Incense

by Bart Arondson

submit your photo


Hall of Fame
View past winners from this year

Please participate in Meta
and help us grow.

Take the 2-minute tour ×
Photography Stack Exchange is a question and answer site for professional, enthusiast and amateur photographers. It's 100% free, no registration required.

Cheaper zoom lenses usually are faster at the wide end and slower at the long end (for example, the $150 Canon EF-S 18-55mm f/3.5-5.6). More expensive constant-aperture zoom lenses have the same aperture regardless (for example, $800 Canon EF 17-40mm f/4.0 L).

My question is: are these good lenses sandbagging at the wider settings, or do they have a different optic system that allows them to maintain the same aperture throughout the zoom range?

share|improve this question
    
I bet that it is the former option. The potential wider aperture is sacrificed to get the convenience of constant aperture across the zoom range. –  ysap Jan 21 '11 at 0:41
3  
@ysap are you serious? Do you think that Canon designed a 17-40 f/1.7-4.0 lens and decided to limit the aperture at the wide end for the sake of keeping it consistent? A 17mm f/1.7 prime lens would impressive enough, but a zoom... –  Matt Grum Jan 21 '11 at 11:34
    
@Matt Grum - not exactly; What I think is that, contrary to your rather extreme example, the design of the lens is such that they could make it wider at the 17mm end, but rather, they think the constant aperture is more preferable. The actual potential aperture difference is rather smaller than your example. Even much much cheaper lenses have a better aperture ratio on a bigger zoom than the numbers you provided, so obviously there is some design techniques to widen the tele end. –  ysap Jan 21 '11 at 13:08
    
@ysap In that case I misunderstood your comment, I apologise. It sounded like you were suggesting that all lenses changed aperture ratio when zoomed and that Canon were artificially limiting the aperture at the wide end for the sake of consistency! –  Matt Grum Jan 21 '11 at 15:24
2  
In any case I think its more like they designed the lens to be f/4.0 at 17mm (as any wider would affect performance) and then designed it to not get slower as it zoomed in, rather than designing it to be f/4.0 at the long end and deciding not to make it faster at the other end for consistency... –  Matt Grum Jan 21 '11 at 15:40

5 Answers 5

up vote 9 down vote accepted

They have different optics and are usually substantially bigger lenses for the same focal range (compare a 70-200mm f/2.8 to a 70-300mm f/4.5-5.6 and see that the latter is small in comparison). To get the constant aperture at the long end, you need to have a bigger barrel because the aperture is a ratio versus the focal length. However, if you do the math for your examples:

18mm f/3.5 means a 5.14mm opening 55mm f/5.6 means a 9.82mm opening

17mm f/4.0 means a 4.25mm opening 40mm f/4.0 means a 10mm opening

It's clear that the aperture's physical diameter can be bigger in both cases. So, in either case, you would theorize that at the widest end you should be able to be f/2.0 or thereabouts and your sandbagging scenario would then apply to both. On the other hand, for the latter, the optics may be simplified and thus approaching prime quality in result. So... Tradeoffs.

In any case, zooms have pretty complex construction involved, much more so than a prime lens ever would, and so there are a lot of considerations around optical correction at various focal lengths, the effect of the aperture on that correction, and so on. It may be, given the lens design and costs associated, that attempting get wider on the short end would result in a hugely unacceptable softness in the image or some other forms of abberation.

Finally, for certain there are different optical constructions between the two. Heck, there's different optical constructions between lenses of the same configuration but different manufacturers. It all comes down to cost versus benefit and, in the end, what price the market will bear for a lens of a given construction.

share|improve this answer
3  
You hit the nail on the head when you did the sums on the 18-55 and realised the effective apertures at each end were 5.14mm and 9.82 the figures are similar for the constant aperture lens. The optical designs aren't that different actually, both change the size of the effective aperture when they zoom, the constant aperture one simply changes it slightly more, enough to keep the ratio of effective aperture to focal length the same. –  Matt Grum Jan 21 '11 at 11:54
5  
Don't forget that both 17 and 18mm represent extreme retrofocus lenses -- the actual physical aperture size may be substantially different from the apparent aperture. That is, the hole you see will appear to be a different size depending on which end of the lens you are looking through. The same goes for the compact tele zooms, which are true telephoto lenses at the longer focal lengths (ie, their optical length is longer than their physical length, which is not true for most of the lenses we colloquially call "telephoto" simply because of their long focal length). –  user2719 Jan 23 '11 at 12:32

Unfortunately, neither the accepted answer nor any of the others seems to do more than hint at the motivation that might be involved in choosing to design one or the other. None of them really even hints at how fixed and variable aperture zooms really differ.

There is actually a fairly fundamental difference in the design. The diaphragm (the part that forms the aperture) in almost any lens is somewhere around the middle of the lens. In a fixed aperture zoom, only the elements behind the diaphragm move around to do the zooming. In a variable aperture zoom, elements both behind and ahead of the aperture move around to do the zooming.

At least in the usual case, the diameter of the aperture does not change as you zoom. This is fairly easy to verify -- take pictures at different zoom ratios and maximum aperture with some out of focus highlights. At least with your typical zoom lens, the out of focus highlights will remain round at all focal lengths, indicating that the aperture is remaining wide open (where it's round). Stop down the lens a few stops, and you'll start to see the shape from the aperture blades closing (though lenses with lots of blades, especially rounded ones, will retain nearly-round looking highlights somewhat more than others).

When/if the elements in front of the aperture move around during zooming, you're changing the (effective) focal length of that part of the lens. You're then transmitting light through a fixed-diameter aperture, meaning the (effective) f/stop changes. Since it's only affected by the change in effective focal length of the elements in front of the diaphragm, the change doesn't (usually) correlate exactly to the change in overall effective focal length -- moving the elements behind the diaphragm changes the effective focal length without changing the effective aperture (e.g., my 28-135 has nearly a 5:1 zoom range, but the aperture only changes from f/4.0 to f/4.5).

share|improve this answer
2  
I can't say I really understand all of this, but that would mean that Canon makes unusual lenses, since all fixed zooms I have (17-40 f/4, 24-70 f/2.8 and 70-200 f/2.8 IS) appear to have parts in front of the aperture moving with zoom. –  che Jan 25 '11 at 11:46
    
I' skeptical to this. All of my constant aperture zooms from Canon and my Sigma 70-200 2.8 APO DG HSM does having moving parts while zooming in front of the aperture. Can you provide a source that supports your claim? –  Hugo Mar 5 at 12:38

Simply put, the ratio f/4.0 means the effective size of the aperture is the focal length divided by 4 - for a 600mm f/4.0 this doesn't mean there is literally a 150mm hole where the aperture blades are, only that the lens behaves as if there is. (if you look at the design of the Canon 600 f/4.0 it's clear there isn't space for a 150mm opening in the middle of the lens).

This is the principle behind constant aperture lenses, the size of the virtual aperture changes throughout the zoom range, despite the face the physical aperture clearly remains the same size.

All zooms change the size of the virtual or effective aperture while zooming, "constant aperture" (really constant f-ratio) simply change the aperture enough to keep the aperture to focal length ratio the same. The design of "constant aperture" lenses is not radically different, just the degree to which the apparent aperture changes.

To steal the numbers from John's answer (to save working them out again) the size of the virtual apertures for the two lenses mentioned are:

Canon EF-S 18-55mm f/3.5-5.6 5.14mm @ 18mm - 9.82mm @ 55mm

Canon EF 17-40mm f/4.0 4.25mm @ 17mm - 10mm @ 40mm

If the optics in the 18-55 made the virtual aperture 15mm at the long end it would be a constant aperture lens (@f/3.5) this would be very expensive however due to the [relatively] large zoom range, which is why as a cheap lens it remains f/5.6

There is no sandbagging going on in the constant aperture lens, at the wide end the lens is trying as hard as it can, it's just that it has been engineered to behave faster at the long end!

share|improve this answer
    
Actually the 600mm f/4.0 is 168mm in diameter at the end of the barrel... the-digital-picture.com/reviews/… –  Nick Bedford Jan 25 '11 at 1:11
1  
@Nick the front element has to be as big as the apparent aperture, I meant the part where the aperture blades are is narrower than 150mm. –  Matt Grum Jan 25 '11 at 12:08

My question is: are these good lenses sandbagging at the wider settings, or do they have a different optic system that allows them to maintain the same aperture throughout the zoom range?

Remember that when using an f-number to represent aperture, it's expressed as a fraction of the focal length, so as you zoom, the same aperture effective diameter is represented as a different number. f/2.8 at 20mm is half the aperture effective diameter of f/2.8 at 40mm. So your constant aperture zoom is not actually "maintaining the same aperture throughout the zoom range" as such. In fact, a 18-55 zoom which maintains the same aperture effective diameter throughout the zoom range would be something like f/3.5-10.7.

So neither type of zoom lens really maintains the same aperture effective diameter. Note that the effective diameter is not necessarily the true diameter of the aperture ring, either, since part of the zooming effect is that the aperture ring itself is magnified. But the effective diameter is what is relevant.

Lens designers battle to solve a number of problems including chromatic aberration, distortion, sharpness and vignetting. With a zoom lens, this is all the more difficult because they have to solve these problems not just at a single focal length but throughout the entire zoom range. However, all lens design makes compromises simply because there are so many opposing forces. For a zoom lens, the lens designers decide what aperture they can get away with at each focal length in the zoom range, without too much softening or other issues such as vignetting.

It's desirable for a zoom lens to have a much wider aperture effective diameter at the telephoto end than at the wide end, because as the image is magnified you need more light for the same amount to fall on the sensor/film. That is, you need it to be much wider just to reach the same f-number.

Cheaper zooms often just make more compromise on speed at the tele end than more expensive ones.

Constant aperture zooms like the Canon EF 17-40mm f/4.0 L you mentioned make a different compromise; they put a lot more effort into getting a wider effective aperture at the telephoto end. As a result, though, they use more glass and create a heavier lens. Since everything is a compromise too, they don't want their effort into getting a wider aperture at the telephoto end to increase the softness or vignetting at the wide end, so that limits the wide end's maximum aperture. So you get a different balance of aperture sizes compared to the cheaper, lighter "variable" (in reality actually less variation in terms of actual aperture diameter) aperture zoom and all it really depends on is what sort of trade-offs have been made in the lens design.

share|improve this answer

It's a matter of having the correct combination of concaved and convexed elements in place to reduce the lose of light throughout the zoom's range. While the f/4.0 might seem like your being cheated on the faster aperture, it's more a result of getting as sharp an image without chromatic aberration while keep a consistent timing and lighting throughout your zoom and focal range.

Canon has some very good literature explaining all this, along with how Optical Diffraction is being used in some of their newer lenses to counter all the prior downsides to regular optics. I'll post it as soon as I find it again.

share|improve this answer
2  
Canon's DO (diffractive optics) lenses were heralded with great marketing fanfare, but so far have failed to make a significant dent in the market. When was the last time you saw a Canon lens with a green ring? –  gerikson Jan 21 '11 at 5:55

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

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