Lines Appear When Staring at Bright Lights - What's the Explanation?

[DrDu]

Why doesn't anybody of you guys take out his cool iPhone and takes a photo of these "interference" fringes?

 

[DrDu]

I think a similar phenomenon is encountered in microphotography, the so called "Coolpix ring artefacts". See here for an explanation:
http://www.klaus-henkel.de/coolpix-artefact-report.pdf

 

[mfb]

No, you see diffraction pattern on a screen or on a place that images that diffraction pattern. You do not see it when it passes through your eye lens directly to be imaged onto your retina. That is what is meant by that statement.

Zz.

This is true if you focus your eyes on the fingers, but if you focus it on the light source behind them you do not get an image of the slit.Just to check some numbers:

10cm distance eye<->fingers, 200µm slit width (+- factor 2), ~80cm distance from an LCD monitor, light of ~500nm wavelength, ~3mm width of the eye, ~3cm visible area on the screen.

3mm/10cm corresponds to the maximal angle between two light rays hitting the eye, this is close to the ratio 3cm/80cm. As further confirmation, the visible area does not depend significantly on the slit width, it just changes the brightness. Moving the fingers closer to the eye increases the visible area.

Using those values, single slit interference pattern have a separation of ~0.25mm at the lense. However, even the first side-maximum has an intensity much smaller than the main maximum (see here, for example). This could be somehow compensated by the nonlinear response of the eye to light, but with a bright light source (like a white area on an LCD screen), we would get the superposition of many single-slit patterns. In addition, different wave lengths would diffract in different angles, so you should see colors by looking at an LCD screen.
In the setup, the eye is adjusted to produce constructive interference for a light source 80cm away. First the regular vision, then the vision through the slit only:

As you can see, the condition for positive interference remains the same, with and without slit. No new diffraction happening anywhere. But we get a reversed image of the lense and other parts of the eye. This lead me to another test, move the fingers closer to the eye to see it clearer: Keep the separation of the fingers constant, move them in your field of vision. You will see that the lines "move" quicker than your fingers, in confirmation with the inversion in the eye. At the same time, this rules out any diffraction/interference effects from the slit between the fingers, as the position of lines would be fixed there.

I do not know what exactly produces lines. But it is certainly a part of the eye. Apparently absorption/scattering in the eye is not the same everywhere.

 

[sophiecentaur]

Why doesn't anybody of you guys take out his cool iPhone and takes a photo of these "interference" fringes?

You called my bluff so I had to do something about it!

Naturally, I couldn't get an effect when I simply tried photographing the slit. The slit image came out fringe-free and with variable degrees of fruzziness, depending upon the camera focus. I did this with a Pentax DSLR with a range of lenses and got the same result. I decided it must be something to do with the aperture. Even with an f32 qperture setting, the image looked ok. But, of course, the light path through a camera lens is complex and I am not sure how the iris actually works. So I improved the experiment with a bit of effort - but not too much (I am not a total nerd).
First, I cut a neat(ish) slice in some thin black card for the slot and made a 'pinhole' in another piece of black card.
The pinhole was taped across the front of the lens and I held the slot at various distances and used various focus settings. The best result I got (it needed to be against a white wall in bright sunshine) is attached. It shows the same sort of pattern that I see between my fingers. I guess I could improve on it with a cleaner pinhole aperture, a more appropriate slot width and, possibly a different lens.
I think the picture is good enough to show that the phenomenon is not just 'an eye thing'.

 

[sophiecentaur]

Woops - here is the atttachment.

 

[Jeremy87]

I took a few pictures too before I saw it was already done. Anyway, the more the better I guess =).

Found it works really well by looking between two credit cards or similar. For my picture I had a piece of paper pressed between the cards to force a very thin opening.
It looked *much* better in the preview image, then the camera decreased the exposure a lot when taking the actual picture, so most of the pattern disappeared. You can still see it though.

 

[sophiecentaur]

I took a few pictures too before I saw it was already done. Anyway, the more the better I guess =).

Found it works really well by looking between two credit cards or similar. For my picture I had a piece of paper pressed between the cards to force a very thin opening.
It looked *much* better in the preview image, then the camera decreased the exposure a lot when taking the actual picture, so most of the pattern disappeared. You can still see it though.

It's a lot harder to get a decent photo than to see the effect with your eye and I am sure it must be to do with the apertures of the eye vs camera. I could get nothing on a 18-55mm lens until I put a pinhole aperture in front.