Putting Together a Lens System
by Sarah Fox

Introduction

Back in the 1970's, when I involved myself seriously in photography, I started off with an old, beat-up Honeywell Pentax H1 and a 55mm f/1.8 semi-automatic lens. Remarkably, I could actually take pictures with it. I soon needed short telephoto capabilities, so I picked up a 135. Years later I also needed wide angle, so I picked up a Tamron 24 (Adaptall 2). I didn't stress over the fact that these lenses weren't the end-all and be-all of photography. I didn't know what "bokeh" meant. I simply made photographs and enjoyed what I was doing. Other people also liked and bought my photographs, and nobody ever complained about the bokeh. Wow, they didn't even care what film I used. Times have certainly changed.

Now our equipment has gotten so good and so advanced that people are absolutely crippled with angst over all their available options. They are no longer capable of taking usable pictures except with the latest, most advanced outfits that I never could have afforded when I was first naively doing photographic work back in the day. A lens that is incredibly good by today's standards can quickly be relegated to "garbage" status once it becomes unpopular. We didn't even have such lenses until recently. And with all these superb, advanced optics, it seems nobody can take a decent photo. Everyone wants and expects more from their equipment, and they pour thousands upon thousands of dollars into their camera systems in order to overcome their limitations

Where most people fail is that they assume good photography is the result of good equipment. Back in the day, we understood that good photography is the result of a skilled photographer. The equipment was of secondary importance, except for any limitations which might have been encountered. We typically bought gear only when we had a specific need for it. Nowadays, it seems people take shotgun approaches and collect all the latest, greatest lenses without the slightest clue how to use them. In fact a thread came up recently on a discussion forum to that effect: Now that I've bought all these (insanely expensive, top-shelf, professional) lenses, how and where should I use them? Another popped up yesterday: A physician who had bought a 1DX and a 24-70/2.8L II (very nice outfit) thought his camera was focusing incorrectly when he was attempting to do portraiture with mirror lockup at 1/4 sec. (If you don't immediately recognize a problem, please stop reading here. Don't buy lenses. Instead, buy books, so that you can learn to use the gear you already have. More lenses won't do you any good.)

What I propose here is a path towards building a lens outfit based on common sense. It is not written for compulsive lens shopaholics with insane disposable incomes to blow on trophy lenses. Rather, it is written for the budget-minded photographer who wants to make the best of the many superb options available, with the end goal of taking good pictures. But first, we must dispel with a few myths:

Lens Myths and Partial Myths

PARTIAL MYTH #1: Some lenses are sharp and some are not. The truth is that most lenses become sharp at smaller apertures. It is mostly at wider apertures that there are noticeable differences in sharpness between lenses. Anyway, I would say almost every lens is capable of delivering a sharp picture, as long as you know how to use it. The better lenses deliver sharp images over a broader range of apertures.

PARTIAL MYTH #2: Sharpness is what makes a lens good. Well, yes and no. The often ignored question is, "sharpness where?" Lenses are typically much sharper in the middle of the image than in the margins. The mark of a good lens is its ability to deliver corner-to-corner sharpness. However, this is not always necessary. For instance, consider a head and shoulders portrait. Sharpness in the center is important. Sharpness on the edges generally is not. What's more, a lens that is sharp in the middle and soft in the margins often also has a rather pleasing bokeh. (Ick! There's that word "bokeh." We'll get to that later.) Even beyond sharpness, there are so many aspects of a lens' image that matter. Remember that lenses get sharper when you stop them down to smaller apertures. However, there are other issues that don't necessarily go away. One of them is chromatic aberration (CA), whereby there are color fringes on high contrast edges. Some CA issues are easy to correct in image editing software, and some are not. There's also the contrast of a lens. Contrast is a result of many things, including good sharpness, good lens coatings, and good design. A contrasty lens is less likely to show flare (glare) when hit by stray light and is less likely to show internal reflections (ghosting) when strong lights are a part of the picture. Simply put, some lenses are much better behaved than others.

PARTIAL MYTH #3: It's important to have a lens with good bokeh. Bokeh? What's that? Most people will rant and rave about bokeh without having a clue what it is. Bokeh is simply the pattern by which out of focus objects are rendered. I've seen people rave about the bokeh in a photograph in which every single object is in perfect focus. It's rather funny, and I suppose people do it to impress others. I won't go into much detail on bokeh, except to say that even professional photographers don't often fret over whether bokeh is "good" or "bad." They are more concerned whether it is approproate for what they are photographing. For instance, if you want to take a photo emphasizing city lights in the background or showing a blurry object in the foreground, your best lens choice might be a lens with bokeh that most people consider "bad." This comes down to knowing your lenses and what they will do. In the end, the whole concept of bokeh is overblown. I suggest most photographers ignore this concept altogether. If you decide you should care about bokeh, first learn what it really is, learn what it comes from, learn what the down sides of "good bokeh" are (e.g. compromises to sharpness), and learn how bokeh characteristics are reversed for near- vs. far- out of focus objects. Otherwise it's an involved topic that is better discussed elsewhere.

MYTH #4: All the best lenses are prime lenses. The word "prime" is sometimes thought to mean "premium," like a "prime" cut of beef. That's not how the term is used in photography. It simply means "fixed focal length" (e.g. 50mm). The other type of lens, the zoom lens, which has variable focal length (e.g. 24-105mm), is necessarily more complicated than a prime lens, and it incorporates many more compromises to the optical formula. While it is true that prime lenses generally have higher image quality than zoom lenses, this is true more technically than in practice, and often it's not true at all. Modern zooms of good quality are essentially indistinguishable in image quality from modern primes of good quality. The differences, if any, can only be appreciated in enormous prints of a size that few people make anyway. Primes do have an important role in many camera bags, but so do zooms.

MYTH #5: Fast (large aperture) lenses are better than slow lenses. Many virtues are attributed to faster lenses (e.g. Canon's 50mm f/1.2 over the 50mm f/1.4, or the 70-200mm f/2.8 over the 70-200 f/4.0). Of these, faster lenses are thought to be sharper (but sharper how/when?) and to have better bokeh (there's that word again). They are certainly heavier and more expensive, so they must be good for something. What they give you in general is a brighter viewfinder image, almost imperceptibly better center sharpness at larger apertures, almost imperceptibly worse center and edge sharpness at smaller apertures, the ability to shoot a bit faster, and a more comfortable wallet, not stuffed with so much uncomfortable cash. In practice the differences are tiny, other than offering you larger apertures.

PARTIAL MYTH #6: It is difficult to focus with a slow lens. Certainly auto-focus (AF) is less accurate with a slower lens, but it's certainly not impossible or even difficult. And in fact focal inaccuries aren't even that consequential except at larger apertures. So while a very fast lens might give you more accurate AF, it also demands better AF in order to achieve that razor thin focus, at least when shooting wide open. When shooting at f/5.6 or f/8, it probably doesn't much matter, as long as the focus is pretty good. Generally speaking, the AF is going to be as good as the lens demands for its widest aperture.

MYTH #7: One always needs the best image quality one can afford. Perhaps this is an unstated myth, but I can tell it goes through people's heads. A thread once came up on a discussion forum in which a man was asking which of an exotic array of lenses he should use to take macro shots to illustrate his wife's book. The photographs were to be 2" x 3" in print. Fortunately I was able to convince him to save his money. I told him that almost ANY lens that would get him in closely enough to the subject (perhaps a lens on an extension tube) would perform brilliantly for those purposes. Even if the photos wouldn't stand the scrutiny of a 20 x 30 enlargement, a 2 x 3 image is tiny, and it will look sharp even if it isn't. The same holds true, to a lesser extent, for most images on the web, snapshots, etc. One only needs a very sharp lens for a very large print. If you don't intend to hang large prints on the wall, you should do some serious soul searching as to why you really want that enormously expensive Zeiss lens.

PARTIAL MYTH #8: You get what you pay for. While this is often true, often it is not. The camera manufacturers (or at least Canon -- probably the others too) often pump enormous R&D into their kit lenses, which they offer at incredibly low prices with little, if any, profit margin. They want those lenses to make the initial offering (camera + lens) as affordable as possible, while still providing high quality images. High quality images sell the camera and its entire ecosystem. So I say the kit lenses are generally very good lenses to own -- even more so because of their low pricing. In the Canon EOS ecosystem, I would say these are some of the great bargains:

• EF 24-105 L IS (the only L kit lens)
• EF 28-135 IS
• EF 50mm f/1.8
• EF-S 18-55 IS (not the non-IS version)

PARTIAL MYTH #9: Canon L lenses are superior to non-L lenses. There's a lot written about this. The L lenses (meaning "Luxury") are Canon's "professional" offerings. Many professionals use them, but they also use non-L lenses. There are many factors that must be satisfied to classify a lens as an L. Some lenses are not so optically demanding that they need all of these design features to deliver excellent optical performance. Among these are many of the prime lenses. Even some of the consumer zooms can give the L zooms a run for their money. For instance, the 70-300 IS consumer zoom can produce some stunningly impressive images with image quality very similar to the 70-200 L class of lenses, at least up to 200mm. From 200-300mm, the image quality falls off somewhat.

There are also, as yet, no EF-S (crop frame) L lenses, even though many of the EF-S lenses are very fine and well regarded. The EF-S lenses, on crop frame camera bodies, can sometime produce sharper images than their L counterparts, in large part because they do not have to produce as large an image circle. While the L's image might provide more image detail overall, much of that detail is simply "cropped" away by the smaller sensor.

There are additional factors that make an L an L, most notably the build. An L lens is built to tighter specs, has weather sealing, has some nicer design features, and overall is just a more nicely built lens. That might or might not matter to you.

MYTH #10: Prime lenses are good because they make you think. Zooms are for photographers too lazy to "sneaker zoom." These sorts of arguments have become a bit like propaganda lately. If a lie is repeated often enough, it becomes the truth.

An acquaintance of mine put it very well when he said that a zoom is a tool that divorces perspective from framing. The experienced photographer will often stake out a vantage that has an ideal perspective, by which there is a certain compositional relationship between foreground and background, and then he or she will use a focal length of lens (either prime or zoom) that will be appropriate to that vantage, framing up the shot as needed. That often means a change of prime lenses or a turn of the zoom ring. The ease of turning a zoom ring is considered unvirtuous and lazy for reasons that elude me. However, in the end, the zoom's ability to dial in 67.8mm to exactly frame up an image, vs. the nearest prime's inability to get any closer than 50mm, is a deal breaker to me. In the end, a 50mm shot must be cropped to just the 67.8mm center in order to get the desired perspective AND framing (i.e. the desired composition), and that crop more than loses any advantage the prime might have had in sharpness in the first place.

But divorcing perspective from framing is not always useful or important. Consider a large aperture, blurry background photo. When the background becomes sufficiently indistinct, perspective issues become less unimportant. On the other hand, large aperture capabilities and performance become MORE important. That's where a prime lens may shine more than any zoom ever could.

There is maybe some truth to this "laziness" myth in the way that some people use zooms. They might see something interesting from a sidewalk and zoom in on it, when maybe the better shot would require them to step off of the sidewalk and get a bit closer. However, in the way other photographers use zooms, there is actually more to think about (one additional variable) and more to do. This doesn't make zooms more virtuous, though. It simply delivers more options to the photographer who knows how to use them. And remember that "more options" can include framing up a scene from a mountain trail, when stepping off the trail would mean falling to one's death!

PARTIAL MYTH #11: Prime lenses are lighter, smaller, and cheaper. This is only true if you do not want many lenses. If you want to be a Henri Cartier-Bresson and shoot with only at 50mm, then it is true you can do it smaller, lighter, more cheaply, and better with a 50mm prime than with a zoom. However, if you want to use maybe a 24, 50, and 100 instead of a 24-105 zoom, be prepared to find that your kit of 3 primes is actually bulkier, heavier, and more expensive than the one zoom would be.

My Advice

My advice is to start small. Start with the kit lens that offers good quality at a good value. Use it for a while, and get to know it. If you're like many photographers, you may discover that your lens cannot do something you want to do. Maybe it won't give you closeups of the zoo animals. So then you research out your next lens based on that REAL need you've encountered. Buy your lenses one at a time, and buy just the capabilities that you really need. You can save your money not spent on lens buying sprees to acquire better lenses (that you really need) in the future.

When you acquire lenses, think about where you might be taking your outfit in the future. Above all, consider whether there might be a full frame camera in your future or whether your needs will always be better suited by APS-C (crop frame) cameras. To help you make this decision, I wrote this article comparing formats. If you size yourself up as a crop-only photographer, you should probably consider only crop-frame lens offerings wherever there's a choice. I discuss this in the article on formats. However, if you're going to be using a full frame camera, it's probably better to pony up for a full frame lens, even if you don't have the camera yet. Unlike camera bodies, lenses will stay with you for a very long time. Buy what you will eventually need, so that you don't have to buy twice!

Before you go acquiring lenses, though, you should have a strategy in mind as to how your collection is going to fit together. I've been very tight in my own strategy, and although it's certainly not "the correct" strategy or the "only" good strategy, it might still be useful for you to understand it. Then you might develop your own strategy that is based on hard decisions as to what you do and do not want and need. Here's the thinking that goes into my own strategy:

First, I examine what I want to shoot. This is a critically important first step. Usually when a novice asks an experienced photographer what lenses to get, the experienced hand asks the novice this very question, and the answer is usually of the kitchen sink variety: portraits, landscapes, sports, and general sorts of pics. Well, that's fine, but the answer needs a bit more refinement. One could simply throw a couple hundred dollars at a cheap 18-200 hyperzoom and shoot all of these things. However, the real question that should be asked is which of these things you want to photograph well. That's an entirely different question.

My own answer to this question is rather broad, because I'm a professional. I have to meet a variety of other people's needs. However, certain themes emerge: I don't really do weddings. I don't often do much sports photography. I don't often shoot little birds. These are the things I may do on occasion, but I don't need to do them well. On the other hand, I often do events and documentary coverage. I do portraiture. My artwork is mostly candid portraiture and landscape work. And I also do architectural photography. These are all the things I need to do well. My longest telephoto needs for the work I want to do well perhaps extend to 200 mm, for candid portraiture. My widest angle needs would be for landscape and architectural work, and they can be quite wide -- often really wide. So my upper tier lenses need to cover from silly-wide to about 200mm. Also as a professional I need backup lenses, so that if my main lens is out of commission, my clients aren't potentially left in the lurch. I satisfy my backup needs with second-tier, consumer grade lenses of a decent, but not stellar quality.

Am I interested in photography beyond 200mm? Sure. However, it's not much of what I do professionally. I'm not going to blow $13,000 on a 600mm f/4L IS lens, because I'm not a bird photographer or a sports photographer. Sure, I'd enjoy using one if you gave it to me, but I'm honestly content to fiddle around with an old, manual-focus Takumar 500mm f/4 lens with a Komura 2X teleconverter, for a total investment of around $500. I probably wouldn't even invest that if I didn't have a view out my back window that birders would die for. But that's it. I couldn't even begin to justify a serious, professional-grade supertelephoto lens in my equipment budget. I also have a 70-300mm IS (non-L) lens as a backup to my much nicer 70-200mm f/4L IS lens. This gets me to 300mm for my "goofing around" photography, while simultaneously satisfying my backup needs.

Now that I've identified the focal lengths I need, the question remains of what lenses I select in that range. The answer to that falls back on a matter of personal style. Finding one's style takes time and experience. Remember when I suggested you start out with a kit lens and see where it takes you? Well, this is where it pays off. Take a look at all of your photos, and note what focal lengths and apertures you most commonly used. This information is contained in the EXIF data of each photo. You will want to put most of your money where you do most of your shooting. For me, my preferred focal length range is silly-wide to normal, and my preferred apertures tend to be rather small -- f/8 and smaller. However, I do sometimes shoot telephoto, and I do sometimes shoot with large apertures.

Because my work is most typically smaller aperture (with abundant depth of field), the sharpness of the lens is relatively less important, because most lenses start looking a lot alike from f/8 and smaller. So this tiny, almost imaginary advantage prime lenses have over zoom lenses is irrelevant to me. So is the speed of the lens. The compositional aspects of having just the focal length I need is extraordinarily important to me, and that need can only be satisfied with a zoom. Because I hate weight and shoot with smaller apertures, an f/4 zoom is a good fit for me. Other aspects of performance are also quite important, such as a control of chromatic aberration. I always pay very close attention to this factor in selecting a zoom. Finally, I hate being constrained by tripods, so I feel that image stabilization is the greatest invention since the light bulb. Oh, and remember how my work is mostly wide angle? Then it might not be surprising that my primary lens lineup includes 1 wide prime (fisheye), 2 wide zooms, 1 normal zoom, and one short telephoto zoom.

I'm pretty happy with f/4, image stabilized zooms, but what happens when a client says, "Hey, can you take one of those pictures of me with a really blurry background?" Am I going to say I think that style is somewhat trite and is getting long in the tooth? No, I'm going to say, "Absolutely! I can do that!" Can I do it with an f/4 zoom? Well, yes. I can find a place with plenty of space for separation between photographer, subject, and background, mount up my 70-200 f/4 IS, dial in 200mm at f/4, and snap away. However, this need is probably better served with a fast lens -- a prime lens. So I have a 100mm f/2 lens for just that purpose. It's a good general portrait length for the fuzzy background type of shot, and it creates plenty of blur -- more than a fast zoom. Do I care about the compositional relationships between foreground and background when doing this? Absolutely not. So I don't need every focal length of prime. I just a good length of prime for a head and shoulders shot to fill that one need.

What about all the apertures in the middle? Might I want a slightly blurred background at f/5.6 that might be slightly sharper with a prime than a zoom? Nah, at least not enough to commit to a bag full of primes for it. First of all, slightly fuzzy looks somewhat like a focusing error or a bad decision in depth of field to my eye. It's seldom a good thing. Second, when a background is only slightly blurred, there is still a background/foreground compositional relationship that often demands a zoom.

So in a nutshell, my strategy is to have high quality f/4 zooms from silly-wide to 200mm, with image stabilization wherever I can get it, and with the addition of occasional primes for special needs. What makes this strategy efficient is that I recognize the different compositional demands of smaller aperture and larger aperture photography, and I fill what I recognize to be only occasional demands as surgically as possible.

Is it OK to make different decisions from what I've made? Absolutely! There's no right or wrong here. Not everyone composes pictures the way I do. If you don't give a flying fig about foreground/background proportioning, which most people don't, then a zoom becomes considerably less important. If you want to photograph little birds in distant trees, you'll need a very long lens of very good quality -- most likely a prime. If you only shoot from tripods, don't waste your money on image stabilization. If you love keeping your aperture wide open, consider investing your resources into lenses specialized for the highest optical performance at wide apertures, such as the 24-70/2.8L II. There are so many considerations that are highly individual in nature.

The only rule I would make for you, if I could, is not to waste your money buying capabilities you don't actually need. Along those lines, I would urge you to ignore the dogma, to view the dizzying array of lens offerings with fresh eyes, and to pay attention to the lens myths that I've briefly discussed. Above all, always understand what you're buying and why you're buying it. If you don't know the reason why, then you really shouldn't be buying it.