Question for 17-40mm f/4L shooters

[slejhamer]

Images I take with this lens seem to need much more sharpening than images I take with my 70-200 f/4L or with my 50mm f/1.4 (well, not surprising on the latter), even with the lens stopped down and the camera on a tripod. Similar results at all focal lengths. Camera is a 10D.

The resolution seems to be there, but the 17-40 seems to have much lower contrast and the shots need an additional dip in the deep fryer to get them comparably crispy. It's not bad, just unexpected.

Is your experience with this lens similar? If someone can post some 100% unretouched crops of what they consider sharp images taken with this lens I'd appreciate it.

[alan sh]

I always use +1 on the sharpening menu in my Canon 20D. I had not noticed any difference with this lens and my 28-135.

What settings do you have in the camera ?

[KennyG]

Here is a sample with 100% crop taken with my 17-40L, 1D MK-II, 1/100 @ F4 and FP flash. Hand held, shot in RAW and processed by Capture One with default settings, which are not the sharpest. The lighting was pretty awful, as it is at most of these events, but you can see it is pretty sharp. Sorry you can't read the small text on the poster, but that would be pushing my luck.

I have no issues with this lens at all.

http://www.motorpix.co.uk/misc/MotorPix-1.jpg

http://www.motorpix.co.uk/misc/MotorPix-1-crop.jpg

[cmM]

Yes, I also experience similar results.
It delivers good quality shots either way, just... as you say, not the same as coming out of a 70-200 2.8IS for example. I think the focal lengths influence this as well.

[KennyG]

I always use +1 on the sharpening menu in my Canon 20D. I had not noticed any difference with this lens and my 28-135.

I would be very worried if my 17-40L was no better than a 28-135, which isn't one of Canon's better lenses. My old 28-135 has sat on a shelf for nearly two years gathering dust in my 'why did I buy it' collection. The 17-40L should be streets ahead of the 28-135, and if it isn't, I would have it checked.

[HKFEVER]

Send it to Canon, same as my 16-35. Tech in Canon told me that they will fix it in very short time.

[roanjohn]

Not mine.......my copy is sharp from corner to corner........even wide open....with excellent contrast.

Maybe you got a bad copy???

Some proof.......this shot was with the 10D, ISO 800, f4 at 1/5 of a second......handheld!!!

http://www.pbase.com/roanjohn/image/38755404.jpg

Even from a resized photo.......you still see a lot of detail........I'll post some 100 crops if you like.

Ro1

[Cadwell]

I always use +1 on the sharpening menu in my Canon 20D. I had not noticed any difference with this lens and my 28-135.

What settings do you have in the camera ?

:eek: One of the reasons I replaced my 28-135mm was that it produced soft looking images compared to my 17-40L...

[Jim Larson]

Hmmm.

I have to say it depends: I just went through a whole pile that needed massive correction and another pile that recieved practically nothing.

[booggerg]

Not mine.......my copy is sharp from corner to corner........even wide open....with excellent contrast.

Well, when you're using it on a 1.6x body like the 10D, it's actually not "corner to corner"

[roanjohn]

Well, when you're using it on a 1.6x body like the 10D, it's actually not "corner to corner"
:p ........alright Mr. Technical...........:D

Ro1

[alan sh]

I knew that would raise some questions.

Of course my 17-40 is better than the 28-135. But both give me nice photos and I do not worry about the 'softness' of either.

[Jon]

Well, when you're using it on a 1.6x body like the 10D, it's actually not "corner to corner"
What? Are you saying it doesn't cover the full 10D sensor frame? If it covers the full sensor, that's "corner to corner" WRT the 10D which was the subject of the original post and the cited body in the reply.

If you're going to pick nits, make sure they're ripe. ;{)#

[Lesmac]

Most of the images in my gallery were taken with this lens , at full width and all apertures, both on a 10D and a 1DS MKII

Les
http://lesmclean.photoblink.com/

[RichardtheSane]

Well, when you're using it on a 1.6x body like the 10D, it's actually not "corner to corner"

But doesn't corner to corner refer to the image and not the lens

Strictly speaking since the lens is not square then it doesn't really have corners.... :p ;)

[slejhamer]

Thanks for your posts.

I've given up on this lens, and will be returning it to B&H. The more I use it, the less I like it. AF in anything but the brightest scenes is horrible. DOF considerations aside, it's no sharper at f/8 or f/11 than it is wide open at f/4, and even stopped down it's really only acceptably sharp in the center. Also, it is notably softer on the left side (when shooting horizontal) than on the right.

Looking back on some old pics, I can see that my old Canon 28-135 IS and Tamron 28-75 Di lenses were notably sharper at comparable focal lengths. From what I can tell, the 17-40 isn't much better than the POS Sigma 17-35 that I sold last year.

In other words, I got a bum copy!

So back it goes...

[LexLuther]

I think this pretty much speaks volumes. http://www.luminous-landscape.com/reviews/lenses/canon-17-40.shtml

I feel the same way about my 17-40 f4L. In bright daylight the lens is pretty amazing. But once things get dim everything gets soft, quickly. I was shooting indoors with poor lighting and my 70-200 really out shined the 17-40. Throw the flash on and things gets better.

[Tom W]

But doesn't corner to corner refer to the image and not the lens

Strictly speaking since the lens is not square then it doesn't really have corners.... :p ;)

I believe that Richard has cornered you on this one. ;)

[Tom W]

Thanks for your posts.

I've given up on this lens, and will be returning it to B&H. The more I use it, the less I like it. AF in anything but the brightest scenes is horrible. DOF considerations aside, it's no sharper at f/8 or f/11 than it is wide open at f/4, and even stopped down it's really only acceptably sharp in the center. Also, it is notably softer on the left side (when shooting horizontal) than on the right.

Looking back on some old pics, I can see that my old Canon 28-135 IS and Tamron 28-75 Di lenses were notably sharper at comparable focal lengths. From what I can tell, the 17-40 isn't much better than the POS Sigma 17-35 that I sold last year.

In other words, I got a bum copy!

So back it goes...

If its softer on one side than the other, there's obviously a lens element or two that are misaligned in there. Its a dud. Tell them you want another copy, and explain why. They'll take it back.

[booggerg]

What? Are you saying it doesn't cover the full 10D sensor frame? If it covers the full sensor, that's "corner to corner" WRT the 10D which was the subject of the original post and the cited body in the reply.

If you're going to pick nits, make sure they're ripe. ;{)#


Dude, obviously the context was in that lenses tend to lose resolution at it's outer edges. In the case of 1.6x cameras, the outter edge of the lens never makes it on the image so you don't really see this flaw. My point is ripe as it can get. Take your best shot boye.

[Mitch]

Don't assume that you are getting consistantly poor images because a specific lens (i.e., a 17-40 vs a 16-35) is an inferior lens, first assume that the lens needs adjustment or, maybe your camera may need adjustment. If many other photogs are getting great shots from the lens on the same camera that you have, but, you are getting poor images, it probably needs adjustment.
Further, it is very possible that both your camera AND your other lenses may be prone to either back focusing or front focusing but are both out of alignment in coincidentally compensating ways. Then, all of sudden you buy a new lens and you can't get a decent photo no matter what but everybody else does. Most definitely one, most, or all of your equipment is out of alignment...but don't assume that the new lens is not a good lens, especially if it is L glass.

[Jon]

Dude, obviously the context was in that lenses tend to lose resolution at it's outer edges. In the case of 1.6x cameras, the outter edge of the lens never makes it on the image so you don't really see this flaw. My point is ripe as it can get. Take your best shot boye.

Dude, Roanjohn's post you quoted specifically said both 10D and corner-to-corner. Slejhammer also said 10D and "The resolution seems to be there". The context was clear. The original lens was exhibiting low contrast across the board and across the image (on a 10D) at all apertures and focal lengths. Give it up.

[abel]

ugh... makes me want to take a closer look at my 17-40


hmmm