Lens Special: Behind the Glass

Have a question about today's best lenses? Here are all the answers you need.

Since Pop Photo began its improved lens-testing program in 1988, we have done full Lab tests on 444 interchangeable optics for SLR and rangefinder cameras, both film and digital. What we've learned could fill a book-but we'll spare you, and instead answer 10 of the most common questions photo enthusiasts ask.

1. What's better, an image-stabilized lens or a fast (bright) lens?
An IS lens counteracts only your motion, but a fast lens can counteract both your movement and your subject's by transmitting more light, letting you use a faster shutter speed to freeze motion and reduce blur when handholding. IS, no matter how effective, won't sharpen a running tyke photographed at 1/15 sec.

If you shoot a lot of action, you'll want a fast lens. If you take a lot of handheld, low-light candids of stationary subjects, you'll want stabilization. Best of all, of course, is a fast lens and stabilization (in-lens or in-camera). Tests of lenses such as Canon's 17-55mm f/2.8 EF-S IS USM ($1,000, street) prove the point.

2. Does a high price buy a better lens?
If "better" means simply sharper, no. Many inexpensive lenses test with sharpness comparable to (and sometimes slightly better than) pricier lenses with equivalent focal lengths. But if "better" entails all factors-optical performance, build quality, brightness, convenience-yes, you get what you pay for.

Case in point: two Canon EF zooms with nearly the same focal range, the 55-200mm f/4.5­-5.6 EF ($210, street) and the 70-200mm f/2.8L EF IS ($1,700, street). A glance at our SQF charts shows the pricier optic earns great scores at three tested focal lengths-Excellent (70mm), Excellent (135mm), and Very Good (200mm)-while the bargain lens scores just one grade down-Very Good (55mm), Very Good (135mm), and Good (200mm). The more expensive lens doesn't win by much on distortion, either.

Is that worth a $1,500 premium? Maybe not, but consider: The L lens gives you 2 full stops' more brightness at longer settings. Besides allowing you to shoot in lower light, this lets you limit depth of field to blur out distracting backgrounds.

And, as photographer Michael Soo notes, the pricier lens has more pleasing bokeh-out-of-focus circles. See how his shot taken with the L lens (below, right) has regular, rounded, blurry spots of background light, while the bargain lens makes lumpy, asymmetrical ones (below, left). It may seem a minor point, but not to serious shooters.

Then there's build-the metal-barreled L lens will stand up to far more abuse than the plastic-barreled 55-200mm. Plus, it has image stabilization. And the rotating front filter ring of the 55-200mm will drive users of filters mad. Still, Soo says, "the 55-200mm is really easy to carry around and handle, because it's light-the 70-200mm L is at least five times heavier."

Another interesting comparison: the 18-70mm f/3.5-5.6 Sony DT ($190, street) versus the 16-80mm f/3.5­­-4.5 Zeiss Vario-Sonnar ($690, street) for Sony DSLRs. The standard Sony kit lens is such a standout that we wondered if a Zeiss could outdo it. Well, the mighty metal-barreled Zeiss was an order of magnitude sharper than the plastic Sony, due to superior control of chromatic aberration and less color fringing and edge softness.

3. Are lenses by the independent makers as good as those by the camera makers?
Often, yes. Over the years we've seen many examples where the lenses from the top three independents-Sigma, Tamron, Tokina-perform on a par with equivalent lenses from the camera makers, and occasionally even better.

For instance, we recently tested the Sigma 18-200mm f/3.5-6.3 DC OS AF zoom, which performed optically a notch better than Nikon's 18-200mm f/3.5-5.6G DX Nikkor VR AF-S ($700, street) and about on a par with Canon's most comparable zoom, the 28-300mm f/3.5-5.6L IS USM ($2,200, street). The image stabilization of the three lenses was competitive (Canon and Sigma, 2-3 stops' advantage; Nikon, 3-4 stops'). The Sigma, at $549 (street), costs $130 less than the Nikon and $1,650 less than the Canon.

Of course, like the camera makers, the independents have multiple lens lines, and their premium-grade optics can be nearly as expensive as the camera makers' top glass. Similarly, kit lenses and other inexpensive zooms from camera makers can challenge the independents on both price and quality. So know what you want in a lens, and consult our lens tests.

4. Are macros really sharper for close-up work than close-focusing zooms?
Yes. True macro lenses (which we define as able to focus to at least 1:2, or one-half life size) are more than optics that can simply be focused to less than a few inches. Their optical formulas have been devised to make sharp images at very close distances. Those designed for general photography, on the other hand, are optimized for moderate distances.

That said, the difference in performance has become narrower in recent years. A major reason: the use of floating-element systems that modify the optical formula of the lens at different focusing distances.

For instance, Nikon's 105mm f/2.8G VR and Sigma's 70mm f/2.8 EX DG Macro AF both turn in stellar performance at normal distances as well as closeup. (In fact, Sigma's may be the sharpest optic in its current lens line.)

The truly compelling reason to use a macro, though, is that they are just more convenient. You can focus to high magnifications without any extra tubes or front-mounting lenses-exactly the sort of accessories that can degrade performance of general-purpose lenses-and they are more compact than most tele-zooms. And while some zooms may provide pretty good macro performance, it's often at a single focal length, and often a very inconvenient one.

5. Is it true that single-focal-length lenses outperform zooms?
No. That was true years ago, but no longer. Within any given price range, zooms can equal or better the performance of primes at their various focal lengths.

An example is the recently tested Tokina 50-135mm f/2.8 Pro DX AF full-frame zoom (equivalent to about 75-200mm on the most popular consumer DSLRs). SQF results were Excellent at all three tested focal lengths, across the full aperture range; distortion and light falloff were very well controlled. It would be very hard to find three f/2.8 primes with that kind of performance at a budget of $700-the Tokina's street price.

But in long-lens wildlife and sports photography (and also in macros), primes still rule. While there are good, long tele zooms, above 300mm many serious shooters choose a single focal length. If you go this route, consider the 1.4X teleconverter matched to your specific lens, which gives you 40 percent more magnification with a loss of just 1 stop in brightness.

6. What's generally the sharpest lens aperture?
Usually it's right around the middle of the aperture range; traditionally, the rule has been 2 to 3 stops down from maximum aperture, and our tests support this pretty much across the board.

For example, if you look at our SQF chart for the Pentax 70mm f/2.4 Limited DA AF lens (the test is on www.PopPhoto.com), you'll see the best (that is, highest) numbers fall around f/5.6 to f/8, and they drop off from there. Turn the chart sideways, draw a line over the red zones, and you'll get close to a classic bell curve.

This surprises people who confuse absolute sharpness with depth of field, the zone in front of and behind the subject of a photo that is acceptably sharp. Depth of field is maximized at small apertures, such as f/16 and f/22. But if you examine the image closely, you'll see the point of focus is less sharp than in pictures taken at middle apertures. This is due to diffraction, an image-degrading effect exacerbated by small lens openings.

7. Do heavier lenses perform better?
Not necessarily. Some lightweight (and light-price) lenses perform very well, especially at their optimal apertures. For a typical f/3.5-5.6 kit lens such as the 18-55mm Zoom-Nikkor AF-S DX, that's around f/11-perfect for landscapes.

But there are drawbacks. Says Ed Nuñez, who used that lens for the photo at right, "It doesn't have a focus-distance scale, like pro lenses have." So he couldn't save time by setting the lens to hyperfocal distance. "I had to manually focus and use the depth-of-field preview button on my camera, or just autofocus in a section of the scene and manually move the focus until I got the feeling that everything would be right." Still, he got some nice pictures with a 7-ounce lens.

8. What is "equivalent focal length"?
In a DSLR, a sensor the size of a frame of 35mm film (24x36mm) is "full-frame," while the smaller APS-size (17x25mm, with variations) is more common. (The Four Thirds system format, 13.5x18mm, is even smaller.)

Put the smaller imager behind a full-frame lens, and it'll capture only the center of the image-in effect, cropping. Since most shooters still remember 35mm film, it's accepted practice to state a conversion factor to show what the equivalent lens for 35mm would be. (Soon people will be so familiar with their digital format that this will be unnecessary. We'll know that an 18mm lens is a moderately wide angle without converting it to "27mm equivalent.")

Equivalent focal length is not the same as coverage-the size of the image circle a lens throws onto the sensor. Coverage is determined by lens design, independent of focal length. "Digital-only" lenses designed for APS-size sensors throw a smaller image circle than full-frame lenses. If you use a digital-only lens on a full-frame camera, the corners of the picture are likely to be dark because the image circle won't cover the whole frame. Using a full-frame lens for an APS-size sensor is no problem, because the full-frame lens has coverage to spare.

9. So shouldn't I buy full-frame lenses for a DSLR with an APS-sized sensor?
Our test results here aren't very conclusive. We recently retested the full-frame 28-135mm f/3.5-5.6 Canon EF IS USM ($410, street) to see how it would do using just the APS-size "sweet spot." The difference? Statistically insignificant. Since that's a legacy lens designed for film, we also tested the new 14-24mm f/2.8G Nikkor ($1,800, street) at APS size as well as full frame. Again, the difference was meaningless.

Then we tested the new 28-300mm f/3.5-6.3 Tamron VC lens ($600, street). While it provides full-frame coverage, it's intended mostly for APS-sensor DSLRs. We found insignificant differencecs at the two shorter focal lengths, but improvements of one full grade at 200mm and two grades at 300mm.The Tamron lens also showed improvement in flare and light falloff when used on APS format.

But this isn't a given for all lenses, either. The fact is that APS digital-specific lenses can and do deliver comparable performance to the "sweet spot" of a full-frame lens. So, if you own full-frame lenses, you'll get fine results from them on your APS-sensor DSLR. But it won't pay to buy full-frame lenses for your DSLR just to get better image quality.

10. Which lens-based image stabilization system is best?
While the details differ among Image Stabilizer (Canon), Vibration Reduction (Nikon), MEGA Optical Image Stabilization (Panasonic/Leica), Optical Stabilizer (Sigma), and Vibration Compensation (Tamron), the principle is the same: A motorized lens group counteracts motion detected by gyro sensors in the lens. Our tests show that most give you about 2 to 3 stops of extra handholding advantage. Stabilization is more effective with telephotos than wide-angles, where you may gain only 1 stop. Why? Short lenses magnify less and thus show little motion blur.

Stabilization improves with each successive generation. We've gotten the best test results (up to 4 stops) from the latest Nikon VR II technology. Do not expect its rivals to let the matter rest.