AF 28-105 f/3.5-4.5D IF is BETTER than AFS 17-35 f/2.8!

Could it be true? I posted a newspaper on the wall and using a D100 mounted on

a tripod i tested these bothe lenses.

Test set-up

Focal length: 28 mm

Aperture: f/3.5 for 28-105 and f/2.8 for 17-35

 

What a surprised! The picture from 28-105 shows significant sharpness more

than 17-35.

Tell me guys, what can be wrong here? I thought 17-35/2.8 is the SUPER LENSE

at all apertures and all focal lengths.

 

I extracted two pictures from upper right corner.<div></div>

AF 28-105 at 28 mm at f/3.5<div></div>

The comparison. Overall picture from 17-35 is blur. Here is the comparison which shows picture from the other lens is much clearer.

If you shoot a lot of newspapers, then you have your answer!<g> There are so many factors to sharpness, like distance, aperture, focal length, camera, what monitor you're viewing it on, etc. As I've swapped lenses over the years I'm really wondering if different batches or individual lenses are better than others. Some of my best lenses are those that others say are dogs. I own some "legendary" lenses that are no where near as good as what others think.

 

For some comparison, I really like the 18-35mm f/3.5-4.5 lens. It's very nearly as good as the 17-35mm f/2.8 I've shot with. This image in my portfolio: http://www.photo.net/photo/4191418 is difficult to tell at such small magnification, but the very finest print under "Cents Per Gallon" is still razor sharp in larger versions. But that lens is highly under rated. I think you just have to find what works for you in the real world and not worry what others say. Few have extensive experience enough with every lens to tell you what is good or bad.

you're not doing it right by comparing the shots from 2.8 and 3.5. Further, you need to take 2-3 shots at different apertures and different focal lengths to see how each lens behaves in comparison to the others at different settings and see which one suits better your shooting style.

 

ps: 17-35mm is sharper than 28-105 at 2.8 (of course, lol). how sharp it is at 3.5, is yet for you to discover. however, virtually all lens is not as sharp as they are when stopped down. so it's probably better to see how these two lens fare at 4, 5.6, and 8 (with 11 even better).

The difference doesn't appear that much except for color difference and slightly less contrast. The test is of limited use in that we don't know if the camera's AF was employed to select focus and if so was the focus at the center or the edge where you are showing the crop. If you are concentrating on the edge sharpness, the lens should be focused manually for that edge section to make a fair comparison. There does appear to be more barrel distortion with the 28-105 at 28 mm. Also you might compare the lenses stop for stop sice I would expect a zoom at f/2.8 to show less contrast than one at f/3.5.

I can't really sleep here. I just tested using my older and cheaper lens AF 28-80 f/3.5-5.6D. Using the same focal length at 28 mm at f/3.5 mm. What a surprised! The picture from this lens show excellent sharpness on all texts across the pictures...ALL TEXTS!

"Some of my best lenses are those that others say are dogs"

In my case it is exactly the same. I have two third party lens that I never would sell. Nikon user.

"It's very nearly as good as the 17-35mm f/2.8 I've shot with"

 

After what I saw today, i don't think 17-35 f/2.8 is nearly as good as any other nikon lens.

 

It doesn't worth extra $ 1200. Not even for f/2.8.

Edward, how far was that newspaper from your camera? None of the lesnes you have tested in this thread so far is designed for close focusing from a meter or so from the subject.

Shun..all of the lenses...is ~10 inches from the wall...

10 inches is awfully close, Edward, and shooting at f/2.8 on the pro level lens is not going to offer much d.o.f.

 

Better test of value of these lenses *for you* would be to take them out and shoot the kinds of photos you like to take, assuming that the newspaper on the wall is not ordinarily what you photograph. :)

 

The consumer level 28 - 105mm zoom actually has a close-focusing capability (not true 'macro' perhaps, but still quite useful). I use a friend's 28 - 105 and for me it's been excellent. Not really a low-light lens, and not as fast to focus as the AF-S lens, but still very good, and lightweight, and well priced.

Suggestion, try comparing at something further away, and stop down both to f4. The 17-35 is not optimised for close-up photography wide open.

 

Also the 17-35 is known to be at its best at wider focal lengths. The 28-105 is a very good lens. But it lacks the build, AF speed, large aperture and performance at smaller apertures of the 17-35.

Edward, there are many factors that can potentially contribute to soft images. But I wouldn't test any non-macro lens from 10 inches (25cm) away from the subject; it is not at all surprising that you are not getting good results. The 17-35mm/f2.8 AF-S is not designed for that type of photography. It is like trying to use a Porsche sports car to tow a boat; you'll realize that a Jeep may do a much better job in that case.

<em>Tell me guys, what can be wrong here?--Edward

Ng<br>

</em><br>

A focus error is the most plausible answer. There are a number of

other possibilities. The D100s viewfinder and focus screen

are not suitable for critical focus.<br>

<br>

When I test a lens on charts I use a DW-2, DW-4 or DW-31, 6x high

magnification waist level viewfinder. I focus with a 1,000 watt

quarts flood lamp then turn off all lights and make the exposure

with electronic flash with the camera set to B. This

eliminates focus errors to a considerable extent and it

eliminates vibrations so its as close to a pure test of the

lens as I can possibility do. In consideration of film flatness I

dont let the film sit in the camera and get kinked.

Considerable care is need in setting up the camera to assure that

it is perpendicular to the test target. My test film was Kodak

Tech Pan developed a little flat (N-1) for continuous tone

printing.<br>

<br>

The distance for all lenses must be the same. The test is only

valid for a given distance. For example I tested many of my

lenses at 2m (6.6) as that is a typical portrait distance.

A more general distance would be 5m (16.4). For general

photography you need a much larger test subject than a sheet of

newspaper.<br>

<br>

Is this a real world test? Not for most of my use but it gives

some useful information about a lens. From there you take the

lens into the real world and use it to gain a practical knowledge

of its strengths and weaknesses.<br>

<br>

---<br>

<br>

Its customary to ridicule lens testing but it gives useful

information. As long as testing isnt the whole object its

not a bad thing. Its also customary to downplay the

importance of the sharpness of a lens. Sharpness is the easiest

feature of a lens to check off so why not do that first. Its

far from the only important issue and sharpness in the real world

can easily be more about the skill of the photographer than the

quality of the lens. Sometimes sharpness isnt even desired

but if a lens has it, its easy to loose. With a DSLR youve

got a mild soft focus filter with you at all times if you shoot

RAW.<br>

<br>

---<br>

<br>

<em>As I've swapped lenses over the years I'm really

wondering if different batches or individual lenses are better

than others. --Michael Axel<br>

</em><br>

Michael,<br>

<br>

I tested three 105/2.5 Nikkors, AIS, AI and IC under the

conditions noted above. The AIS was best, the AI next and the IC

last. The difference was greater between the AI and IC, than the

AIS and AI. These lenses all have the same optical formula so the

only plausible explanation is sample variation. All were good

lenses but the AIS and AI were better in terms of sharpness.<br>

<br>

<em>I own some "legendary" lenses that are no

where near as good as what others think. --Michael Axel<br>

</em><br>

Id love to say my 105/2.5 AIS is my sharpest lens but its

not so. The lowly 50/1.8 AI and the 55/2.8 AIS Micro both best the

105(s) by a clear margin.<br>

<br>

Another feature of the 105/2.5 is that its not as sharp

wide open and doesnt achieve its sweet spot until f/5.6

where other lenses I tested achieved their sweet spots about a

half stop or so earlier. This is a result of 105/2.5 being

designed to give beautiful background rendition wide open and at

portrait focus distance. This is a true feature. By sweet spot I

mean general area center sharpness. <br>

<br>

In some cases as one stops down you can see center sharpness

begin to fall off as edge sharpness is still increasing.

Diffraction doesnt just start at f/16 or f/22. Its

there all along. I find with high resolution film that if a lens

doesnt achieve its sweet spot by f/5.6 that its

not going to be very sweet when it does. Lens contrast is a

factor in image sharpness but diffraction tends to take this away.

A an enlarger easel you measure the loss of sharpness when a lens

is stopped down deeply.<br>

<br>

---<br>

<br>

<em>Shun..all of the lenses...is ~10 inches from the wall...

--Edward Ng<br>

</em><br>

Edward sleep in peace: this test distance is only valid for a close-up

lens. Try a distance of 5 meters, about 15 feet, for something

more appropriate to general photography. There is a reason why

brick walls are popular as test targets ;)<br>

<br>

If you dont have a place to test with proper controls

believe you results from general photography. Forgive me but

poorly done tests are much less than useful.<br>

<br>

The reason for formal lens testing is to gain knowledge quickly.

If you have a few months or more of experience with a lens and

the test does not agree with your experience believe your

experience as the test is surely flawed.<br>

<br>

Best,<br>

<br>

Dave Hartman.

I really believe that you should test a lens in the situation you would most likely use it. A portrait photographer would like to see how a lens handles portrait distances and probably wider apertures and may well be interested in how smooth the OOF areas are. A landscape photographer may well want to stop a lens well down to get maximum DOF. If you don't test a lens in the situation that would likely use it in you could be dissapointed with your choice.

Stop reading newspapers.

 

I have the 28-105, I enjoy this lens very much. I keep it for one reason only: quick and dirty snap shots; at that, it excels. I can't justify getting the 28-70/2.8, and the G/DX marvels are out of the question. This little bugger was found for under $150 used, and is incredibly sharp at f/5.6 and f/8-- it also exhibits near to no distortion.

 

Now the bad news, Edward. You either have a fantastic 28-105, a lemon 17-35, a flawed testing procedure, or a combination thereof. Your pic (Pun intended).

Mebbe' better for you, but I'd chew my own arm off and walk on hot coals before trading

my 17-35mm for a 28-105mm.

I don't do intensive lens testing like David did and I did test them only in 1 condition. I hope I flawed the testing procedure (just to make me feel better that i had spent $$ for such expensive lens)but for whatever it is...for close range object...plastic and cheap zoom lens 28-80/3.5-5.6 wins it all...tack sharp and clear. All texts visibly clear and sharp all across the picture. If it is the focusing error caused by the D100, then why is it when paired when cheap zoom lens producing tack sharp picture? Well...I know cheaper jeep does better job on towing the boat than the expensive porsche..but.......well...I thought 17-35/2.8 is THE lense like many other tests say. I am going to test the lens with further object and different aperture and focal length.

 

For your information, i magnified the pictures and study the details. The pic from 17-35 is generally soft in every corner of the frame with softer image at upper right corner. The pic from 28-105/3.5-4.5 is generally better than 17-35 but the center left part of the image is really soft. Then, the pic from 28-80/3.5-5.6 is all sharp in every corner across the frame.

<i>"better than 17-35 but the center left part of the image is really soft"</i><br><Br>

Again, either a lemon, or more likely, you're unleveled and DOF is killing your snapshots.

Edward if the lense has specific corners being soft (upper right, etc.) I'd make certain the film/sensor plane is absolutely parallel. At close distance critical focusing is not trivial. Your best bet is to manually focus and bracket and then pick the best (sharpest). With specific lense/camera combinations, it's not uncommon that the AF consistently focus a little before or behind the test object. This is obviously important when comparing performance nearly wide open at close range. Otherwise you are only testing DOF and it might come as a surprise that every lense is sharpest at f/16 ;)

 

Lenses have special "traits" at close focus. So if all you want is a casual sharpness test, use a larger distance from where you shoot most frequently.