Sharpness Testing

[Skeptic14]

I've been considering upgrading my uwl-04 to a wwl-1 to improve overall image quality with my rx100m2.

I can definitely notice softness in and near the corners on some of my shots with the uwl04, even when stopping down to 6.3/7.1 which seems to be the conventional wisdom with the rx100 and WAL. But I've also read though that the sweet spot for the rx100m2 sharpness wise is 5.6.

So I did a few experiments (both in water and air) to assess if spending $1,500 for a wwl-1 would actually be a good idea.

First, the sweet spot does appear to be 5.6 (no WAL):


Here's at 7.1 (no WAL):



And that appears true with the UWL04 as well, but maybe not as much different:

5.6


7.1



So, what gives. If generally speaking the rx100m2 is sharpest at 5.6, and also the same results appear to match with the WAL, why is the usual suggestion to stop down further to improve the corner sharpness with a WAL?

I did similar tests in water as well and the results were the same; in both instances air or water, the sharpest result was with 5.6 vs 7.1 or 9 with or without the UWL-04.


And finally, while more difficult to compare apples-to-apples the sharpness without the uwl04 is better, so there is definitely room for improvement sharpness wise with the application of a WAL, so the wwl-1 would very well be an improvement, but it appears I should've been sticking with 5.6 vs 6.3/7.1 with my uwl04.

 

[tursiops]

Stopping down too much causes loss of sharpness due to diffraction of the light going through a small hole. The reason to stop down is depth-of-field (DOF). So you are always looking for the best compromise....

 

[Skeptic14]

@tursiops Yeah, that's why I accepted what seemed to be the consensus regarding getting sharper corners by stopping down further with certain WALs and specifically the rx100.

But something easy to see in test photos isn't nearly as easy to discern on a dive shooting different subjects/scenes. And this is the first time I actually tested the theory and it appears I shouldn't stop down any further than f5.6; the overall sharpness doesn't seem better nor sharpness towards the corners.

 

[tursiops]

@tursiops Yeah, that's why I accepted what seemed to be the consensus regarding getting sharper corners by stopping down further with certain WALs and specifically the rx100.

But something easy to see in test photos isn't nearly as easy to discern on a dive shooting different subjects/scenes. And this is the first time I actually tested the theory and it appears I shouldn't stop down any further than f5.6; the overall sharpness doesn't seem better nor sharpness towards the corners.

f5.6 to f8 is one whole stop, i.e. a factor of two less in the amount of light hitting the sensor, but the aperture size for f8 is only the square root of 2 (1.414) smaller in diameter that f5.6, which is what matters for diffraction. F7.1 is the square root of 1.61 (1.27) smaller that f5.6. Not much difference.

If you like the overall sharpness that f5.6 gives you, and can tolerate the lessened DOF and thus lessened corner sharpness, stick with f5.6. It's YOUR picture; you get to decide.

 

[Skeptic14]

I don't see improved corner sharpness in my test shots though with this particular camera; that was my point, my results seem counter to a lot of what I've read aimed for this particular camera.

There's easily discernable overall sharpness decrease and the corners look less sharp too although that's not as obvious, but they are definitely not sharper.

And just to be clear, I'm not arguing against the DOF theory, simply a little surprised to find it appears that the point at which diffraction begins to impact the sharpness on the rx100 is at a lower fstop than a lot of what I've read for best settings; at least that's my interpretation of my test shots.

 

[tursiops]

The way you did your tests is a weak test of corner sharpness, and a strong test of diffraction effects. our image you are focussed on doesn't even go to the corners.....if it did, you'd probably see more loss of corner sharpness at f5.6.