17-50mm Lens on an SD1 Merrill - Sharpness Field.

Scottelly wrote:

xpatUSA wrote:

I'm reading about Lens Field Curvature in articles by Roger Cicala (he of Lens Rentals fame). I've also read comments here about a significant field curvature from the 17-50mm. Cicala mentions that the field curvature of a lens can be shown by shooting grass and then applying Edge Detection to the resulting image:

https://www.dpreview.com/opinion/4042117089/roger-cicala-why-i-dont-use-an-mtf-bench-to-test-my-own-lenses

I intend to do that soon with my copy of the 17-50mm, set at f/2.8 and a few different focal lengths.

But, meanwhile, I tried out Edge Detection on a shot taken at 23mm focal length with the said lens:

The usual test scene ...

Focus was on the far tree tops about 600 ft distant. SPP settings were all at default, including Sharpness. Algorithm was Laplace as found in the GIMP - there are others ...

Not the best scene for Field Curvature, but I'm beginning to wonder why I sold the SD1 Merrill.

Hey Ted, what did you find out?

Nothing definitive, I'm afraid. Couldn't find an homogeneous enough patch of grass.

I found my 17-50mm f2.8 OS EX lens in my backpack. I'll shoot some photos on some flat grasslands for you, and post one or two here. I just might try shooting the flat surface of a parking lot too. I figure that should do it pretty well. I'll shoot a canyon with it too, so you can see how great that lens is for shooting canyons.

Looking forward to those for sure ..

BTW, my Sigma 24-135mm f2.8-4.5 in Nikon F mount has a lot of field curvature too. It works great for some stuff on my D810. Here's a thread I posted with a bunch of samples from that lens:

https://www.dpreview.com/forums/post/62268435

I looked at the edges in the 'excess curvature' example but didn't see why the lamp on the post should be terribly out of focus. 56mm at f/8 is an aperture size of 7mm, so anything in the scene bigger than that will be resolved. One would need to know the distance to the lamp and the distance to the balcony to see if the lamps parts should be well-resolved or not. See:

http://www.trenholm.org/hmmerk/TIAOOFe.pdf

By "resolved" I do not mean in perfect focus - plead read the link carefully especially about Convolution - or "pseudo science" as @jfa3000 often says.

P.S. I just now looked at the lamp in QuickMTF - the whole image is over-sharpened, so I'll refrain from further comment about that particular lens' focus field curvature.