I don't understand the embolded phrase below. I'm unschooled at physics and photography.

Starburst and Sunstar Effects

Starbursts, also called sunstars, are beautiful elements that you’ll find in certain photographs. Despite the odd names – one, a type of candy; the other, a type of starfish – I always try to capture them in my landscape photos. Here’s an example:

A photo of aspen trees with the sun appearing as a sunburst

The sunbeams in this photo are purely a result of my aperture (in this case, f/16).

How does this work? Essentially, for every aperture blade in your lens, you’ll end up with a sunbeam. This only happens if you photograph a small, bright point of light, such as the sun when it is partly blocked. This is fairly common in landscape photography. If you want the strongest possible starburst, use a small aperture. When the sun is in my photo, I typically set f/16 purely to capture this effect.

Also, the starburst effect looks different from lens to lens. It all depends upon your aperture blades. If your lens has six aperture blades, you’ll get six sunbeams. If your lens has eight aperture blades, you’ll get eight sunbeams. And, if your lens has nine aperture blades, you’ll get eighteen sunbeams.

Wait, what?

That’s no typo. For lenses with an odd number of aperture blades, you’ll get twice as many sunbeams. Why is that?

It sounds strange, but the reason is actually quite simple. In lenses with an even number of aperture blades (and a fully symmetrical design), half of the sunbeams will overlap the other half. So, you don’t see all of them in your final photo.

Here’s a diagram to show what I mean:

Sunstars with even and odd aperture blades

When you have an even number of aperture blades, the sunbeams will overlap.

Most Nikon lenses have seven or nine aperture blades, resulting in 14 and 18 sunbeams respectively. Most Canon lenses have eight aperture blades, resulting in eight sunbeams. I took the photo above using the Nikon 20mm f/1.8G lens, which has 7 aperture blades. That’s why the image has 14 sunbeams.


Due to the wave nature of light (light travels as both a wave and a particle), the wave will bend around edges (much in the way that water going through a gap in a wall will bend around the wall). The light bends in the direction orthogonal to the edge. The effect is called diffraction and the "sunbeams" are called diffraction spikes.

The effect is most noticeable when edges are straight ... aperture blades that are flat such that the aperture is a polygon. If the aperture blades are rounded, the effect isn't noticed (or at least not as much) because the curve causes the light to spread out along the curve ... rather than all light bending in the same direction and concentrating the effect. It is possible (or it was ... I used to use one) to get a filter that has very fine wires embedded in the glass. These were sometimes called starlight filtered because they turn every point of light into a "star" with diffraction spikes. There are variations on these that used etched grooves ... but these create spikes with a prismatic effect.

Some reflector telescopes (such as Newtonian reflectors -- but there are others that do this) have a secondary mirror near the front of the telescope supported by a "spider" ... usually four thin vanes that support the secondary mirror, but occasionally three. These telescopes also create diffraction spikes. There are variants that used curved vanes specifically to spread out the diffraction effect so that the spike isn't noticeable.

Every edge causes the light to bend in both directions. When you have an even number of aperture blades, the diffraction spikes from the blades on opposite sides of the aperture are overlaid on each other and they double up the effect -- creating fewer spikes but the spikes are larger and brighter. A lens with an odd number of blades will also create spikes in both directions (hence a 9-blade aperture gets 18 spikes) but since the spikes aren't paired up with a blade on the opposite side of the lens, the effect isn't intensified. This results in more spikes ... but not as large or bright.

| improve this answer | |
  • I upvoted this answer because the diffraction wave description is more accurate than light "ricochet." – Steven Kersting Jul 13 at 12:40
  • +1 This is the correct answer. I would just put in bold or otherwise highlight the phrase about the even number of blades because that is exactly the answer to the OP. – Itai Jul 13 at 15:08

The camera lens is adjustable as to its working diameter. This is accomplished using thin metal leaves that mechanically move -- thus changing the diameter of the circular light entry we call the aperture. This design is taken using the human eye as a model. The colored portion of the human eye has blue, or hazel or brown or green pigment etc. It is called the Iris from the Greek goddess of the rainbow. We attempt to replicate this action by a means of a movable iris located within the camera lens.

When we do this, the mechanism we use is moveable thin metal leaves or blades. The entry way mechanism fails to make a perfect circle; it has serrated edging. The number of scalloped edges is a function of how many metal leaves make up the iris. More expensive and thus more complicated mechanisms have more leaves, making the circular opening more rounded. Lower quality construction uses fewer metal leaves-- thus the iris opening is more scalloped.

The pattern of the sunburst effect is due in part to the number of leaves used to make the iris opening. This opening is called an iris diaphragm.

The stuff about diffraction spikes: As you know, solid objects cast shadows. These shadows are not perfectly clean at the edges. This is because as light skims by the edges of objects some will ricochet thus their paths are altered. These altered rays comingle with others that passed cleanly by. This is called diffraction and it spreads out perpendicular to the knife edge of aperture blades. The result is a diffraction spike. Since the iris is not a perfect circle, there will be a corresponding spike emanating from an edge on the opposite side of the iris. Thus the number of iris blade dictates the number of detraction spikes that appear.

enter image description here

| improve this answer | |

That drawing is misleading IMO. Each straight(er) edge of the aperture causes a diffraction spike that has two points/sides in the resulting image. I relabeled the first half to hopefully make it clearer (the asterisked numbers indicate the edge that caused the spike). In the second half I labeled the spike originations in matching colors.

enter image description here

| improve this answer | |
  • Good point about the relabeling. Also worth noting, the drawing sort of highlights this without making it explicit: the "doubled-up" stars from even-bladed apertures can tend to be visibly thicker or "heavier" (especially if the opposing blades aren't perfectly parallel, causing the overlap rays to slightly spread out as they get far from the light source), whereas the rays from the odd-bladed apertures are lighter, more "wispy", less busy. (I greatly prefer stars from odd-bladed apertures than even-bladed) – scottbb Jul 13 at 20:51

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