Steve Spencer
Offline
• • • • • • •
Upload & Sell: On
|
 p.19 #13 · Sigma 35mm f/2 DG DN Contemporary Review | |
Jonas B wrote:
This is a post following the discussion (or "discussion") at page 15 and 16.
I have now managed to shoot the IKEA shelf with USAF patches one more time. A couple of the patches were slightly moved to make the distance between them exactly even.
This time with the Samyang AF35 F1.8 and the 35GM only as the three Sigmas I have tried all went back to their owners. So, this became just a test of the method used.
Target: 100% crops of the central part of the 1951 USAF test chart
Camera. A7C - 24MP
Sensor size: don't know really, either 35.6 or 35.8mm depending on source
Target width: 810 mm
Shooting distance: 835mm (Si35i), 905 mm (35GM) and 910mm (SY35)
Effective focal length at the shooting distance: approx. 36mm (Si35i) and 39.8-40.0 mm (sensor 35.6mm).
Perpendicularity: The middle of the center patch straight in front of the camera, the camera perpendicular to the shelf. (This was checked the same way as earlier: by measuring the height and then the long legs of the triangle shaped by the center of the center patch and a point at the optical axis at the cameras position. The long legs differed, in all cases, less than 3mm and typical 2mm.)
Processing: ACR, 100% crops, adjustments for exposure and White balance only, no adjustment more than half a stop, contrast, curves, sharpening and so on at zero or neutral.
Here is the resulting graph (Excel chart so a bit clumsy). Remember that the resolving power of each lens is measured by eye-balling the images so some errors are due to the fact one gets tired of doing this. The resolving power is here without considering contrast.
https://photos.imageevent.com/jonas_b/3xsigma35icompared/graph_v4_2x_35i_and_SY3518_and_35GM_compared_resized.jpg
The third copy of the Si35i was omitted as it has some swing (as inksandpapper would call it) and shouldn't be used for lens testing (or anything else). We can see that all lenses, including the 35GM at f/1.4, were reasonably well centered, well focused and perpendicular to the shelf. We can also see that the lenses touches the limit of what we can get from the 24MP sensor.
The numbers are lower for the 35GM and the SY35 than in my previous post. I found I forgot to adjust the number for focal length and shooting distance last time.
If you find the curves odd you can check the patches here below. Each row is from one single exposure and the 100% crops of the patches are moved together to save some space here.
https://photos.imageevent.com/jonas_b/3xsigma35icompared/USAF_patches_v4_2x_35i_and_SY3518_and_35GM_210425.jpg
I still have the 35GM and the SY35. They are good lenses with some flaws. It's indeed a first world problem but I'm not really happy with any of them.
I do want an aperture ring.
The 35GM has one but just as the focusing ring it fails when it comes to haptics. The 35GM is, simply put, a lens for AF and camera control. Super "sharp" and great when shooting against the light. Easily the best AF of the three lenses.
The SY35 has either an aperture ring or a focusing ring. What a stupid idea! It also has more LoCA (in some situations) than I care for. "Sharpness" is not a problem. AF works fine but a bit slower than the Sony.
The Sigma renders nicely, has the least amount of LoCA, has best haptics with three things actually working pretty well: at the same time a good focusing ring (for an AF lens that is), an AF/MF button possible to use also with gloves and an easy to find aperture ring. However, out of three tested lenses not one was good - if a good lens is defined as a lens capable of taking an image that is sharp from border to border also at a distance around one meter. Well, not a definition but a box that must be ticked for my use. The three Sigmas had serial numbers as follow: 5528xxxx, 5530xxxx and 5531xxxx.
I liked the Sigma and may try a forth copy some day. For now I'm set but at some point I would like to sell the 35GM and get a good Sigma instead....Show more →
Jonas, thanks for your contribution to the thread and for posting your results. I respect your thoughts and your contribution on the forums here a lot. Here is what I see: 1) Your first copy is a dud. No way around it, that is pretty bad performance on the right hand side in #6; 2) I think your second copy might be acceptable to me. Here is what I see in the 7 shots. The best 2 to my eye are #3 and #5; followed fairly closely by #2, #4, and #7; #6 and #1 (which is the worst to my eyes) bring up the rear. I don't see any of the crops being too far off. Yes, there is a fair amount of SA in all the crops, but that has its plusses and minuses. I would love to see how this lens does at f/2.8 and f/4. If it is just SA it should clear up nicely stopped down a bit.
For me this #2 copy if I am right and its problem is SA and it cleans up stopped down as it should if it is SA, would be fine with me. This lens is never going to be great at 1m. They did leave some residual SA and that is going to be worse at close focus distances and wide open. If I can use it at this distance at f/2.8 I would be happy and maybe even at f/4. I often want a bit more depth of field at this close of distance anyway. I don't see the variance in performance across the frame to be that large. Further that this lens has a bit more SA may well have the benefit that at longer distance the bokeh and especially the transition zones might be a bit better. I would like that in this lens.
I think I may well have kept this copy, and been happy with it. That is just me, however, YMMV and very well does.
|
.
), I could see some softening in the center when shooting below 0.5m I think, but I never really checked the borders.
