My impressions, 20 F2.8 USM vs 17-40 F4L USM

[isaac sibson]

I posted this in another forum, but feel it is of wider interest...

 

<snip>

 

Having a good rapport (based on the spending of large amounts of

money) with your local camera shop can be a very useful thing.

 

This proved so today, as my local shop let me out with one of their

EOS 1VHS (since my EOS 3 is at my parent's place) and a brand new 17-

40 F4L without any security. Only for half an hour, but still....

 

So, I shot a roll of film (the roll of Kodak High Definition that

I've had kicking around since it came free with Amateur photographer

a while back), just of various things around the Cenotaph in

Southampton, but still worthwhile...

 

I also compared some shots to my 20 F2.8 which I had with me at the

time. Unfortunately, since it was film, I have no samples to show

for now, and comments are based on some (not brilliantly made, since

it was a 1-hour job) 7.5"x5" prints. However, some things are

evident...

 

1) Vignetting. This is the most serious "problem" I could find with

the 17-40. At the wide end, wide-open it vignettes. Badly?....

 

....well, no actually. At F4 (wide open for the zoom) it vignettes

less than the 20 F2.8 does at F4 (one stop down, of course). No

vignetting is apparent at F8. No vignetting is apparent at 40mm,

even at F4.

 

2) Contrast. Excellent. As good as the prime, despite the fact that

I didn't have the hood for the 17-40, and had the hood on my 20

F2.8. Flare resistance seems good.

 

3) Sharpness. Hard to judge to a truly critical level from this size

of print, but... VERY good. It appeared to match or even exceed the

20 F2.8 at the 17mm end. F4 is very sharp except in the corners (way

outside the 1.6X FOV and outside the 1.3X FOV of the 1D). That's not

to say that there isn't a notable difference from the corner of 1.6X

FOV to the centre, but it isn't anything like objectionable. The

centre is extremely sharp at all apertures, with no notable

difference between F4 and F11 in the centre. Edges and corners

improve as you'd expect.

 

4) Chromatic Aberration. Observed on a few high contrast edges in

the corners/edges. Clears up as stopped down, but for a lens of this

wide an angle, CA is EXTREMELY well controlled. Better than the 20

F2.8.

 

5) Handling. This lens handles BEAUTIFULLY. It is truly superb. The

only thing that could have made the 1VHS + 17-40 F4L combo handle

better would have been the hand-strap. The zoom ring is super-smooth

and damped, with NO change in resistance as the lens groups change

direction (as can be felt in a great many zoom lenses where this

happens, including, I'm told, the 28-70 F2.8L). The focus ring is

also superb, feeling as good as the focus rings of the 300 F4L IS

and 135 F2L, a real cut above the 70-200 F4L, which in turn is way

better than the consumer zooms. It feels a little better than the

focus ring of the 20 F2.8.

 

On the basis of my results, I am going to replace my 20 F2.8 with

the 17-40 F4L. I felt that the two lenses were pretty much on a par

with regard to performance, with the 17-40 maybe better in some

situations, but the flexibility of the 17-40 takes it for me. It

will replace my 20 F2.8 and 24-85 lenses. At F4 in the overlapping

range, I think that the 17-40 kicks the 24-85 into the weeds, and

that probably applies as far down as F8 or so.

[andr__fisahn]

Thanks for the quick review Isaac. I'm contemplating the same switch - from the 20/

f2 which I have to the 17-40/f4. And I trust you're going to purchase the zoom in

your camera store and thereby further increase their inclination to let you loose with

several grand worth of equipment :-)

 

Best,

 

André

[isaac sibson]

Indeed. They give me excellent prices as well, almost as good as is possible to find in this country (UK), and I'll pay that little extra for the level of service.

 

That was really the purpose of this "test drive", and when I get it (when funds allow) it will definately be from them.

[anand_n._vishwamitran]

On a different note, Im curious about your impressions of the camera you used. How did the EOS 1v handle compared to your usual EOS 3?

 

Thanks

[isaac sibson]

Entirely the same. I shot the whole roll on single-shot, single-frame drive with centre focus point selected and CF 4-3 set. Looking through the viewfinder I didn't know what camera I was shooting except for the different noise and even shorter viewfinder blackout.

[night73]

Thank you Canon and Isaac for pointing out the fact that while I thought I was finished buying lenses, I am back in it again!

[jim_larson1]

NOW **THIS** WAS THE 17-40/4L REVIEW I WAS WAITING TO READ!!

 

I am now wondering how the 24/2.8 would have stacked up, since it is reputedly sharper than the 20/2.8.

 

Hmmmm. Maybe I will need to start saving the extra 40,000 pennies for the "L" zoom.

[samir]

Thanks for your review. I just got one for the week end ... I will compare it to a 17-35 2.8 I had 2 weeks ago!

 

If the vigneting at 17 is less than the one you observed with the 20mm, this is a great performance. My only reservation to your review, is that you used a one hour lab and small prints which, as you say, limits the scope of the review.

 

Thanks for this good review (which complements one I have seen on luminous-landscape and which was very positive as well!)

[isaac sibson]

Indeed Samir, but I never intended it to be a completely rigourous, scientific test. I will put the negs through a film scanner when I get the chance and see how things look then.

 

Certainly anyone who has the 20 F2.8 shouldn't be itching to get rid of it on performance, since it is still a perfectly fine lens. I don't feel that the difference between the two is enormously significant, and neither one is shamed by the other.

 

But the new zoom does look to genuinely be to the 16-35 F2.8L as the 70-200 F4L is to the F2.8 versions, in that it is a smaller, lighter, cheaper, slower lens of the same (or similar) range which is still very good indeed optically.

[John Crowe]

This is not the kind of lens test one really wants to use in deciding to put aside one expensive lens to purchase a much more expensive one. If you like the flexibility that zooms have to offer then go for it. If you want to purchase based on resolution, and contrast then I recommend using fine grain slide film for testing purposes and view them under at least 10x magnification.

[jim_larson1]

John: Agreed. But this is the first review I have seen of the new zoom vs the old primes. I can wait for awhile till the jury comes back.

[ben_rubinstein___mancheste]

Which shop Isaac?

[isaac sibson]

Of course one should make one's own tests to satisfy yourself that a change is or is not worthwhile. I will be looking much more closely (ie neg scanning) my results, and seeing whether I still think the same, but on what I personally have seen for the sort of uses I personally would have for such a lens, the 17-40 F4L is overall the better choice.

 

The shop in question is London Camera Exchange, Civic Centre road branch (there's two), Southampton, UK.

[samir]

Picture from 17-40 ... so far I am not thrilled...

[samir]

Previous picture hand held, wide open... f4, iso 400

[samir]

Another pic, no PS adjustment... f5.6, iso 100, @40mm. 1/125s

[samir]

One more ...

 

Tv(Shutter Speed)

1/90

Av(Aperture Value)

6.7

Metering Mode

Evaluative

Exposure Compensation

0

ISO Speed

100

Lens

17.0 - 40.0mm

Focal Length

25.0mm

[samir]

If I go through the original review above. Please note I am essentially a

primes user (50mm1.4, 100mm 2.8 macro) and that I had been impressed by

the old 17-35mm 2.8 I had for a week. Also, no scientific review, just some

impressions.

 

1) Vignetting. No problem .

 

2) Contrast. Very good. Could not check for flare, it was rainy here in DC with

a cloudy sky.

 

 

3) Sharpness. I do not know the sharpness of the 20mm, but sharpness is far

from excellent on distant subject. Overall I would say good, and better than

the old 17-35mm 2.8.

 

 

4) Chromatic Aberration. A lot on high contrast edges in the corners/edges.

Sharpening with PS will ake it more apparent.

 

 

5) Handling. This lens handles BEAUTIFULLY. I agree...

 

What I do not like, is that all pictures are flat, there is feeling of 3D ....

 

I interested I have more pictures, indoor with flash. I want to compare it with

the 16-35 2.8, to see it that one has that 3D sensation or not...

[samir]

I have some pictures done with the 17-35. They are not comparable but show,

I think, the special look of that lens...

[samir]

another from the old 17-35

[lostinhigh]

I had pretty much the same experience as Isaac. I did compare it to my Canon EF 24/2.8 lens and in terms of sharpness, distortion and vignetting, it held its ground. <br><br>

 

I did quickly test it on my 10D (so vignetting was a non-issue). Distortion is very comparable (@ 24mm and f/4), at least to my non-scientific eyes. I do mostly landscape and travel photography, so my definition of distortion may not stand up to some of the more critical eyes. In terms of sharpness, and I checked this repeatedly, "my" copy of the 17-40/4L was better (from f/4 to f/5.6) and more or less equal (@ f/5.6 and higher) to "my" EF 24/2.8. I was so convinced that I decided to sell it. (Interestingly, I had a EF 24-85/3.5-4.5 zoom as Isaac and sold also.)<br><br>

 

PS: Most of you are probably aware of the test performed by Michael Reichmann. If not, check it out here:<br><br>

 

http://www.luminous-landscape.com/reviews/lenses/canon-17-40.shtml

[charles_dunlap]

The series of photographs by Samir Jahjah do not, in my view, contribute much to the

discussion. When I loaded them into Photoshop and viewed them at 100%, I noticed

obvious jpeg compression artifacts, and these became very apparent when

performing a modest unsharp mask. The jpeg compression is going to mar any

critical evaluation of the fine details that might distiguish the lens from others.

 

Furthermore, the lack of pictures taken by another lens under the same conditions

makes the test essentially worthless in judging the performance of the 17-40. This is

a common pitfall in subjective, informal lens comparisons. We are asked to view

heavily jpeg-compressed images of uninteresting subjects and to compare them to

photographs of a compelling subject with much better composition shot under

entirely different lighting conditions. This demonstrates nothing about the 17-40.

 

If I had to make an observation based on the hodgepodge of snapshots posted here, I

would say that the 17-40 appears to have better contrast than the 17-35, but the

MTF published by Canon tells us that this should be the case at certain aperture and

focal length combinations.

 

What we really need is a properly designed comparison if we are to gain anything by

the posting of photographs.

[jaïbee]

I am debating which lense to get between the two. Anyone else have any more helpful comparision between the two lenses?