far more distant than the apple. When focused on the apple, the reflected image of the light source will go out of focus. On the other hand, if the highlight is tiny, i.e. a tiny point, the point is focused correctly along with the rest of the apple's surface.

  • So do I understand correctly that the model of specular highlights is not of a "imperfect mirror" but rather more like "glitter" scattered on the surface? I don't follow the second paragraph, would even a tiny, far away point of light still not be blurred to a "bokeh ball" in a mirror if we set the focus close (like with street-light bokeh photography)? – Chris Novak Jul 11 '17 at 21:21
  • Mostly they imperfect reflections of the distant light source i.e. this light source image has dimensions (its size). However a tiny point of glitter emanates from the surface of the object, thus it's focus distance is the same as the surround object. – Alan Marcus Jul 11 '17 at 23:08
  • 1
    @ChrisNovak it may help to read a little about "BRDF," the bidirectional reflectance distribution function, to get a better idea about the makeup of specular reflections and glint. – Carl Witthoft Jul 12 '17 at 11:22

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.