Starburst seen from points of light in night photographs

[MotoMike]

Hi
been a while, but you guys always square me away.

I've recently noticed that when a night time photo is taken with the aperture closed down, (higher f number) seen in the resultant image is a starburst radiating from points of light. The general wisdom is that it is cased by diffraction over the aperture blades. accepting that that is true, why does the resulting starburst exhibit the same number of spikes as the number of aperture blades if they are an even number, but double the number of spikes as the aperture blade number if they are an uneven number.

I've looked in the photo forums and though discussed it is not really explained.

Kind regards,
Mike

 

[jerromyjon]

why does the resulting starburst exhibit the same number of spikes as the number of aperture blades if they are an even number, but double the number of spikes as the aperture blade number if they are an uneven number.

I would have to say that in the even case they are overlapped, meaning it is always double, it is just that the opposites coincide.

 

[MotoMike]

this makes some sense to me. Can you explain just what the light is doing to present this result? I can see it bending around the aperture blade and creating one spike but why the opposite? is it a function of the optics beyond the aperture blades? In lenses that have nine blades, the aperture seems almost round to me and I wonder what it is about the intersection of the blades that light finds different than the space where they do not intersect?

 

[Andy Resnick]

Yes, the reason why you see the even/odd effect is from symmetry: even numbered means the two opposite sides of the bright spots' diffraction patterns coincide. The math is not too difficult to work out, the diffraction pattern of a square aperture is standard (advanced) undergrad homework problem. Going to hexagons, heptagons, etc involves a little more geometry but isn't an unreasonable problem. Going from straight blades to curved blades, OTOH, would be rather tedious.

 

[MotoMike]

Andy I appreciate your help. I hate to admit that the advanced undergrad math is probably beyond me. I am just an old curious guy. I am not in a physics program, just curious about what causes this to happen. In my mind I likened it to rf passing over a knife edge being diffracted into the shadow area behind the knife edge. My previous readings on that subject never indicated that it diffracts in other directions. Light being a higer frequency form of energy, I expected it to diffract and result in the one spike of the starburst when passing over the aperture blade, but was surprised to see it also appear on the opposite side. When Jerromyjon suggested that it always produced the two spikes, it made perfect sense to account for seeing double the number from an odd blade count. All that said, is it possible to explain what is happening without doing the math?