Frogfish wrote:
Since you are using these on a crop sensor then I can say that the distortion is considerably less than the Sigma 10-20 I used to use on my Pentax K5 cameras. The sharpness is not acceptable for any crop - though you are not usually cropping with an UWA and I'm sure some work in PP can raise the levels.
The benefit of the distortion is that you are going to lose less mm in correction so it will act more like a 9-10mm with straight lines. For interiors (I took hundreds of interior shots with the 10-20, it was good enough for my needs) the lens looks fine though and an improvement on any of the 10-20, 10-24 UWA crop sensor lenses....Show more →
Improvement in focal length, or image quality. Some of the shots posted here show very good IQ, but I don't think it surpasses the Sigma 10-20/4-5.6 or the Tokina 12-24/4.
Agree on the distortion btw. I need to test it at 10-12mm to compare to the other lenses, but I want to get a good copy before I start liking this lens too much.
Update. The nice folks at B&H helped me with a smooth exchange. Kudos Henry, I know you're watching. I took a few snaps with the replacement lens and could see some improvement. I set up and took the same test shots again. These are at 8mm and f/8, which is probably the only way I'll be using the lens.
So again, I see improvement. There's still not a lot of IQ change when stopping down, which is indicative of an alignment issue. But I'm not going to exchange again, so I decided to keep the lens rather then return it. Usable images at 8mm is still a major benefit. It will take the place of the 10-20 in my architectural interior bag.
New f/8 (top) vs. old f/8 crop. Taking a second look I think the new one is more processed.
Edit: It's actually the light that's mostly different. The old is bounced off the back wall and the new off the ceiling, which produces higher contrast.
There's still not a lot of IQ change when stopping down, which is indicative of an alignment issue.
The Photozone review also found that central resolution didn't increase when stopping down. The primary benefit was improved consistency of sharpness across the frame.
infimum wrote:
The Photozone review also found that central resolution didn't increase when stopping down. The primary benefit was improved consistency of sharpness across the frame.
Interesting. I am actually seeing an improvement from f/4.5 to f/5.6 to f/8 with this copy, but it's not major. Still, their charts show a decline in IQ as the lens it stopped down. And then there's this. Wow!
Sample Images
Sorry, not this time. We went through no less than 4 copies of the Sigma lens to find one without significant decentering (not counting the also decentered Canon version we reviewed earlier). The decentered units were obviously of no use to shoot sample images with, the 4th and final unit, which is reviewed here, was well centered but unfortunately suffered from a quite strong backfocus issue.
Well, I must say, sigma has not been the leader in QC over the years, but it seems to be improving. Their latest "global vision" seems to acknowledge the need they had to improve that area.
This lens is not within the "new vision" production timeline, so we can still expect to see some mixed results I fear.
All said, I am using mine for the exact purpose you are, and so far, I am pretty pleased.
Elan I don't think there was much difference in two copies of lense , only contrasty lighting in second. Can you compare your copy at 8 mm to your favorite, 16/17/18 mm by stepping back for same FOV?
tjny wrote:
Elan I don't think there was much difference in two copies of lense , only contrasty lighting in second. Can you compare your copy at 8 mm to your favorite, 16/17/18 mm by stepping back for same FOV?
Not much difference, I agree. The combination of it being just a little better and me really wanting the 8mm focal length is what made me keep it.
As far as other focal lengths, this might as well be an 8mm prime for my purpose. Where I used my 10-20, it was always at 10mm and never quite captured everything I wanted to fit in. I'll be using the 8-16 on the wide end only and crop if I have anything to spare, a must technic in architectural in case you have to make perspective corrections, with a side benefit of reducing distortion. I don't need and didn't test the 9-16mm, but no doubt my Tokina 12-24 beats this lens in any overlapping focal lengths.
Generally you shouldn't expect center sharpness to increase as you stop down and approach diffraction. 5.6 is usually a sweet spot, so I wouldn't expect f8 to help in the center. Corners can be a different story.
I have some pics taken w/it on my zenfolio, but it was a while back since I'm on full frame now. I think there might be more copy variation than with some other Sigma glass, simply because of the precision required w/the lens (look at the design and such ahh!) So maybe test in store before you buy.
Lens was def. sharper in corners than the Canon 10-22 and the Tokina 11-16mm (At least, on the wide end) . However, there could be field curvature so use live view for critical corner focusing.
I took a few snaps with the replacement lens and could see some improvement. I set up and took the same test shots again. These are at 8mm and f/8, which is probably the only way I'll be using the lens.