Fast, useful lens test.

[Belgarchi]

To evaluate field curvature is very important to rate and to use a lens.

If, as often, especially for wide-angles at large apertures, the corners of an image are less sharp than the center for a flat subject, you should determine if it is due to field curvature or due to other aberrations.

Field curvature is less of a problem if you know it is there, allowing you to get objects in front of the focusing plane sharp.

The procedure takes less than a minute: put a detailed subject in one corner of the image, focus on it, look at the same subject placed in the center of the frame. With my Fuji X-T3, no need to take a photo, the viewfinder is good enough to see immediately if there is field curvature (you must change the focus to reach maximum sharpness in the center) or not.

 

[Tomasg71]

It s a matter of seconds for me (Sony a7mk2), if i do it correctly that is. I focus on the subject, which is usually never at infinity, using the max magnification, i then move the magnification to a corner that has some detail in it, usually one of the bottom corners. When i see it s soft i refocus toward infinity, if it s field curvature it sharpens up, i then move it back to the subject and evaluate how much softer it is....based on the importance of what should be sharper i then choose a compromise between the center/subject and corners. I find that all older wide angle lenses that i own have field curvature even at f5.6 and f8 so i tend to stop down to f11 or even f16 sacrificing overall sharpness for a more homogeneous image when there is something of importance in the corners, which is seldom. Of the modern lenses i own the Samyang 35/1.4 and 24/1.4 have almost no field curvature but the 14/2.8 has a lot of it.

 

[E Dinkla]

For not too far distances and wide open; a level gravel path in front of you and the camera level but forward tilted, curve shows more or less in the gravel detail. That is what I see used by some people and it is quite good based on the images I have seen.

Edit: I see Roger Cicala does not object to that method either: https://www.lensrentals.com/blog/2019/11/stopping-down-some-bargain-primes-and-zooms/

Met vriendelijke groet, Ernst
No photographer's gear list is complete without the printer mentioned !

 

[RCicala]

You guys are going to love my DPR editorial coming out in a couple of days. Basically describes using both tests and why I think they're the best way to test lenses; with a bit about why I actually prefer lenses with field curvature.

 

[E Dinkla]

You guys are going to love my DPR editorial coming out in a couple of days. Basically describes using both tests and why I think they're the best way to test lenses; with a bit about why I actually prefer lenses with field curvature.

All ears/eyes for that editorial. Then there are VFC lenses ,,,,,,,


Met vriendelijke groet, Ernst
No photographer's gear list is complete without the printer mentioned !

 

[ProfHankD]

For not too far distances and wide open; a level gravel path in front of you and the camera level but forward tilted, curve shows more or less in the gravel detail. That is what I see used by some people and it is quite good based on the images I have seen.

Edit: I see Roger Cicala does not object to that method either: https://www.lensrentals.com/blog/2019/11/stopping-down-some-bargain-primes-and-zooms/

I'm also a fan of the road surface photo test (well, any textured surface will do).

My blacktop driveway (which was only really black for the first year or so) works great. Curvature is pretty darn obvious from live view peaking.

However, there are two ways to use it. One photographs the surface at a low angle, like Roger Cicala suggests . The other way points straight downward at it -- which doesn't tell you which way it curves, but makes it easier to spot tilts and other optical misalignments. In any case, even a sloppy shot of such a texture shows field curvature really clearly in the focus-peaked live view and in images; here's one with my Fujian China TV Lens GDS-35 35mm f/1.7, which has crazy field curvature on APS-C:

 

[Photodog2]

I prefer lenses with field curvature too and sometimes add it in post.

 

[Lumens77]

I tend to prefer flat field lenses which facilitate landscape at infinity. But I must admit that a lens perfectly corrected for field flatness might have sacrificed another strength.

My Minolta MDIII 50 f2,0 has very nice field flatness compared to MD II 45 f2,0 I recently bought. But I am discovering there is a kind of micro-contrast delivered by the 45mm that is quite pleasing so this lens might have been less corrected on a parameter enhancing some other aspects. This is on A7II and these points might have been irrelevant in film days.

I must add that I had to buy three copies of the 45mm to obtain an homogeneous field curvature. The other lenses had a tilted focal plane.

 

[E Dinkla]

You guys are going to love my DPR editorial coming out in a couple of days. Basically describes using both tests and why I think they're the best way to test lenses; with a bit about why I actually prefer lenses with field curvature.

The two recent articles on field curvature Roger refers to.

https://www.dpreview.com/articles/1351719699/roger-cicala-field-curvature-for-fun-and-profit

https://www.dpreview.com/articles/1351719699/roger-cicala-field-curvature-for-fun-and-profit

For a practical solution off axis focusing is mentioned in the first article but little more at which point then, one example there at halfway to the long edge with a 2:3 aspect ratio sensor. I have used another rule in the past that fits the rule of thirds composition as well, so 1/3 of the frame length off the edge. Neither method does justice to the different shapes curved fields can have and worse, with a slight tilt in the field off axis focusing can be worse than focusing in the center. You have to know your field curved lens damned well and remember its particular flaws for the cases that require more of the field to be in focus. On the other hand it will hardly be needed for night scenes that ask for a wide open shot. Astronomy as an exception. With portraiture etc the field curvature can be positively used but then you have to know your lens as well.

Most of my efforts to get it right for tilt and field curvature was actually with a scanner, my Nikon Coolscan 8000. I made a wet mount film holder of a converted Nikon film holder and with the way better film flatness I discovered there were flaws in the tray design the film holder was kept in during the scan. For the first 6x9 frame scanned the holder would hang a bit lower at the front, for the second frame a bit at the rear. Tweaked that as good as possible so both would have an equal difference but really flat was impossible. The other axis, short one, left right, the tilt was easier to get correct. All finished I noticed that there was field curvature over the short width 56mm, so I used the 1/3 from the side focusing. I used Vuescan software as it had finer units for focusing distance than the Nikon software. Static situation with one fixed lens so worth the trouble and less demanding of my discipline.

There is a brick wall near the back of my house where I glued a laser aligned mirror in the center, a bit recessed in a brick. So whenever I do a brick wall test I can put the camera in front of it, make sure the EVF shows in the center the camera lens front reflected from the mirror. Take four shots with one corner focused. It may not be scientific enough but some vintage Olympus OM wide angles were sold again when their tilt was unacceptable.

Sometimes I think there must be an old church or cinema somewhere to convert to a pixel peepers paradise with an online cashier to make the appointments. Customers have to be devoted lens lovers. Pixel peeper not being a curse there.

Met vriendelijke groet, Ernst
No photographer's gear list is complete without the printer mentioned !