Pixel Peeping The Fuji 18mm XF

Don't Worry Safe For Work For Once.

I've mentioned the Fuji 18mm f/2 XF a few times here. I think I even posted a few random images and mini-review from the first time I ever used it. In any case yesterday's first impressions of the Fuji XPro-2 initiated an email conversation based on what I referred to as a dream kit of sorts. The XPro-2 paired with the 18mm f/2 and the 35mm f/2.

My comments regarding the 18mm f/2 the first time I used it were all about look, as in how do your pictures look. Seems the eBuddy that emailed me wanted me to break that down a bit and quantify why I think the 18mm is so good when the common wisdom is that it's so bad. Honestly I cannot answer that but I can break down why the 18mm is so good far beyond "it looks good to me".

Starting With Some Numbers

Hey, us photographers like numbers right? Graphs are like numbers they're pictures made out of numbers so we should like them even more. It just so happens Fuji gives us a bunch of pretty graphs. Let's take a look at four of them okay.

Okay we're looking at wide-open data here I assume since Fuji doesn't say and folks like Zeiss that used to test every lens and include the tests in the box not only tell you but include other apertures as well. What is this crap? Well it's the resolving power at various frequency of detail, specifically two frequencies. Ya have to sort of extrapolate for crap in between. How About that S and M shit (no pun intended, really, just happened). Well in a nutshell it's the direction of the line detail, google it if you want.

So take these notes in for a second okay?

  • Compare the universally lauded as touched by god 35mm f/1.4 vs the 18mm f/2. WTF? Overall the Fuji says the 18mm is better. Well except for that M line at low frequency detail. Maybe everyone likes low frequency M detail??? You may say no no no look how fast the 18mm falls off and look at that edge all the way over on the graph. Ummm not really. I'll refer you over to the high frequency detail chart. Hmmm.

  • Now compare the universally even more wonderful, or at least just as wonderful 35mm f/2 WR. Well it seems like wide open it really is quite a bit better in the center isn't it. Hold on for a second it's terrible according to Fuji at the edges. I mean really terrible. Horrid. Isn't that what is universally "the problem" with the 18mm????? How come nobody has cast such dispersions on the 35/2?? It's way worse based on Fuji's measurements and I trust their lab better than I trust everyones StupidCrap™ test shots for sure.

  • Moving on just for laughs let's take a look at the 18-55mm. It may actually be marginally better in the exact center but that kind of difference is absolutely not going to be detectable by anyone out in the real world. Looking a little deeper you'll see it's also better at the extreme edges with a giant caveat. It's got this strange dip very near the center then higher at 10mm out. Strange and may be the reason I was initially not very impressed with my own StupidCrap™ test shots.

Also remember we are comparing wide-open only. The 18mm behaves a lot like the 35/1.4, possibly even more drastically with the exception of the far-far-corners. One stop down vastly different, two stops crazy vastly different.

So, let's assume Fuji is not lying. They say it's pretty stinking good. In the same ball park as the 35mm lenses at least sorta. From my observations it's really damn good especially for a wide.

What do edges mean any way vs. the center?? Oh yea we don't take circular pictures do we. I think people get confused about MTF charts and think anything off the center of the picture is somehow going to be bad. Well the edges that are only 8mm from the center are not going to be bad. The top edges at more like 12 might compromised if you actually put something important literally on the edge. I use the edges more than most people and for me it's more like 9 or 10mm which is a different story isn't it. So we really are talking about the corners right? Let's see how this might matter.

As Promised Pixel Peeping.

Let's take the above photo where Melanie wasn't moving around like crazy. Of course I was hand-holding it at 1/125th and sort of laying on the ground backwards-ish and uncomfortable think I focused on her black top if I recall since I didn't give a hoot about her head given 92% of what I made was with M. swinging it wildly.


Feel free to pixel peep by opening in a new window if you want. Looks fine doesn't it. Well that's at f/2.8 while certainly a measurable difference in performance the DOF may be a bigger deal than the absolute performance. Even now not much is actually in perfect focus at f/2.8. Even at f/5.6 not a lot is in perfect focus only 4 or 5 feet away. Not only that but hand holding at 1/125 is probably more compromising that ultimate lens performance.

Just because I promised here's another one at f/5.6


Holy moley, that looks pretty damn good too. Sure, by f/4 or f/5.6 the 18mm is awesome. Even at the edges. It's actually pretty damn good one stop down at what normal photographers on not flat rectangular 3:2 things shot to exactly fill the frame with important high frequency detail in the lower left corner. Again that tiny bit of DOF 4 or 5 feet away helps a lot too maybe more than performance in this case since EVERY one of these is shot upwards and the focus plane is not even close to parallel with Melanie.

One more for old times sake k?


The Real World

As opposed to arm-chair photography that is. Consider the following picture. Hell consider any of the pictures here.

Here are a few points.

  • For all intents and purposes that "center performance" is far more important than that extreme edge performance in any picture remotely like this. It's practically all the center, at least the important part which is Mel.

  • I mentioned DOF a few times. I probably said something along the lines of "there is no such thing as DOF" meaning perfect. Even Fuji kinda agrees. Check your XPro-2 for a dandy new option to set "print DOF". Ever notice the "pixel DOF" default looks nothing like what old manual focus DOF scales looked like. It's virtually nothing. Moving on Mel is obviously not flat. Pick your point of focus, pick your aperture, anywhere not exactly in that plane of focus is nowhere near the performance limits of the optics in terms of rendering detail.

  • As a caveat to the above riddle me this, what's going to be the problem with that lower right corner? Absolutely perfect focus or corner performance? Can you actually stop down enough to really give you great focus of both that corner and Mel?? If you could you would really hate the MTF graph on that not to mention the visual loss of acuity.

  • The one exception to all this is the one guy (definitely a guy, not a female) that takes landscapes where the entirety of the frame is actually at infinity or thereabouts. No body does this. Oh, and the guy that does microfilm like photos of documents that are absolutely flat and 3:2 aspect ration. I don't know anyone attempting that with a frigging 18mm do you?

Bottom Line

The 18mm is a fantastic lens and a fantastic bargain around $250 used for people actually getting rid of it. I love that little lens. You could probably fit four of them inside a 23mm f/1.4 or 16mm f/1.4. You can definitely fit four of them inside an 18-55mm. It's also one of the best rendering 28mm equivalent lenses I've ever had in terms of it's overall look. I've owned a boatload of them too. I would take the 18mm f/2 over the 16mm f/1.4 99.9% of the time all things considered. I know, 28mm is boring, and stodgy, and it's not expensive and impressive looking.