The inconvenient truth about zoom lenses. The best amateur lenses for Canon cameras

17.08.2013 10569 reference Information 13

We have already talked on our resource about lenses with a fixed focal length -. And we completely undeservedly ignored the most common lenses among photography enthusiasts, the lenses with variable focal lengths, which are standardly equipped with the vast majority of cameras - zoom lenses. These optical devices for photography allow you to zoom in and out of the shooting scene without leaving the spot. Despite the obvious convenience of this function, you should know about the features of their use in order to get maximum results.

According to the principle of operation, zoom can be optical and digital. Digital zoom has nothing to do with the lens, and we won't cover it in this article. The focal length range of zoom lenses (ZZ) is indicated on the body as follows:

Which means: the focal length can vary from 14 to 42 mm in the first image and from 18 to 55 mm in the second. Dividing the second FR value (as photographers say in slang - at the long end) by the first, you get the zoom value or lens magnification. (For example, for the lens in the first picture - 42/14 = 3.5). However, we are more interested in the angle of view, which determines what is visible that will fit in the frame. The viewing angle is strictly related to the focal length and the physical size of the matrix (not to be confused with the number of pixels!). The longer the focal length, the narrower the angle of view and the greater the zoom:

Nowadays there are a large number of cameras on the market with different physical matrix sizes: full-frame (FF - “full frame”, exactly corresponding to the frame size of 35 mm film), APS-C (“cropped” matrices of amateur DSLRs and prosumer compacts), etc.

Physical dimensions of digital camera matrices

In order to compare the viewing angles of lenses of different cameras, it was necessary to introduce the concept of “focal length equivalent to 35 mm film” or “equivalent focal length” (EFL). At the same time, the characteristics of the lens indicate two focal lengths - real (which is written on the lens) and EGF - for comparison with other lenses.

For cameras with matrices smaller than the size of 35 mm film, when choosing a lens, it is necessary to take into account the coefficient that increases the focal length written on the lens - the Crop Factor. For APS-C matrices, for example, this is usually 1.5 (1.6 for Canon cameras), which means that a lens with a focal length of 50 mm indicated on the body, when installed on a camera with a reduced APS-C matrix, will work as with a 50x1.5 AF =75 mm.

The table shown below shows approximate focal lengths for shooting scenes. These values are arbitrary; the boundaries of the indicated ranges should not be perceived as dogma.

Table - Focal lengths are equivalent to 35 mm film

Shooting plot
Landscapes, panoramas	16-28 mm
Portrait, still life, reportage	50-85 mm
wild nature	100-500 mm

Thus, for most scenes, one zoom lens with an EGF of 24-70 mm is quite enough. If you pay attention, it is exactly this, or with a similar range of FR, that is offered in kits for cameras with interchangeable lenses (whale lenses with a FR of 18-55 mm for amateur DSLRs have an FR of 27-82 mm).

How does a lens change focal length? By changing the position of the lenses. Because of this, even in small cameras, lenses with a large FR range have to be pushed far out of the body. This is not always convenient, as it takes additional time and reduces reliability.

It is impossible to select lenses in a lens that work ideally when changing positions with each other. Those. A zoom lens can show acceptable performance only in a narrow FR range, despite significantly greater capabilities. In the rest of the range, distortion is inevitable. To ensure sufficient quality for any lens position, the optical formulas of zoom lenses are complex, with a larger number of lenses compared to prime lenses. In addition, the extension mechanism can break, especially if sand gets into it, not to mention moisture and its companion - mold. The sealing elements of a zoom lens tend to wear out and lose their elastic properties over time. You've also probably heard about play in the retractable parts of the lens, which does not contribute to image quality due to distortion of the geometric axis of the lens group in the lens.

Optical designs of fixed and zoom lenses. Pay attention to the number of lenses

If the FR range is large - the so-called super-zoom lenses (18-120, 18-200, etc.) - a price for versatility is inevitable: a decrease in light transmission, an increase in internal scattering, a decrease in picture detail and significant geometric distortions of shooting objects at wide corners. You may have noticed that when photographing a group of people in a cramped room at a wide angle with a zoom lens, the faces of those standing on the edge become elongated and rounded, and the people themselves become tilted. Who will like this? The same applies to photographing architectural structures. The cheaper the lens and the larger the FR range, the more distortion and other defects listed above when shooting. This is especially true for universal superzooms.

The above equally applies to a camera with a non-replaceable lens that has a large FR range - the so-called all-mirror cameras (or ultrazooms).

So, you should keep in mind: The greater the range of focal length changes (as they say, the greater the zoom or degree of magnification), the greater the likelihood of distortion at wide angles, internal light scattering and deterioration in image detail.

It is worth noting that if you are shooting for posting online or the size of the pictures for printing does not exceed 10x15 cm, then you don’t have to worry too much about detail - the quality of the pictures will be acceptable. But light scattering and distortion cannot be eliminated by any computer modifications.

The use of long focal lengths (close zooms), regardless of the principle of operation of the zoom, is fraught with two more dangers.

1. The closer the zoom, the less stable the camera position. This is especially felt when the camera is small, which is difficult to hold level. The subject begins to jump, dance, and disappear from the frame, which when shooting will lead to blurred images and defects.

To avoid this phenomenon, fast shutter speeds should be used to obtain a stable image. But here lies another pitfall.

2. As the focal length increases, the light transmission of the lens decreases. That is, the photographer finds himself in a double hole: the shutter speed must be shorter, and the light transmission must be low. As a result, the camera will not have enough light to make the correct exposure. And if there are no problems on a sunny day, then in cloudy weather, at dusk, shooting with long focal lengths poses a big problem.

Manufacturers have taken care of users by offering lenses with built-in image stabilizers, which is usually indicated in the description and. These devices allow you to expand the range of illumination by approximately 3 steps, at which high-quality images are obtained.

When choosing a zoom - is it worth buying universal superzoom lenses, preferring mobility, or high-aperture optics with a moderate range of photons, putting quality first - there is no definite answer. Everyone decides for themselves, based on shooting conditions, required quality and other criteria.

Explore the capabilities of your equipment by taking test images with different phased arrays and for all the listed situations. This way you will know exactly when and how to use different settings to get high-quality photographs.

All the photography to you!

A reporter's mid-range, high-aperture zoom lens is the standard tool of a reporter's work.
How many of them have passed through your hands? ... and then I decided to understand their history: I’ve been trying to understand long-focus zoom lenses for a very long time, super-fast, portrait-reportage and ultra-wide-angle lenses too. Let's define the group boundary: aperture above f/4.0, range of standard focal lengths.
(Once again, this is a “debriefing” for myself, so that it’s in order in my head; if anyone disagrees with anything and can correct and complement me, I will only be grateful).

The first such lens was built for a 35mm movie camera in 1936 by
Bell and Howell Cooke "Varo", 40-120/3.5.
Cooke Optics Ltd. and today he builds amazing cinema lenses, many on the PL mount, with which digital Canon and Nikon are now compatible.
The first European serial photographic lens that can be called a “fast standard zoom” was the Kilfitt Voigtlander-Zoomar 36-82/2.8 built in 1959 for Bessamatic/Zeiss Icarex/Alpa/Exakta; it already had a jumping aperture with an external mechanical drive.

In 1963, the first standard Zoom-NIKKOR Auto 43-86/3.5 was hard-wired into the Nikkorex Zoom 35 SLR camera.

Interchangeable lens with variable focal length P. Angenieux zoom 45-90/2.8, developed in 1968 for the Leica-R - Leicaflex SL DSLR. It was with him that everything began in earnest, apparently...
Appears in December 1973
OPF from
Canon FD35-70/2.8-3.5 S.S.C.
The delight of reporters and the misunderstanding of build editors... it’s very convenient to work, but there are many complaints about the result of the shooting. The lens actively catches glare despite the progressive coating: the glare is not visible at point-blank range, although the color is good.
Nikon answers 1976: Zoom Nikkor AI Auto 43-86/3.5 was not originally designed for professionals and artists, it is a convenient universal lens. But in the same year Nikon “caught up” with Canon - they released the Zoom-Nikkor Ai 35-70/3.5 Macro.

This lens had a “long life and glorious offspring”: the Ai 35-70/3.5 Macro laid the foundation for the competition between Canon and Nikon, which we still see today.
In 1979, Canon changed the FD mount to the newFD: there was more plastic... but a serious response to Nikon’s reportage 35-70/3.5 was starting. Without plastic, as before:
New FD 35-70/2.8-3.5 caused very positive emotions.
The problems are gone, the convenience remains.

Then another lens of the group we are interested in appears:
New FD 28-50/3.5. In the same “serious” design, but with a more “wide-angle” zoom range. I understand that many readers are now going to throw something at me, but I am quoting words and impressions from 1979...
Around the same time, standard compact zooms appeared for the Olympus OM system
Olympus OM Zuiko AUTO-Zoom 35-70/3.6, and,
in 1.5-2 years - Olympus OM Zuiko AUTO-Zoom 35-80/2.8 ED.
The first caused considerable criticism, but its newest version survived until the entire OM system was discontinued, remaining one of the best zoom lenses in the world in terms of color reproduction.
In 1979, the French concern named Pierre Angenue released a photo lens
Angenieux 35-70/2.5-3.3 (-3.6), showing HOW to build such lenses. Based on cinema experience, a compact and ergonomic lens was built with such resolution and resistance to backlight that all manufacturers had to put engineers to work.

In 1982, the improved autofocus Angenieux 28-70/2.6 appeared. The cost was very high, the price was at the level of developments of German optical monsters. But the quality of the result is the highest. Focusing speed is slow...
However, Angenieux did not intend to produce a lens that was not his own profile for a long time:
the attempt to enter the photo market was not particularly successful, just at this time all the European manufacturers of photographic equipment became “severely ill”, and Japanese products took their place on the market pedestal. The respected Japanese company Tokina approached Angenieux with a proposal for licensed production of 28-70.
In 1988, the first-born professional with manual focusing went into production:
Tokina AT-X 280 Pro I/II 28-80/2.8, the flaws of licensed production disappeared in the model with Roman II after a year.

And in 1990, production of the autofocus Tokina AT-X Pro 280 AF 28-70/2.6-2.8 started. It is from this period that the legend dates back to the fact that there was a poster hanging in the workshop (the habit of rationalizing production every minute was already in the blood and driven into the heads of Japanese workers and engineers) with a reminder in Japanese:

"For attempting to make improvements to the design of the AT-X Pro 280 AF 28-80/2.8 - dismissal within 24 hours."
In 1993-94, a somewhat simplified Tokina AT-X 287 PRO SV AF 28-70/2.8 appeared - apparently, the “hard” license had expired.
Simultaneously with the Angenieux 35-70/2.5-3.3 (-3.6), in 1979 Nikon launched the ultra-compact “youth” camera Nikon EM and introduced a line of simplified and lightweight optics. In 1982, a Nikon Series E 36-72/3.5 lens was added to it - at the same time the magnificent Nikon FG was introduced and, most likely, it was a gift for him.

And after 2 years, a very small and short-range Zoom-Nikkor Ai-S 28-50/3.5 appeared. Like the 36-72, it had a “trombone” control circuit. But - the most excellent optics! In 1981, the 1979 lens was improved to Ai-s: Nikon Zoom-Nikkor Ai-S 35-70/3.5 Macro. It has become more compact (62mm filter instead of the previous 72mm) and a little lighter.

On its redesigned base, the Nikkor AF 35-70/2.8D was built in 1987, a neat and accurate reporter’s instrument. Unless - not for sports. It is still respected and popular today. Canon finds itself in the role of catching up. In April 1989, the Canon EF 28-80/2.8-4L USM lens was added to the EF optics series for EOS cameras. And today the “four” is remembered by professionals and connoisseurs as a parfait from a French restaurant - by gourmets. Its portrait characteristics were excellent, and most of the complaints revolve around the not-so-perfect autofocus system of the early 90s. However, users liked the convenience of “electric manual focusing” - without direct mechanical connection.

So, Nikkor AF-S 28-70/2.8D IF-ED appeared. Fast and accurate. The championship in the reporting field was taken away from Canon in 1999 and, with the growth of the SWM line, it began to return to Nikon: the “white” photo reporter stands at sports competitions (due to the color of Canon long-toms) began to rapidly turn black.

Canon responded by expanding the zoom range:
The Canon EF 24-70/2.8L USM was popular, but due to vignetting and distortion, it was not very popular. Nikonovsky 28-70 turned out to be better...

And at the end of the summer of 2007, Nikon AF-S FX N Nikkor 24-70 / 2.8G ED was released, which turned out to be ideal for both film and digital photography. In the hands of the reporters, the 24-70G seemed to be firmly attached. The only thing that puts me off about it is its excessive length: comparable to the Nikon AF 180/2.8D IF-ED.

We had to wait another five long years for Canon's answer: the Canon EF 24-70/2.8L II USM is now on sale; it appeared in Moscow only in November 2012, 7 months after the announcement of the start of production. The lens is magnificent and fully corresponds to the high level that Nikon set in 1999. But its price is appropriate.

Perhaps because of the price, it was decided to release the “younger brother” 24-70II onto the market - more compact and less fast, but (!) - a stabilized universal zoom

Canon EF 24-70/4L IS USM. Digital cameras have gained the ability to shoot decently at 6400ISO, so there is something to balance the loss of aperture.

So, you just bought the camera of your dreams, received all the due congratulations on this event and asked the logical question: what’s next? Which lens should you choose first? And which one for the second? How to form the best arsenal and maintain balance in the categories of “price” and “quality”? We have accumulated 5 years of experience, several dozen hours of detailed study of the issue and several interviews with professional photographers in order to recommend the four best lenses for a beginning photographer with good ambitions.

In more detail:

In this article we will talk about lenses for a regular Canon DSLR camera (from the English. Digital single-lens reflex camera - that is, qi DSLR camera). Manufacturers of DS LR cameras and third-party companies produce a huge number of lenses for their cameras, and you can choose between them for as long as you like. But that’s why we’ve gathered here, so as not to drown in the swamp of choice, but to take a confident path in the formation of an optical arsenal, avoiding mistakes and wasting time and money.

Your Canon camera will probably come with the EF-S 18-55mm f/3.5-5.6 kit lens - it's a good standard everyday zoom lens for exploring all aspects of your camera. It makes sense to shoot with it for some time in order to understand all the features of the “body” and identify your format, i.e. favorite genre of photography. There are traditionally four main formats: macro photography, close-up photography (the so-called zoom), landscape (wide-angle) and portrait. Each genre requires a lens specially “tailored” for it, and here are our confident recommendations from the top four.

The best primer for a beginning photographer

The easiest way for a beginning photographer to develop his skills is to work with lenses whose focal length remains constant. Yes, you lose the ability to zoom out or zoom out on something, but this design allows you to achieve amazing photography quality at a very reasonable price. All other lenses are essentially a search for a compromise between the ability to change the focal length (bringing the subject closer or farther away) and the desire not to degrade quality. As a result, the better the photo quality becomes, the higher the price per lens. In the case of primes, the photo quality is immediately excellent, and the price is one of the lowest.

Separately, it should be noted its appearance: the white color of Canon lenses is usually characteristic of the professional line of lenses, which are very, very expensive. However, in the case of the Canon EF 70-200 mm f/4L USM, we have professional shooting quality for quite amateur money. This is a very, very good offer.

Average price in Russia: 44,000 rubles

Wide angle lens

For wide-angle shooting. When will it be useful? For photography in tight spaces (review of a hotel room, for example), parties, architectural, landscape photography.

This lens is small, lightweight and has excellent sharpness. Its image quality and autofocus accuracy are a very strong competitive advantage over currently available lenses in this class. In addition, it contains a stabilizer, which allows you to avoid “smudges” in photographs.

The compact size of this wide-angle lens makes it suitable for everyday use, when traveling, or you can simply keep it in your pocket as a backup just in case. Average price in Russia: 16,000 rubles

Macro photography

This type of photography is considered highly specialized, but often becomes a favorite genre among photographers. In this category, we consider the Tamron AF 90mm f/2.8 Di SP to be the best

The build quality, lenses and viewing angle of 145 mm make this lens an excellent solution for a more or less reasonable price (reasonable compared to the Canon branded lens, of course, because the Canon EF 100mm f/2.8L Macro ISUSM costs 30% more). The lack of image stabilization may require the use of a tripod.

Average price in Russia: 45,000 rubles

Universal lens

At the very beginning of the text, we say that the kit lens needs to be replaced with something better. Meanwhile, the very idea of complete glasses is quite reasonable - a novice photographer is given a lens with universal focal lengths that works in most genres. Yes, the quality of the whales is not the best. Therefore, to the high-quality, but no less versatile Sigma AF 18-35mm F1.8.

As one of the commentators on this lens on Yandex.Market writes:

“This is the best zoom lens for crop at the moment. It replaces 3 prime lenses: 18, 24 and 35 mm with an aperture of 1.8. Consider that you have several lenses on your camera at the same time and you don’t need to waste time changing them.”

An important nuance - this lens has problems with autofocus in low light. Therefore, after purchasing a lens, we recommend finding a USB dock. This device allows you to update the firmware and fix the focus. In addition, on thematic forums it is advised to come to the store and try several copies, because... The quality of lenses even in the same batch may vary. Yes, there are possible pitfalls here, but you won’t be able to find any other lens for the money that can shoot like this.

Average price in Russia: 48,000 rubles

Another translation of an article by Roger Cical from Lenzrentals.com. The translation was made for www.photogora.ru, but let it be present in Vlador as well.

It’s interesting when the scientific intersects (although perhaps “collides” would be more appropriate) with the creative. The scientist states: “Facts are more important than feelings.” The Creator objects: “Only my vision and design matter.” This is exactly the case in fine art. Whether a photographer or videographer achieves the intended shot is the yardstick, and the equipment used is secondary. That's why I don't argue when an artist tells me that all the tests in the world can't influence his choice of equipment. I accept his words that “this” lens is perfect for him.

But, while not challenging the right of shooters to have their own opinion about what equipment is suitable for them, I continue to believe that inaccurate information and lack of knowledge are harmful. So I'm going to tackle a recurring statement online that drives my inner scientist into a rage: This zoom is as good as a prime. (And following from the first one - give me the best copy of this zoom).

We will use a scientific approach, which involves familiarity with my testing methods and interpretation of results. (Without a scientific approach, I would simply state that this lens scored 82.7 in our rating, the meaning of which you do not understand, and the other - 79.2, and the article would be short, which our editors like so much. By the way, the editors hate me ).

Let's start with MTF charts, which I know many people don't understand and refuse to learn to understand. This part will be mercifully short, and then we will move on to the long-awaited pictures. Just be patient, I’ll make it easier for you by using our experimental method of subliminal encouraging text: it works unnoticed, but the positive charge goes straight to the subconscious, instilling a lasting feeling of well-being and achievement of goals.

The science has started, but so far without mathematics, so we’ll be tolerant.

You can do it, I promise.

You've probably seen MTF charts. Even if you don't understand them, I think you have the idea that the higher the lines on the graph, the better. You've probably already evaluated lenses by comparing the placement of lines on graphs. MTF graphs show the average performance of real lenses (if the graphs were created by Zeiss, Leica or myself), or a computer model of the theoretical best possible scenario (for all other graphs). The graph shows how half of the lens performs, from the center (left side of the graph) to the edge of the image (right side).

Here are the MTF graphs of the two lenses, from which we can conclude that the resolution of the left one is higher than that of the right one. The graphs say a lot more, but we’ll limit ourselves to the main thing - the higher the line on the graph, the sharper the lens.

There is one “but”: the graphs display either the average performance of many lenses, or ideal performance simulated by a computer. Now we take on testing. Firstly, you see that our graph shows not half, but the entire field of the lens. The center of our image is in the center, and not on the left edge of the graph. Secondly, it is immediately clear that one side is clearly different from the other. With mass production it is impossible to make everything perfect. By the way, this is a chart of the MTF model, the average chart of which was above on the left.

It can be seen that the left side is different from the right. But if one side is different from the other, how do the top and bottom of the frame relate? Or corner-to-corner? If we want to truly test a lens, we need to take readings multiple times, rotating the lens to get data from different areas of the frame. Here's a graph of our lens tested at four rotations.

We're almost done with the boring explanations. There is very little left. You can do it!

This is a graph plotted over four rotations. You could do it at 8 or twelve, but the pictures will be very small, and you’ll already be bored. You’re probably thinking: “Just say 79.2 out of 100 in a simple way instead of all this drudgery!”

What if, instead of lines, we graphically display how MTF is distributed over the surface of the frame? Below is a map of the sagittal MTF, where the area of greatest sharpness is shown in blue, slightly worse in yellow, and slightly worse in red (in this case it is absent) where the sharpness is not very good.

Do you agree that this is much clearer? It's immediately clear that our test unit is well centered (highest MTF in the center) and the right side is slightly less sharp than the left. More details later, but if you test it yourself you won't notice such a minor difference. The MTF test bench is much more sensitive than any camera (at least to date).

Maps of other lens metrics can be made. Below, for example, is a map of the astigmatism of the same lens.

MTF is over for today. Now you can look at cute color pictures, achieving peace and relaxation.

It is obvious that the lens in question has increased astigmatism on the right edge. Such cards are an easy and reliable way to evaluate a specific lens at a glance. There will be a whole bunch of these pictures later, so I tried to explain how we get them.

Will this be noticeable in real photographs?

I already said that our optical test bench is much more sensitive than a camera. It distinguishes very minor deviations, which in real photographs will be masked by other variables present in the frame - lighting, focusing, location of objects and many others. Serious deviations will be noticeable. How serious? Let's look at the cards of two copies of the same lens model, one of which meets all the requirements, the other is worse. (Actually, it's not bad at all and the only worrying area is the red area at the bottom). If you were to shoot with this lens, you would most likely rate it “good” or “a little soft.” If you shot it second, you would describe it as “outstanding.”

(The map appears cropped compared to the one shown above. This lens has a built-in limiter to reduce reflections, and the image it produces is a rectangle, similar to a camera sensor, rather than a circle, like the one above.)

Olaf Optical Testing, 2017

I understand that it’s more pleasant to evaluate lenses based on colorful photographs, but a stage shot has too many variables, and we strive for scientificity. You'll have to make do with photographs of test tables.

Let's compare the top part above the center of the frame, which was excellent on the right lens and good on the left. To fit 100% of the crop into the format of this unfortunate blog, you will have to place them on top of each other: the right lens on top, and the left one on the bottom. These are crops of RAW images of high-quality test patterns taken on a 36-megapixel camera, without sharpening. On a camera with a higher resolution the difference will be more obvious, with a lower resolution it will be less noticeable, but for our purposes these images will be suitable.

I can see the difference, and I think you can too. If I had shot in JPEG rather than RAW, the difference due to in-camera sharpening would have been less noticeable. Don't forget that test charts are easier to evaluate than photographs, and in real photographs the difference would only be noticeable in a side-by-side comparison. If you bought the lens, the map of which is shown on the left, you would hardly complain about blurriness at the top of the frame, especially after some post-processing and when posting JPEGs with a resolution of 800 pixels on the long side.

Let's look at the lower left area. As before, the right lens is on top and the left is on bottom.

Here the difference is greater. It can be assumed that there is something wrong with this angle. Tangential test lines (those from top left to bottom right) appear gray on gray, indicating low detail. That's it, let's limit ourselves to this. I just wanted to show that our MTF maps reflect the real state of affairs.

How long? Isn't it time to talk about zooms?

It's coming soon, my patient friends. We're done with the conceptual part, there's not much left to do.

Many people know that a zoom, even if comparable to a prime lens in terms of sharpness in the center of the frame, can rarely boast the same sharpness at the edges and corners of the frame.

But few people think about how much more complicated a zoom lens is than a prime lens. The optical design of a zoom often has about 20 elements, versus 6–12 for a lens with a fixed focal length.

In comparison with one moving focusing group of elements in a fixed lens, a zoom has moving ones: a focusing group, an element or several that are responsible for changing the focal length; a compensating element is often added to them. Increasing complexity leads to increased scatter from specimen to specimen. Increased complexity leads to increased variability.

Let's look at the MTF cards of several quality prime lenses. I show maps of 9 real lenses tested using the method described above. I’ll add (because someone will definitely notice) that these are f/2.8 lenses, not f/1.4. No f/1.4 lens can resolve 30 pairs of lines so well. By the way, one copy from this group was dropped during rental, but “there was no visible damage.” Can you guess which one?

Calming blue means razor sharpness.

If you ask me to send you a really good copy of this group of lenses, I will, without a shadow of a doubt, choose any of those located diagonally from top left to bottom right.

(I’ll make a reservation right away so as not to return to this: believe me, if you estimate my labor costs for testing 9 lenses just to choose the best one, then you won’t be able to afford it). Even if I send you one of the other three copies, the card of which does not have yellow on the card, I am more than sure that you will not feel the difference in real pictures.

Now about zooms

Let's look at the cards of several copies of good, expensive (priced at $2000) zoom lenses. You probably already realized that the difference from instance to instance, even with a good zoom, will be greater than with a prime lens. But have you ever thought that zoom needs to be tested at several focal lengths? We are used to evaluating lenses in terms of “good / bad copy”. This works with primes, but is not always applicable to zoom lenses.

I present the results of tests of eight copies of the 70-200mm f/2.8 lens, carried out at three focal lengths.

I warned you that the truth would be uncomfortable. But everything will be fine. Exhale.

Firstly, I assure you that this picture is not unique to this lens, this range of focal lengths or anything else. We've tested thousands of zoom lenses. Everyone's behavior is similar with very rare exceptions. Some are generally harsher. Some work better at one end of the focal range. Good performance of a specimen at one focal length does not mean a similar result at another focal length. However, I will add that a failed result at one focal point actually suggests poor performance at others.

The thing is that even a good example of a zoom lens can be slightly decentered at one focal length, have a slightly tilted element at another and tilted in the other direction at a third.

Take a closer look, you will notice.

For example, if you had the opportunity to compare copies No. 6 and No. 4 side by side, then you would definitely choose No. 6 - it is better than its rival at 200 mm focal length. But without that comparison, you'll probably rate #4 as decent. The owner of example #6 would rate the lens as much sharper at 200mm than at 70mm, and the owner of #4 would claim that it is slightly sharper at 70mm. The owners of No. 1 and No. 8 would join the dispute and call their opponents incompetent photographers, because the lens clearly works the same across the entire focal range. The owner of #8 will probably be happy with his lens unless he gets to compare it to #1.

Wait to draw conclusions, we only considered the sagittal graph. And it’s worth looking at the tangential one (or the astigmatism graph showing the difference between the first two). Let's say, according to the maps above, the third number looks like one of the leaders at a focal length of 200 mm, but if you look at its astigmatism map, then it turns out to be one of the outsiders at this focal length.

Understand correctly, Zooms are by no means “horror and suck.” These are excellent and very convenient lenses. But once you know the compromises that go into designing them, you'd be amazed, as I was, that they can be so good at these prices. I will add that based on images posted online by forum fighters with a resolution of 800 or 1200 pixels on the long side, you will not only not be able to see the difference between zoom lenses, you will also get confused between zoom and prime.

I just want to emphasize that, in general, zoom lenses have greater variation from one example to another, plus each instance of a zoom lens will also differ at different focal lengths. These are the laws of physics and the inevitable tolerances of mass production. The more variables in the lens, the greater the difference and deviation. That said, are zoom lenses any good? Without a doubt, yes! Can they match the quality of fixes? No. But even the best prime lens will not give you the convenience of changing the focal length. Each tool has its own tasks.

What are the conclusions?

There are few conclusions, rather information that reminds us of reality. Here are a few points for photographers:

At comparable aperture values, even an excellent zoom cannot compare with a good prime lens, but it will be convincingly good, especially in the center of the frame.
Zooms are characterized by greater scatter from specimen to specimen, which is superimposed by the difference at different focal lengths. Ask me about the best zoom lens and I will inevitably ask: “Which focal length?” After all, the sharpest at 200 mm may not be the best at 70 mm.

Now about testing. I have to say: testing a single copy of a zoom lens is often simply pointless. Differences that are barely noticeable in real photographs stand out during testing. If someone were to test specimen No. 6 from our group, the numbers, and most importantly, the conclusions would be seriously different from those that would have been made when testing specimens No. 1 or No. 8.

A certain reviewer tested one copy of the zoom lens and gave it the highest rating for himself. Readers disagreed, argued with this assessment and asked me what I thought about it. This article is an attempt to explain why I see no point in getting involved in such disputes. An attempt to fit something as multifaceted as the operation of a lens with variable focal lengths into one single figure, and even after testing one single copy, has no scientific meaning or value. And I don’t care what rating is awarded - 3.1415926, 2.718281828 or 1.61803398. Unless the rating is 42. Then it will make sense.

It was funny. It's time to laugh. And eat an avocado.

Roger Cicala and Aaron Closz