55
$\begingroup$

I took this photograph a few days ago, during sunrise, using my smartphone camera and digital zoom:

It seems that the sun is in front of the building. How is this explained?

$\endgroup$
  • 3
    $\begingroup$ I think the diaphragm of your camera was much wide that the glow around the sun has become white and you think its the sun itself. $\endgroup$ – AHB Jan 21 '17 at 11:24
  • 1
    $\begingroup$ If you tweak the levels and knock the contrast way up, you can see that @AHB is right; the sun itself is right next to the building (the edges are just touching), and there is a distinctive glow around it (produced by the camera) which is overlapping the building. $\endgroup$ – lemon Jan 21 '17 at 11:57
  • 23
    $\begingroup$ I guess that building is much farther away than it looks. $\endgroup$ – wavemode Jan 22 '17 at 1:21
  • $\begingroup$ This is also answered on Photography $\endgroup$ – JDługosz Jan 22 '17 at 4:52
  • 7
    $\begingroup$ @JDługosz: I'll take your word for it, but it would've been a lot more useful if you'd linked to the actual Q&A answering it there. $\endgroup$ – Ilmari Karonen Jan 22 '17 at 9:38
69
$\begingroup$

You are seeing the effect of a combination of blurring - caused by an imperfect optical system - and saturation.

Blurring: ideally, a lens should map a point source into a point source. In reality, it is mapped into something slightly wider than that. If the image is well-focused, the lens is free of imperfections, and diffraction due to the finite size of the lens (aperture) is minimal, then that approximation is close to true. In your case, it's clear by looking at the image of the building that there is significant blurring: probably this is due to the image being slightly out of focus; perhaps atmospheric blur (scatter - light fog?) is contributing. There is also something called blooming where excess exposure on one pixel causes some signal to appear on neighboring pixels. All these things contribute to a blurring of bright objects in your picture (like the sun).

Saturation: the sun is much, much brighter than the building. The exposure setting of your camera is such that an intensity that is a small fraction of the intensity of the sun causes a "fully on" pixel in the image.

In the following image I try to illustrate how a point source that is imaged through an out-of-focus lens becomes slightly blurred:

enter image description here

I believe that in your image, the saturation is sufficient that the edge of the sun really corresponds to part of the blurred image of the sun, as illustrated in this next image which shows the convolution of a top hat function ("the sun") with a Gaussian of different widths ("good" and "out of focus" lens).

enter image description here

The "saturation" line shows a possible intensity at which pixels would appear white: as you can see, for the values I chose the sun would appear larger than it really is - and therefore it would appear to "cut through" the edge of the building. Which gives the illusion that it must be in front.

Here is a 2D demonstration of what is going on - I hope the picture and legends are self-explanatory:

enter image description here

I had to use log scale a couple of times because the sun is so much brighter than the "building" and background in this image.

$\endgroup$
  • 1
    $\begingroup$ This is just a great answer! $\endgroup$ – tfb Jan 21 '17 at 17:22
  • 3
    $\begingroup$ Good answer, but would not diffraction also account for this illusion? Just like the edges of a shadow from an distant object object would be blurred. Or are you accounting for that in your description of blurred? $\endgroup$ – mharr Jan 22 '17 at 18:01
  • 1
    $\begingroup$ How did you manage to generate that image? $\endgroup$ – Tomáš Zato Jan 23 '17 at 9:18
  • 1
    $\begingroup$ @TomášZato I used a few lines of Python. Unfortunately I am on the road now and the code I wrote is on my home computer... $\endgroup$ – Floris Jan 23 '17 at 11:52
  • 2
    $\begingroup$ @Lior it is possible - but don't try it! The human eye has a strong nonlinear component to detection which makes it much less prone to this kind saturation. At the same time there is all kinds of habituation - when you look away from a bright image you see it "burnt in" to your retina for a while. The sun might stay burnt in because of damage. I repeat: Don't try it. $\endgroup$ – Floris Jan 23 '17 at 11:58

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.