Interesting Question

[bossel]

I just came across this post on dpreview:

http://forums.dpreview.com/forums/readf ... d=29665760

We all know this effect of the light rays when we photograph light sources at night. Now where do they come from? I've checked the pics, it seems all lamps have 15 or 16 light rays. This can't be random! But then, I doubt the lens he used has 15 aperture blades (In some pics the artefacts you see depend on the number of blades). So what about these lamps photographed at light? Given my lens & camera settings, could I predict how many rays I'll get?

[KevinBarrett]

I just put this image up in another thread and wondered the same thing.

Sony a700 + Tamron 17-50

There are 14 points which seem to be staggered in strength; every stronger point has a weaker point opposite it on the star. I plotted the points trajectory and they don't orient toward the center of the image circle but stay in the same orientation in relation to each other. To me, this suggests that they are an effect dependent on the aperture blades as my Tamron has seven blades. It would be interesting to compare two images at identical settings taken with, say, the Sony and Tamron versions of the 18-250 lens, as one has rounded aperture blades and the other does not.

[harveyzone]

We all know this effect of the light rays when we photograph light sources at night. Now where do they come from? I've checked the pics, it seems all lamps have 15 or 16 light rays. This can't be random! But then, I doubt the lens he used has 15 aperture blades (In some pics the artefacts you see depend on the number of blades). So what about these lamps photographed at light? Given my lens & camera settings, could I predict how many rays I'll get?

I count 14 on all of them. The only ones that aren't appear to be two lights together.

It is definitely based on the amount of blades you have in the lens apperture. I read somewhere that an even number of aperture blades gives the same number of rays, while an odd number gives double, but this could be completely wrong. I don't know the science behind it, but smaller apertures, going beyond F/16 increase the likelyhood of seeing the effect.

[KevinBarrett]

Even numbers of aperture blades produce the same amount of star points because there are aperture blades immediately across from one another--the extra points are overlapping the others. Each star point extends to either side of the highlight, but stronger in one direction than the other (at least in my Tammy), showing the staggered effect where there appears to be a weaker 7-point star inside a stronger 7-point star.

That's the "why," but I'd love it if someone could tell us the "how!"

[alphaPDX]

That's the "why," but I'd love it if someone could tell us the "how!"

The same thing happens in astronomy, with reflecting telescopes. Each 'spoke' of the the secondary-mirror support creates rays along its entire axis. Ironically then, if you have three equidistant spokes at 120 deg you get more diffraction rays (six) than if you have four spokes precisely centered (four). Telescope makers get around this by making the spokes curved, or having them intersect off center. The diffraction still happens in these methods, but the light is dispersed across more of the image rather than concentrating in the rays. Since the aperture is set to contract on-center, guess that isn't an option!

[01af]

These star points around bright light sources come from a diffraction effect that occurs at the aperture blades. Each blade will create two opposing points so the number of points will always be even.

If the number of blades is odd then there will be twice as many points---a lens with e. g. seven aperture blades will create 14-pointed stars. Straight blades as well as smaller apertures will create sharper, more pronounced points; curved blades and wide apertures will create less pronounced points. A lens wide open, i. e. with a round rather than a polygonal aperture, will create 'an infinite number of points'---i. e. just a soft flare around the light source, spilling smoothly into the dark.

If the number of blades is even then the points from opposing blades will overlap so the points will be brighter but the number of points will be only equal to (as opposed to twice) the number of blades.

-- Olaf

[bossel]

Many most helpful answers. Thanks to everybody!