After getting my Sigma 18-50 2.8 and 24-70 2.8 and being very impressed, I was still curious to see how they competed with the L equivalents side-by-side. I have owned a 24-70L, but sold it when I got the Sigmas as the quality of the Sigma seemed good enoug. But I never involved them in a direct test.
Yesterday, I met up with a friend, who also brought his 24-70L and 17-40 f/4. The tests were all handheld, but with adequate shutter speeds to eliminate shake. All the shots were taken within 15 minutes of each other. ISO 200. All focusing was on the tower on the left, using AF. (I do see that in the 24-70 tests, there must have been some AF error in two shots.) The tower was about 100 metres away from the camera.
These are all 100% crops, (give or take 1%). Please realize that this is very demanding on any lens. Of course, the unresizes originals on my computer monitor look much sharper than here, but, again, all the shots were subjected to equal treatment, so it is a fair test and comparison.
These are not precise, scientific tests. They are informal and real-world. Whatever their limitations, both lenses were subjected to IDENTICAL conditions and treatment. There was no bias whatsoever toward any of the lenses. If you do percieve any differences, it was unintentional and random. There was nothing done in Photoshop other than cropping, adding text, resizing and saving.
You can make your own conclusions as to the results and discuss it here, if you wish. You can ask any questions you want and I'll try and reply.
20Dshooter wrote:
You can make your own conclusions as to the results and discuss it here, if you wish. You can ask any questions you want and I'll try and reply.
I have also tested these lenses, and after having done so I saved for three months more and bought the Canon 24-70 2.8 L. It has a better build quality and faster autofocus.
The sigma 24-70 doesnt look to bad but you can see what the extra $800 or so buys you in the 24-70L, (aside from USM, weather sealing etc)
The 18-50 test though is really hard to make a judgement due to the fact that the sigma is in warm direct sunlight and the Canon is in diffused light. Additionally the difference of 10mm presents quite a difference in subject size
I would really be intrested in seeing more between the 17-40 and the 18-50, particulary in corner sharpness if you still have the ability to redo some test
I have the Sigma 24-70 and find it to be a great lens for the money. Never found the need for weather sealing on it (and use it a lot outside). The AF motor is more then enough for the shots I take (mostly buildings or objects sitting still) When I do use it for moving subjects it keeps up just fine.
Thought about the Canon 24-70L, but every one I tested (3) I was not impressed enough with, especially for the price it costs.
Is the Sigma perfect? NO, but at its price it is more then capable of producing sell-able photo's.
gfiksel wrote:
The top line of Sigma images at 24 mm - the shot at f/5.6 is much less sharper than at f/4 and the sharpest is at f/2.8 Is the order reversed?
bogatyr wrote:
I have also tested these lenses, and after having done so I saved for three months more and bought the Canon 24-70 2.8 L. It has a better build quality and faster autofocus.
These are my impressions of the differences between the two.
Canon built the L's like tanks, but Sigma EX are built like cars. So, well, yeah, I guess Canon has a better build quality. Just keep it in perspective.
Same thing with the Autofocus. Canon's AF is instantaneous, whereas the Sigma's AF is lightning-fast. In absolute terms, Canon has faster AF, but in the real world, I didn't notice any significant difference.
The Big Bad wrote:
The 18-50 test though is really hard to make a judgement due to the fact that the sigma is in warm direct sunlight and the Canon is in diffused light. Additionally the difference of 10mm presents quite a difference in subject size
True on both counts. I only noticed that when you mentioned it. At the time, the sun was going down and I was racing against it.
Also, I originally wanted to conduct the tests so that all focal lenths would be of the same field of view. In other words, move closer for wide angles and further away for the tele ends. In the end, there just wasn't enough time to do that, or I would have lost too much light to get good shutter speeds.
I would really be intrested in seeing more between the 17-40 and the 18-50, particulary in corner sharpness if you still have the ability to redo some test
Unfortunately, not at this time. There isn't much detail in the corners of these shots. Perhaps another time.
Remember also, that the Sigma 18-50 is a 1.6x crop lens. Unless you're asking for 20D corners.
I went through two Sigmas to get a good copy; first was softer that the Rebel kit lens (even with the 18-55 wide open!) My second copy is much sharper and is comparable to my 50 1.4 by f4 and still sharp wide open. That said, I'm still very tempted by the 24-70L.
gfiksel wrote:
The top line of Sigma images at 24 mm - the shot at f/5.6 is much less sharper than at f/4 and the sharpest is at f/2.8 Is the order reversed?
I looked at the originals, but, no. These are the corect order.
BTW, f/4 looks sharper to me. 5.6 is probably an AF error.
I've been quite happy with my Sigma 18-50 f/2.8. It was my first Sigma lens and has made me more confident in Sigma products, at least the EX line.
The build quality is excellent, though is not weather-proof like the Ls.
Sharpness is excellent in the centre at all apertures. Corner sharpness is soft wide open at f/2.8. Perhaps this is where the 17-40L is better since it is a FF lens. However, by f/4 the Sigma is fine in the corners, but at least you can get down to f/2.8 to make the shot or get some shallower DOF (in which case corner sharpness is unlikely to be a concern).
I do wish it had USM and supported FF. But then again, it wouldn't be $200 or so cheaper than the 17-40L.
