Macro lens aperture changes at different distances. Why?

[john_meyer14]

Hi

 

I recently bought a Tokina 100/2.8 macro lens. It seems great! I especially like the build and feel of the lens-

very smooth.

 

My question concerns the changes in aperture at different focussing distances.

 

Apparently, unlike other macro lenses, the Tokina, on a Nikon mount, shows the changing aperture in the

viewfinder - I don't know if others do as well. That's a good thing, but the thing that I find strange is that

when set to 2.8, it only shows 2.8 in the viewfinder when the focussing distance is roughly more than 12 feet.

Anything less than this, and the aperture readout keeps reducing to 5.6 at 1:1.

 

When I use a normal 2.8 prime or zoom, and it's set to 2.8, the readout in the viewfinder stays at 2.8 even if I

shoot something at, say, 5 feet. Unlike the Tokina.

 

Any ideas on why this should be so?

 

Regards

 

John

[User_276104]

It's perfectly normal. Macro lenses lose light as they focus closer, much like a variable aperture zoom.

 

I have a Nikon 105 f/2.8D, and it's really only f/2.8 at infinity. As soon as I focus closer, the aperture begins to narrow and it's f/5 at 1:1 magnification.

[ellis_vener_photography]

The elements are moving farther away from the lens, spreadign the projected image over a larger area. The result is a less light hitting the sensor or film format limited area.

[john_meyer14]

Oh, I see.

 

So, a 2.8 macro should really be labelled 2.8-5.6? Just like a variable aperture zoom? Why isn't it, then?

 

So, is my 2.8 normal zoom really at 2.8 when I focus at, say, 5 feet? If so, its faster at 5 feet than the macro, then?

 

John

[Michael R Freeman]

> "Any ideas on why this should be so?"

 

Physics. More specifically, exactly as Ellis stated. Same amount of light coming into the lens is spread over a much larger area (circular) at very close focus. You can actually see this happen with the lens removed from the camera if you project the image formed by the lens on a piece of paper.

 

All lenses do this to some extent, but with normal primes or zooms that don't focus into the true macro range, the change is usually less than 1/3 stop, so you don't see the difference.

 

Also, the reason that you see the aperture display change in the camera with some macro lenses is because they have an encoding strip in the lens that detects the focused distance, and thus can output a true effective aperture value to the camera viewfinder display. As an example, the AF 105/2.8D Micro-Nikkor focused to 1:2 will lose approximately* the same amount of light as the AIS 105/2.8 Micro-Nikkor at 1:2, so both will have approximately* the same true effective aperture at 1:2, but only the AF version will report that value to the viewfinder display.

 

*approximately because both of these lenses shorten their effective focal length as they focus closer, so at 1:2 they may each have a slightly different effective focal length.

[leif_goodwin8]

"So, a 2.8 macro should really be labelled 2.8-5.6? Just like a variable aperture zoom? Why isn't it, then? "

 

No, the F2.8-5.6 refers to the maximum real apertures of the zoom lens at the two extremes of the zoom range. The real aperture of your macro lens is F2.8 even at 1:1. What your camera displays is the effective aperture, and not the real aperture. This is the real aperture, plus the light loss due to image magnification.

 

The fact that G macro lenses have no aperture ring means that you cannot set the true aperture, only the effective one. That is not good IMO.

[lex_jenkins]

Most lenses experience some loss of effective maximum aperture at closest focus when the barrel is at maximum extension. Auto-exposure and TTL flash technology will compensate so you'll never notice with most lenses. We just don't notice it with most lenses because the loss is so minute. But you can observe a 1/3 EV loss with many lenses at closest focus. Test it against a gray card or even white wall with a consistent light source (daylight or tungsten, not fluorescent or other flickering source) in manual mode.

[alex_lofquist]

The chip built into the lens must reflect the effective aperture in order to properly calculate the exposure. Otherwise you'd have to figure the exposure loss based on the reproduction ratio of the lens. (This is the way we had to do it before cameras and lenses had all these built-in microprocessers.)

[joseph_wisniewski]

John - "So, is my 2.8 normal zoom really at 2.8 when I focus at, say, 5 feet? If so, its faster at 5 feet than the macro, then? "

 

It's probably less. As Lex pointed out, virtually all lenses, macro or not, have this problem. I don't know the mechanism of your particular "normal zoom", so I'll pick on a "classic" 50mm f1.8. Five feet is 1524mm. If you focus the lens to that distance, it will extend to 1/(1/50mm - 1/1524mm) = 51.7mm.

 

If you set to f2.8, the f stop is now 2.8*51.7mm/50mm = 2.895. That's a loss of 1/10 stop, not enough to make a difference in pretty much any kind of photography.

 

Focus the same lens to it's close focus limit of 320mm, and it's now at 1/(1/50mm - 1/320mm) = 59.3mm, and the f stop rises to 2.8*59.3/50 = 3.31. That's a difference of a 1/2 stop, and would make a minimal difference on a color or B&W print, but would screw up a color slide.

 

This is known as the "bellows factor", going back to the days where cameras used to extend their lenses using an accordion like contraption called a "bellows".

 

A macro lens extends a lot more than a "regular" lens, so you have to pay attention to the bellows factor if you're using a light meter other than the one built into the camera. Otherwise, you'd be two full stops off at 1:1, and this would ruin even a color print shot on film, let alone a slide or a digital shot.

 

As Alex points out, the chip in the modern lens just automatically displays something photographers have been calculating or looking up for 150 years...

[joseph_wisniewski]

"The fact that G macro lenses have no aperture ring means that you cannot set the true aperture, only the effective one. That is not good IMO."

 

In what way is that "not good"?

 

The effective aperture is the one you use to set exposure, the one you use to calculate depth of field, and the one you use to determine whether you're stopped down enough to cause image damaging diffraction. It's much more "true" than what you refer to as the "true aperture".

[lex_jenkins]

Ditto what Joseph wrote. I've never experienced the fluctuations with G lenses that I've observed for years with manual apertures and even non-G autofocus Nikkors.

 

The worst I've ever seen was one particular sample of the 20-35/2.8 AF Nikkor, which was probably a lemon. The stop-to-stop variations showed errors up to 2/3 EV in actual exposures. I've never seen anything remotely approaching that sort of error with G lenses. Only a couple of my more worn manual lenses show such variations, especially when I try to set the aperture between detents.

 

BTW, John, the fun ain't over yet. If you experiment carefully you'll also observe a loss of effective focal length with some lenses at minimum focus. For example, my 18-70/3.5-4.5 DX Nikkor is closer to 50mm in true focal length at the 70mm setting at minimum focus. Some complex and expensive engineering is required to minimize such variations in effective aperture and focal length.

[john_meyer14]

Guys, thanks for all that info. I found Joseph's calculations facinating. Really!

 

Is there a Table available anywhere on the web for calculating (if only for curiosity) these aperture changes with different focal length lenses, distances, and apertures?

 

John