Are Micro Four Thirds lenses sharp enough?

ATMX observed this his high-res photo taken with the E-M5 II has soft corners (although exactly how soft is not specified).

There is no lens I have used that has corners as perfectly good as in the center of the photo, but if there is one lens I’ve used that comes close to perfection, it’s the lens on the Ricoh GR. Of course the Ricoh GR is a rather limited camera. You have only a single focal length (equivalent to 28mm), there’s no viewfinder, there are “only” 16 megapixels, and I also find it kind of hard get the perfect exposure (although you can use exposure bracketing).

As far as I can tell, the sharpest Micro Four Thirds lens for center-to-corner image quality is, believe it or not, the Olympus 12-40mm f/2.8. I say believe it or not because most people would assume prime lenses would outperform a zoom lens, but that’s not the case here. The 12-40 is definitely sharper than the 12mm f/2.0 and the 14mm f/2.5, it’s a close call with the 17mm. I have not carefully compared the corner quality with other prime lenses that I have (25mm and 45mm), so it’s possible that one of those lenses might outperform the 12-40mm.

Even with the sharp 12-40mm, I can tell that the corners don’t quite have the same quality as the center, and I am sure that would be even more obvious if I used the high-res shot mode on the Pen-F which creates an 80MP raw file, although I have not experimented with the high-res shot mode.

It would certainly be very interesting to compare the 12-40mm with images from various full-frame systems like Canon, Nikon, Sony and even Leica, but I don’t have any of those other systems.

Even though Henri Cartier-Bresson said that sharpness is a bourgeois concept, if you are going to be doing bourgeois stuff like making really large landscape prints, then you need the sharpest lens if you want your really large bourgeois prints to be “superior” to other people’s really large bourgeois prints.

* * *

When comparing lenses, it should always be noted that there’s a lot of sample variation, and my 12-40mm might be better or worse than someone else’s 12-40mm. I think that the 14mm f/2.5 and the 17mm f/1.8 are especially prone to sample variation and that I have a good copy of the 17mm and a bad copy of the 14mm. My opinion of my 14mm is that the quality is unacceptable for landscape photos.

Does the Olympus 17mm f/1.8 lens render unrealistic and unnatural images?

A blog post by Robin Wong reviewing the new Olympus 17mm f/1.2 lens states that the f/1.2 lens “manages to render realistic and natural looking results, something I feel is a step up from the older Olympus 17mm f1.8 lens.”

Wow, that’s quite a statement! Phrased another way, the older Olympus 17mm f/1.8 lens renders images that are unrealistic and unnatural.

I find this assertion pretty dubious without some sort of comparison photos so that we can see the difference. People who praise expensive Leica lenses also say similar unverifiable things, they say the Leica lens has a special “Leica look” that you can’t get with cheaper lenses.

I think this could simply be that people are imagining that the more expensive lens produces better images because of cognitive biases. For example, scientific experiments showed that wine tasters rated the same bottle of wine as tasting better when they were told it cost more money.

* * *

It may be pointed out that all wide-aperture lenses produce unnatural results compared to the human eye. When was the last time you ever saw “bokeh” without the aid of a camera lens? To the extent that people prefer images with “bokeh,” they are preferring the unnatural over the natural.

The amazing benefits of an extra 4 megapixels

The Olympus Pen-F has a 20MP sensor, that’s four megapixels more than the previous E-P5. Does an extra four megapixels make a difference? Well just look at the studio scene comparison tool. The Pen-F definitely shows higher resolution compared to the older E-P5 with fewer megapixels.

However, when I tried to reproduce the experiment at home, I was not successful in seeing a difference. I think that once you have 16 megapixels of resolution, to get the benefits of more megapixels you have to be using the sharpest lenses, have perfect focus, absolutely no camera shake or shutter shock etc. Maybe my tripod wasn’t sturdy. Maybe the focus wasn’t exactly the same. I guess the people at are better at setting up these tests than I am.

I saw a photo with fringing

On another blog (no link because I don’t want another blogger to be pissed at me for dissing their photography) I saw a photo with fringing. It was lateral CA fringing, blue on one side and orange on the other side.

I seldom see that sort of fringing in my own photos because Adobe Camera Raw automatically corrects it (if you’ve checked the box). This is yet another reason to shoot RAW and not JPEG.

There’s a meme which sometimes appears on photography forums, that you don’t need good gear if you are just posting photos on the internet, because the small size of web-sized photos will make any flaws in the photo undetectable. Well that was not the case here. The photo was less than 700 pixels wide, but that CA fringe was in my face.

People say you shouldn’t pixel peep (and I agree, in theory), but is it pixel peeping if I could see the problem in a small web-sized photo, even though I wasn’t trying to find any flaws?

One may ask, did it matter? Did it make the photo unworthy? I think, in this case, the answer is yes. The photo was posted primarily for the purpose of showing off. “Look at my awesome photography!” But it wasn’t so awesome with that ugly fringe.

* * *

Reminder: Just because a photo is free of fringing and other technical imperfections doesn’t mean it’s a good photo.