 |
| |
As beginning or amateur photographers,
we do not have the luxury of going out and purchasing
a lens, only to find out that it does not meet our specific
needs in terms of sharpness. Also, we typically do not
have the budget to simply throw our money around and buy
the most expensive lenses, although I have seen many amateur
photographers spend as much as they would on a car for
their equipment. Personally, I would rather spend my money
on a trip to the Caribbean or Hong Kong, where I would
be able to actually use my cheap lenses to take real photographs
and not just test photos of porcelain figurines on my
desktop.
Luckily, there are many inexpensive
lenses out there that can give lenses several times their
price a run for the money. You may be surprised to find
that the most expensive lenses are not always the sharpest
lenses. In fact, one of the cheapest lenses that both
Canon and Nikon make, the 50mm f/1.8, also happens to
be one of the sharpest lenses they make. Don't get upset
that it's not a zoom lens. With a 50mm prime lens, you
only need to walk backward or foward several feet to get
the same zoom range as the Nikkor 35-70mm f/2.8 lens that
costs 6 times as much! The 35-70mm itself is less than
half the price of the Nikkor 28-70mm f/2.8, yet, it actually
outperforms the 28-70mm at f/8. The 28-70mm only surpasses
the 35-70mm when it is wide open at the telephoto end.
Depending upon your photography style, this may not even
make a difference to you.
The most objective tool we have
available to us in finding these "gems" is MTF
charts. This article provides the basics of what you need
to know regarding MTF charts and provides a comprehensive
method for helping you choose a lens that will best suit
your needs.
|
 |
 |
|
|
Introduction
This is not an in-depth article on Modulation Transfer Frequency
(MTF) theory. To learn about MTF theory, take a look at Norman
Koren's excellent article here.
Instead, this article simply provides the basics of what you
need to know regarding MTF charts. More importantly, it provides
a comprehensive method for choosing the best lens to suit your
needs.
Why Subjective Lens Evaluations and MTF Ratings Can Be Inaccurate
Many people, when choosing a lens, do not seriously consider
what they will be using the lens for, at least not as seriously
as this article proposes. Most people simply want a general
lens, a "sweeper", if you will, that they can glue
to their cameras and forget about. After looking at several
subjective lens evaluations and perhaps even comparing MTF ratings,
they finally purchase a lens. However, after examining their
photographs, they are suprised to find that they aren't as sharp
as they would've expected despite the lens having a relatively
good MTF rating. How did this happen? Aside from proper technique,
one of the problem is that the lens they purchased may indeed
be sharp, but not at the focal length and aperture they tend
to use it at. In other words, for their style of photography,
the lens is not sharp.
It is simply not enough to read subjective lens evaluations
and reviews. First of all, as you might expect, they are very
subjective. The reviewer may have a completely different style
than you. For example, in Ken Rockwell's informative website,
he has much praise about the sharpness of the AF
Nikkor 28-85mm f/3.5-4.5. If you look at that lens, it is
a lot sharper than you might expect for its price, especially
at f/8. However, when it is wide open, only the center remains
sharp. Well, it just so happens that Ken Rockwell is mainly
a landscape photographer, and when you shoot landscapes, you're
more likely to find yourself shooting at a wide angle with the
lens stopped down to f/16. So, depending upon your style of
photography, you may come up with a completely different conclusion
about the same lens, although, in this particular case, I find
Ken Rockwell's evaluation to be very accurate.
To me, subjective evaluations are great for getting a general
idea about build quality, how a lens handles, and its flaring
characteristics. But for determining sharpness, there is nothing
more objective than understanding a lenses MTF. But even MTF
ratings can be deceiving. This is so because they only provide
an overall weighted average of lens performance. You need to
look at the entire MTF chart, not just the overall rating. For
example, the Nikkor AF 24mm f/2.8 has an overall MTF rating
of 3.7. This is slightly higher than the 20mm Nikkor, which
has a rating of 3.5. However, by looking at the entire chart,
you will see that while the 24mm Nikkor is sharper in the center
than the 20mm, at the corners, the 20mm Nikkor is slightly sharper.
Depending on your photography style, you may prefer the more
even sharpness of the 20mm to the center sharpness of the 24mm.
Figure 1: AF Nikkor 24mm f/2.8 vs. AF
Nikkor 20mm f/2.8
The Bottom Line
The bottom line is that you need to first understand what you
want from a lens. To figure this out, you must understand your
style of photography. Secondly, to get an accurate evaluation
of a lens, you need to look at the entire MTF chart, not just
the overall MTF rating. Armed with the knowledge of your own
shooting style and a basic understanding of MTF charts, the
following steps outlined below will hopefully help you to make
a more informed decision when purchasing a lens.
Step 1: Determine the style(s) of photography you expect
to use the lens for
The first step in purchasing the perfect lens for your needs
is determining what your style of photography is. In the example
above, we see that the 24mm lens is sharper as long as the subject
is in the center. But what if you preferred placing your subject
off center, in adherence to the rule of thirds? Then, the 24mm
Nikkor may not be the best choice. Of course, for the beginning
photographer, you may not know what your style is. When I say
style, I mean, what combinations of aperture, focal
length, and subject placement do you find yourself
at most of the time? Although the following rules do not always
apply, they may be helpful as a reference point for beginning
and amateur photographers:
1. Landscapes - Most landscapes are shot at a wide angle
with the aperture stopped down to get the most depth-of-field
(DOF) and sharpness. If you expect to photograph mostly landscapes,
you'll probably want a wide angle lens (36mm or less) that is
evenly sharp throughout the frame when the aperture is stopped
down (f/8 or smaller).
2. Portraits - Most portraits place the subject directly
in the center (unless you're doing environmental portraits)
and are shot with a moderate telephoto lens. The aperture is
often fully open or almost fully open to blur the background.
Therefore, if you plan on using your lens for portraits, you
probably want a moderate telephoto lens (70mm-150mm) that is
sharp in the center at its widest aperture (f/4 or larger).
3. Street/Candids - For street photography and candids,
where you must capture spur of the moment action, you probably
need a normal zoom lens (between 35-100mm) that is sharp throughout
the frame when it is stopped down some (f/8). The reason why
you want it sharp throughout the frame is because candids often
include several people or objects interacting with one another,
not just a singular subject. You usually want to have your lens
stopped down because it will give you a greater DOF so that
even if you did not have time to focus accurately due to the
spontaneity of the moment, you are more likely to have the subject
in focus. And, unless you are like Henri
Cartier-Bresson, you'll need a zoom because you may not
have time to compose your images by repositioning yourself.
4. Action - The same as above, except you'll want a
longer telephoto, something that is at least 200mm, but more
likely in the 300m-600mm range. You may also want a lens that
is sharp when it is wide open to blur the background, but it
depends upon what type of action you are photographing and how
you intend to photograph that action. For example, for motorsports,
you may not need to blur the background because panning can
give the same effect. However, if the vehicles are coming directly
toward you, you cannot use panning to blur the background anymore,
in which case, you'll need sharpness at wide apertures (f/5.6
or larger). I typically use f/5.6 for my surfing photos, which
gives me a good balance between blurring the background and
maintaining an adequate depth of field to allow me some room
for play in case my focus was slightly off.
4. Abstract - Most abstract images trick the eye into
judging depth incorrectly. To do this, you must eliminate as
many clues about depth, one of which is focus. In photography,
shallow focus is not only used to emphasize the subject, but
also to add depth to an image, which you are trying to avoid.
Therefore, if you are doing abstracts, you'll most likely want
a lens that is evenly sharp thoughout the frame when it is stopped
down (f/8 or smaller).
5. Nature. Most nature photographs are taken with extremely
long telephoto lenses (300mm or more) with wide apertures (f/4
or larger). This allows you to blur the background to emphasize
the subject. Thus, you want a telephoto lens that has a wide
aperture and is sharp wide open. Coincidentally, this is usually
the most expensive type of lens.
Step 2: Rank the various styles of photography you expect
to use the lens for
In determining your photography style, you probably found that
one or more categories apply to you. This step keeps you from
expecting everything from a single lens, such as what people
expect when they buy "sweeper" lenses, like the popular
28-300mm Hyperzoom. The problem with these types of lenses is
that they are, in general, noticeably softer than the prime
lenses in their range. However, in most cases, it is impractical
to get a prime lens for every focal length you wish to cover
because you may not have time to switch lenses due to the spontaneity
of what you will be photographing. You must find a balance between
the two. Ranking the styles of photography you expect to use
the lens for in frequency of expected use will help you do this.
The very first lens I purchased using this methodology was
my Nikkor AF 35-70mm f/2.8 D. When I purchased it, I expected
to be shooting mostly portraits, followed by landscapes, followed
by candids, and finally abstracts. Thus, I ranked the priorities
of what I wanted the lens to do based on my expected shooting
style:
Priority 1 - For portraits, sharp in the center at
70mm or greater with the aperture wide open (f/4 or larger).
Priority 2 - For landscapes, sharp throughout the frame
at 35mm or less with the aperture stopped down (f/8 or smaller).
Priority 3 - For candids, sharp throughout the frame
from wide angle to telephoto with the aperture stopped down
some (f/8).
Priority 4 - For abstracts, sharp throughout the frame
from wide angle to telephoto with the aperture stopped down
(f/8 or smaller).
In summarizing my priorities, I was generally looking for a
zoom lens that was sharp in the center when fully open at the
telephoto end (for portraits), and sharp throughout the frame
when stopped down from wide to telephoto (for landscapes, candids,
and abstracts). As you know, I finally chose the Nikkor 35-70mm
f/2.8, however, there were several other lenses I looked at
with similar characteristics, as you will see in the next step.
Step 3: Choose possible lenses
The first step I took in choosing possible lenses was to look
at all the lenses covering the focal lengths I required. I looked
at every zoom lens with a focal length between 24mm and 150mm.
I looked at the following AF Nikkors: 35-70mm f/2.8, 35-135mm
f/3.5-4.5, 28-105mm f/3.5-4.5, 28-85mm f/3.5-4.5, and the 28-70mm
f/2.8. I then eliminated several lenses based on subjective
evaluations. What I looked for, in particular, was common complaints.
The more common a complaint, the more likely it was valid. I
also eliminated certain lenses based on my budget. For example,
at around $1200, the Nikkor AF-S 28-70mm f/2.8 was definitely
out of my league. Even the 35-70mm at $400 on E-Bay was beyond
my budget, but I decided to keep it on my list because it was
close enough. I also chose to keep third-party lenses off my
list solely because of resale value. I have found original equipment
manufacturer (OEM) items to hold their value much better. In
fact, I recently sold a AF Nikkor 75-300mm f/4.5-5.6 lens for
$65 more than I bought it for. However, you should not always
limit yourself to Nikkors or Canons. Sigma, Tamron and Tokina
have several excellent lenses that, in some cases, outperform
their OEM counterparts.
Step 4: Rank possible lenses using MTF charts
This is where understanding the basics of MTF charts comes
into play. In this step, I looked at the MTF chart of each lens
that remained on my list. I ranked the lenses by comparing the
relative sharpnesses at the apertures and focal lengths required
for each type of photography I expected to use my lens for.
I created a matrix with the results and finally scored each
lens with the sum of its ranking for each style of photography
multiplied by the importantce of that style for my needs. I
will explicitly illustrate what I did, however, I must first
explain the basics of understanding MTF charts.
|
What you need
to know about MTF
To understand MTF and how it relates
to sharpness, you must first understand what sharpness
truly is, which is nothing more than localized contrast,
or, in layman's terms, contrast between fine details.
When evaluators create MTF charts, they are essentially
measuring the contrast between lines on a test chart as
seen through a lens and comparing it to the true contrast
of the chart by percentage. A perfect lens would be able
to transfer 100% of the original contrast of the the test
chart. Of course, this lens does not exist.
The lines on these test charts
usually fall under two categories. The first type of line
is oriented so that it extends like the spokes of a wheel
from the center of the lens to be tested. These lines
are called radial or sagital lines. The second type of
line is perpendicular to the radial lines. These are called
tangential or meridonial lines. Each of these lines are
typically arranged in test patches, with increasingly
finer thicknesses and distances, which evaluators describe
in line pairs per millimeter (lp/mm). The greater the
lp/mm, the finer and closer the lines are. Figure 2, below,
shows a sample test chart used by the USAF that many other
test charts are based off of.
Figure 2:
USAF 1951 Test Target
When reading MTF graphs
(see Figures 3 and 4 below), you will see two axes. The
vertical axis represents the percentage of contrast a
lens is able to retain compared to the actual test chart.
The horizontal axis represents the distance from the center
of the lens. In addition, there are usually several of
these graphs, each representing lens performance at various
focal lengths and apertures. For every graph, you will
see that there are there are 3 sets of solid lines. These
lines represent the ability of the lens to resolve the
radial lines of the test chart. Likewise, the dotted lines
represent the ability of the lens to resolve the tangential
lines of the test chart. The topmost lines represent the
lens performance at 10 lp/mm, followed by the second lines,
which represent the lens performance at 20 lp/mm, followed
by the lowest lines, which represent the lens performance
at 40 lp/mm. In general, the topmost line (performance
at 10 lp/mm) describes how much overall contrast the lens
will have. The middle and lower lines, (performance at
20 lp/mm and 40 lp/mm, respectively), describe what we
would typically perceive as sharpness. Strangely enough,
however, a lens with more contrast may sometimes appear
sharper than a lens with less contrast, regardless of
its actual performance at the finer lp/mm.
Figure 3:
AF Nikkor 35-70mm f/2.8 D
Figure 4:
AF Nikkor 35-105mm f/3.5-4.5
The Digital Bonus
The MTF charts for
most lenses are measured with the dimensions of film in
mind, which, in the case of 35mm film, is 24mm x 36mm.
The radius of the diagonal of 35mm film is approximately
21mm, which is how far from the center the MTF charts
are measured out to. With the smaller sensors of digital
SLR cameras, which, in the case of Nikon's DX standard,
have the dimensions 23.7mm x 15.5mm, the radius of the
diagonal is only about 14mm. This means that you can pretty
much ignore everything to the right of 14mm on the MTF
charts. As a result, you have essentially cropped out
what is usually the most unsharp region of a lens. This
is another reason why overall MTF ratings can be especially
misleading for digital photographers.
|
Now that we understand the basics of MTF charts, we see that
all we need to do in order to objectively compare lenses is
compare performance at similar apertures and focal lengths by
seeing which lens has a higher graph for a given lp/mm. To illustrate
how I compared the lenses above, I created an overlay of the
MTF characteristics of both lenses at around 70mm with the lens
open as wide as possible (see Figure 5, below), which is what
the style of shooting I gave the highest priority required.
Figure 5: Comparison between 35-105mm
at 60mm (red) and 35-70mm at 70mm (green)
Surprisingly, the 35-105mm seems to do better
when it is wide open as compared to the 35-70mm, however, keep
in mind that the 35-70mm has a wider aperture of f/2.8 compared
to f/4. I took this into account and in this situation, I ended
up giving both lenses the same score. Thus my ultimate decision
rested on lens performances in the other types of photography
I expected to use the lens for. Figure 6 shows the final matrix
comparing both lenses.
Figure 6: Matrix used to rank lens performance
based on MTF characteristics
The lower the total score, the better the lens
could be expected to perform for my specific needs. As you can
see, this is the reason why I chose the 35-70mm f/2.8 over any
of the other zooms covering a similar range.
Conclusion
When I initially did this comparison, I was very
surprised at how well the 35-105mm lens did compared to the
35-70mm. Although you don't necessarily have to go into as much
detail as I did when picking a lens, you should, at a minimum,
consider the type of photography you will expect to be doing.
The benefit of doing the comparison outlined in this article,
however, is that it can result in insight that would otherwise
be overlooked. In statistics, we learn that the most important
insights are the ones that surprise you. Imagine how surprised
your friend who purchased a 28-70mm f/2.8 will be when the photographs
you show him from a lens half as cheap end up being just as
sharp or sharper!
Hopefully, with this article, you have learned
a little more about how to understand MTF charts and also learned
of a way to objectively compare lenses based on your specific
needs. Although this method may seem overly methodical, the
crisp images you are rewarded with are well worth the extra
effort.
Where to Find MTF Data
Unfortunately, Nikon only publishes the MTF charts
for a few select Nikkors on their Japanese website.
Also, www.photodo.com
had some MTF charts for the older Nikkors, but they seem to
have removed the actual charts and left only the MTF ratings.
Niklas Nikitin has several MTF charts on his website,
but he switched to Canon and mentioned that he may remove them
from the web. I have downloaded all the charts and posted them
on my PBase account, here,
in anticipation of them being removed.
For another good read about how to properly use
MTF data when purchasing a lens, take a look at this article
by Dante Stella.
|
