Godfrey
Well-known member
Re: Fun with the A7/7R
I wonder if the better edge/corner performance of the 28mm is mostly because it's a slightly longer focal length. No matter really, just idle conjecture ...
I put the A7+Elmarit-R 24mm on a tripod this morning and set it level (vertically) and square to cover my bookcase so I'd have plenty of corner/edge detail to look at. I did two exposure test runs: first run I focused on an item near the center of the FoV, second run I focused on an item at the corner of the FoV.
f/2.8 - corners are approximately 1EV down from center and somewhat soft. Center is good.
f/4 - corners are about .3-.5EV down from center, sharper to very good; center is excellent.
f/5.6 - corners are within .1EV from center, excellent; center excellent
f/8 - little change
f/11 - little change
f/16 - slight softening overall (diffraction I am sure)
f/22 - slightly more softening overall
I saw no evidence of sensor-induced smearing, just normal wide-open softness and fall-off at the corners for an ultra-wide lens from f/2.8-f/4. By f/5.6, performance was great everywhere.
There's a small amount of lateral CA; Lightroom 5.3 eliminates it completely.
I did a composite JPEG showing a detail clip of the lower-right corner of the frame. The only processing done beyond the LR5.3 defaults was to turn on lateral CA removal and add 0.7 EV exposure to all seven frames. The image file is large (5100x1200 pixels, 1.5Mbytes) so you can see the detail, I'll link to it rather than embed it:
https://dl.dropboxusercontent.com/u/25268645/sony_A7-elmarit-r_24-corner-performance.jpg
Overall, I'm sure this lens is usable even at f/2.8 for less demanding work (say, trying to work with very dim circumstances and more hand-held, people type subject matter), but for more demanding work f/5.6 to f/11 is the sweet range. Very pleased ... I'll have to repeat the experiment with the Leicaflex SL to confirm that what I'm seeing is purely lens and not sensor, but it's darn good enough anyway. Can't wait to get out and do some real shooting with it.
BTW, this was also my first opportunity to test the Novoflex ASTAT/NEX tripod mount. It's a little fussy to set up since you have to remove the mount adapter and fit it, then replace the adapter with the foot loose. To get it aligned correctly, I first aligned and leveled the tripod head. Then I fitted the camera and turned on the camera's two-axis level display. I rotated the camera until all indicators went green with the camera in horizontal orientation, tightened the clamp, and put a paint mark on the device and on the adapter. Then I loosened the clamp, rotated the camera 90 degrees for vertical orientation aligning the heads up display, and added another paint mark on the adapter.
With index marks now in place, it is very easy to rotate the camera orientation between landscape and portrait with virtually no disturbance to the optical center of the framing or the rest of the tripod setup now.
happy happy
G
Thanks!
I wonder if the better edge/corner performance of the 28mm is mostly because it's a slightly longer focal length. No matter really, just idle conjecture ...
I put the A7+Elmarit-R 24mm on a tripod this morning and set it level (vertically) and square to cover my bookcase so I'd have plenty of corner/edge detail to look at. I did two exposure test runs: first run I focused on an item near the center of the FoV, second run I focused on an item at the corner of the FoV.
f/2.8 - corners are approximately 1EV down from center and somewhat soft. Center is good.
f/4 - corners are about .3-.5EV down from center, sharper to very good; center is excellent.
f/5.6 - corners are within .1EV from center, excellent; center excellent
f/8 - little change
f/11 - little change
f/16 - slight softening overall (diffraction I am sure)
f/22 - slightly more softening overall
I saw no evidence of sensor-induced smearing, just normal wide-open softness and fall-off at the corners for an ultra-wide lens from f/2.8-f/4. By f/5.6, performance was great everywhere.
There's a small amount of lateral CA; Lightroom 5.3 eliminates it completely.
I did a composite JPEG showing a detail clip of the lower-right corner of the frame. The only processing done beyond the LR5.3 defaults was to turn on lateral CA removal and add 0.7 EV exposure to all seven frames. The image file is large (5100x1200 pixels, 1.5Mbytes) so you can see the detail, I'll link to it rather than embed it:
https://dl.dropboxusercontent.com/u/25268645/sony_A7-elmarit-r_24-corner-performance.jpg
Overall, I'm sure this lens is usable even at f/2.8 for less demanding work (say, trying to work with very dim circumstances and more hand-held, people type subject matter), but for more demanding work f/5.6 to f/11 is the sweet range. Very pleased ... I'll have to repeat the experiment with the Leicaflex SL to confirm that what I'm seeing is purely lens and not sensor, but it's darn good enough anyway. Can't wait to get out and do some real shooting with it.
BTW, this was also my first opportunity to test the Novoflex ASTAT/NEX tripod mount. It's a little fussy to set up since you have to remove the mount adapter and fit it, then replace the adapter with the foot loose. To get it aligned correctly, I first aligned and leveled the tripod head. Then I fitted the camera and turned on the camera's two-axis level display. I rotated the camera until all indicators went green with the camera in horizontal orientation, tightened the clamp, and put a paint mark on the device and on the adapter. Then I loosened the clamp, rotated the camera 90 degrees for vertical orientation aligning the heads up display, and added another paint mark on the adapter.
With index marks now in place, it is very easy to rotate the camera orientation between landscape and portrait with virtually no disturbance to the optical center of the framing or the rest of the tripod setup now.
happy happy
G
Last edited:
