When I was looking at getting a lighter travel kit I 'ummed' and 'arred' for quite a while about the lenses I should go with. The body was relatively straight forward - the Canon Digital Rebel series fits the bill nicely (and I continue to be impressed by the XTi/400D). But given that I already had a 17-40mm f4'L' and a 24-105mm f4'L', why wouldn't I just stick with them?
As a general, walk-around lens, I didn't think the 17-40mm would give me 'quite' enough reach for everyday use. And as much as I love the 24-105mm on my 5D for wedding work, it's not a light lens, which kinda defeats the purpose of a 'lighter' travel kit. Anyway, to cut a long story short (although this is going to be a long story anyway), I read some very good things about the 18-55mm IS and the 55-250mm IS - both very light (and very cheap) image stabilized lenses that I thought would fit the 'travel' bill nicely.
Of course, now that I've got all four lenses, my curiosity has got the better of me. I do want a lighter travel kit, but I also don't want to sacrifice too much in the way of image quality. After all, why bother lugging any camera gear around with you if you're only going to be disappointed with the end results!
So, armed with the EF-S 18-55mm IS and the 17-40mm f4'L', I decided a lens shoot-out was in order. I was prepared to take the results as they came, and make my decision about which lens to use based on the cold hard facts. Would the 'kit' 18-55mm lens prove as good as many claimed, or would my $1000NZ more expensive 'L' lens kick sand in its puny little face?
First, let me give a quick rundown on how I performed this 'less-than-scientific' test. I didn't shoot lens charts, resolution charts, or colour charts to determine which lens was 'best'. Instead, I took them both outside and did what was important to me - I actually took some real photos with them. I found a subject that wasn't going anywhere in a hurry, wouldn't exhibit any shake from the wind etc, and set up my tripod. I switched the 400D to remote release and for all the images taken I used the RC-1 infrared shutter release so I wasn't physically touching the camera when the image was taken. I didn't use mirror lock up because a: I never do, and b: I think it's overkill. Above is the scene as taken for my first series of test. Both lenses were set to 18mm.
It may be difficult to tell much from a low-res image on a computer screen (click on the image for a larger view), but if you're thinking "gee, there's not much in it" then you'd be right! I've looked at all of them at 100% on my screen (obviously), so let me tell you what I can see (and you'll just have to take my word for it).
On the edges, the 17-40mm is sharper than the 18-55mm, and exhibits a touch more detail at both f4 and f8 - just. At f11 and beyond the 18-55mm sharpens up to be practically identical. There isn't a hugh difference to be honest, and the 18-55mm equips itself very well against its $1000+ more expensive 'L' cousin.
The identical conclusions can be drawn from the central part of the image with both lenses at 18mm. Again, the 17-40mm is sharper up to f11, and then both are identical. Saying the 17-40mm is sharper up to f11 makes it sound as if the 18-55mm is crap - but it's only 'just' sharper. And these results are unedited jpegs straight from the camera. Nothing has been applied in photoshop.
As a quick test, I did apply some sharpening to the 18-55mm files (150% with 0.5 radius) and then compared them with the unedited 17-40mm files. Basically identical. I gotta say - I'm impressed.
I then zoomed both lenses to 40mm and took another series at all apertures. For the sake of sounding like a cracked record, the results were the same. There's no doubt that the 17-40mm has a tiny edge over the 18-55mm at apertures below f11 - but it's a tiny edge indeed. Amazing for a lens that cost literally a fraction of the price!
Of course both lenses are not created equal. The 18-55mm IS is a plastic fantastic with lightweight materials and a plastic lens mount - no depth of field scale, and difficult manual focus. The focus motor, whilst reasonably snappy, is audible - the front lens rotates, and it only has 1 aspherical element in its lens construction (although to be fair it does a very good job of controlling aberration). What it does have over the 17-40mm f4'L', of course, is the IS function.
The 17-40mm, on the other hand, is a solid, well constructed lens with silky smooth manual focus, full-time-focus override, a non-rotating front element, weather sealing, 2UD and 3 Aspherical lens elements, a silent USM motor, with a metal lens mount and distance indicator window. To look, touch and use the two lens is like comparing apples with oranges - or chalk with cheese.
Of course this also goes a long way to the $1000NZ price difference between the two lenses. But is that enough? Because when we look at the actual images taken with both lenses, the $1000 difference become less obvious. You are certainly not getting a thousand dollar image difference between the two lenses. Fortunately, I think that's a good thing.
The EF-S 18-55mm f3.5/5.6 IS is an incredibly solid performer - image wise - for the price. It's an obviously mass-produced, low-end 'kit' lens that is all but thrown in with a new Canon digital SLR, it's lightweight, and won't stand up to a lot of abuse. BUT - the images you can get with it are on a par with images taken with the much much more expensive 'L' series Canon lenses. A quick sharpen and slight curves adjustment in photoshop and you'd be hard pressed to tell the difference.
That makes the 18-55mm IS amazing value for money in my book. Too many photographers are told that their kit lens is 'crap' and to get rid of it asap. If your kit lens happens to be the Canon EF-S 18-55mm 3.5/5.6 IS lens, then you might just want to hang on to it for a bit longer.