There has been a lot of debate about the new 18-55 IS catching up to and surpassing the 17-40L in image quality. Well, I decided to see for myself. Based on the 18-55 IS hype, I was certain that my 17-40L was long in the tooth and no longer fashionable, being beat by a $150 kit lens of the future. Well, let’s just say the results are interesting.
Everything was shot using a tripod and remote shutter release on a Canon 50D. Click the thumbnails for larger versions.
Test # 1
Test # 2
Test # 3
We are talking about a $700 lens in the 17-40L, and a $150 lens in the 18-55 IS. That’s quite a difference in cost.
There are a couple things I realized when looking at these tests. First, the 17-40L soundly whips the 18-55 IS from 17mm to 24 mm. It has an almost surgical contrast. Secondly, the 18-55 IS absolutely trounces the 17-40 L from 30mm on. It’s an embarrassment in all honesty. The performance from 30mm to 40mm on the 17-40 is a very big issue in this comparison. A lens of this cost should perform much better across the range. In terms of 17-24, the 17-40 just delivers a far more pleasing image even if 100% borders aren’t that hot. Also, the 17-40 just does color MUCH better than the 18-55 IS from 17-30mm. The 18-55 IS delivers a darker, mushier color look and a root beer, almost purple cast to all its images. Also, I would say the colors on the 18-55 IS are not really accurate. It ads a bluish/purple cast, or darker cast to some of the textures (see the brown wooden siding in the backyard test). I can tell you that is not the right color of which the 18-55 is representing.
That said, the 18-55 IS really surprised me. It’s light, has IS, and has better range than the 17-40L. The 17-40L has superior build quality and USM, but needs to be chained to a tripod because it’s slow as a dog (especially with a circular polarizer). The only thing I can conclude from this is “toss up”. If you are looking for razor micro contast and freakish color from 17-24mm(nature landscapes, urban landscapes), the 17-40 is a great lens. If you are looking for outstanding resolution and pop from 30mm to 45mm, and anice, versatile walkaround with IS, the 18-55 IS is the lens for you.
In the end, the only thing that matters is the overall look to the image – not the 100% crops. I plan to post the full unprocessed images from each lens shortly.
Here are the resized full scenes:
All of those are unprocessed and resized images from the originals. Finally, here’s a condensed example of the color cast problem that the 18-55 IS brings to the table:
Note how the wood trim on the 18-55 crop has a weird purple/blueish tint. That is not how the wood and paint actually looks. The 17-40 is representing the real color. It goes to show that there is much more to image quality than sharpness. That cast would not be easy to remove in Photoshop. But the final say is up to you. Clearly, the 18-55 IS has a huge cost advantage.
Edit:
Someone brought up a very good point in that perhaps the color advantage for the 17-40 was because of the 18-55 IS slightly underexposing. This would give it that darker, purple cast. I did some exposure tests and these are the results (click them for the large version):
After looking at the exposure results, it’s clear the 17-40L color advantage is not because of exposure. It just gets the actual real color of the scene. This *may* not hold up past 30mm though, and it will be tested soon.
I’m sorry to say, but this is an invalid comparison with an incorrect result. You have have compared 40% of a EF exstreme wide angle zoom to 100% of an EF-S normal zoom. In other words, you have used the EF17-40 on a body that doesn’t use even half of its picture, then you’ve draw conclusions without stating that this is the case.
I could just as easily claim that the 17-40 is the clear winner because the 18-55 cannot be used with a digital camera, without specifying that by “camera” I mean 35 mm Full Frame equipment.
You should do is to clearly express that you are comparing how the two respective lenses act on small sensor, 1.6 crop cameras, and that the test results only tell how these two respective lenses act against each other when you are using an EF-S camera. In other words, you do not put these two lenses against each other at optimal conditions, but you are confining yourself to the EF-S camera mount.
Comment by L.P.O. — May 13, 2009 @ 8:04 am
LP – I don’t see anything wrong with the comparison in all honesty. Both are on a crop sensor, rending the comparison pretty equal. The 17-40 is actually much worse in the corners on full frame.
Comment by Eric — May 13, 2009 @ 8:42 pm
I think that it’s unfair to compare them as the older 17-40L could be better if it was designed for use on a cropped body. It’s still a great lens for full frame. If you consider some might want to progress to full frame at a later date the lens is still better value as can be transferred and used as very wide angle lens.
The 18-55 is a great lens and i agree that there is no way that you can even tell the difference in most images. But we all know no L glass will ever be an ef-s lens. Take a look at the canon 17-55 2.8, now that is sharper than L glass comparing full frame output to cropped. That is a fair comparison and if you look at the cash involved we are talking similar value.
To this day I think the full frame system is just that bit better for Bokeh and film like feel but there are some real amazing cropped lenses that kinda negate the need to switch up.
alex
Comment by alexgowers — August 15, 2009 @ 4:48 pm