UPDATE: I explained my methods poorly in this article. Updated portions will bear the italicised text, 'UPDATE', and a short explanation. My apologies.
Initially, I purchased the X100s in order to stay stealthy when covering hi-resolution audio events where literally any noise distracts from the amazeballs of extra ultrasonic information and placebo. Then, after falling in love with its one handability and sneaky little lens, I started using it for everything else.
It was about the time I reviewed the DSD-compatible LinnenBerg Vivace DAC that I realised all was not right with the X100s. Easy breezy as it may have been to pop the shutter button, it did this thing under manual focus at macro distances that caused no end of frustration.
This thing happens to miss focus at ALL aperture settings.
Putting down a HiFi footprint, the Vivace wasn’t too tough to work with. I could miss the big G in its trademark and still come away with something usable. Real problems surfaced when I put the X100s in front of a pair of custom-fit Noble Audio K10 earphones, and later the custom-fit Vision Ears VE6 earphones (admittedly, which were taken in shitty light).
Quiet-like, the X100s would, post-press of the shutter button, shift focus from what appeared in the EVF to be 100% correct to something soft, and, largely unusable. The solution: focus bracketing 5-10 shots, was, and is, time-consuming, and utterly frustrating.
BTW, here is a typical view from the LCD (the EVF being the same) which shows an out-of-focus subject. This image ended up being the best of the bracket.
Because the focus ring isn’t labelled, I couldn’t off set the focus by a set distance relative to the shift. And because X cameras interpolate information in the EVF at high magnifications, nailing perfect focus is already quite a crapshoot. Still, I’ve been using X cameras for about 2 years. I know what I’m doing. But the X100s has been nothing but a doozy for close-ups of audio products .
Most of my day job (shooting audio gear) is done on bellows, at macro distances. The X100s’s tiny fixed lens should have posed no problems. But it did, and it does, for non-work related close-focus things. I’ve tried three different X100s models at Yodobashi; each behaves the same. I have not tried the X100T at macro distances. As it uses the same optical formula and ridiculous fly-by-wire system, I assume a similar, and detestable, behaviour.
For illustrative purposes, here is the basic setup:
Subject: Leica M240 on a bit of scratched aluminium
Imagine device: X100s (excepting the scene photos, which were taken by an X-T1 / 35/1,4 and iPhone 6)
Tripod: Gitzo 3-series
Tripod Head: Arca Swiss Cube
Trigger: Flashwaves III
Lights: Profoto Pro-B head
Generator: Profoto B2 1200
At longest, the lens-to-subject distance is ~19cm. At shortest, it is ~15cm. In this particular example, I had very little trouble nailing the M240’s accessory shoe. So it goes. The other subjects, the M button, and the ‘LENS’ text on the bottom of the Canon 35/1,5 LTM, were wonderfully sharp through the EVF. Beside the accessory shoe, each turned out shit unless I bracketed.
The behaviour of the X100s lens once focus is apparently nailed is this: beep beep boop, whilst fluttering its aperture blades and moving the lens back and forth a bit. That’s where things myopic. If you’re more thorough than me, you can do you own test, and measure the distance from front of lens to subject, then subtract the difference prior to the shutter opening and closing. The world will rejoice.
For reference, here's a shot of the setup, showing the X100s's bonnet. (I'm afraid it's missing the Flashwaves III radio trigger.)
Here’s one showing the back of the X100s.
The Leica M240 and Canon 35/1,5 LTM really is a beautiful hug of metal and glass, isn’t it?
NOTE: I took 8-9 images per bracket. In order to keep this post short, and not melt your eyes with loads of fuzzies, I removed over half of them. You needn't worry, though; only in the accessory shoe set was more than one image from a set reasonably sharp. Each focus bracket was focused manually through the EVF and/or LCD, and the camera captured the image at f/8 at 1/250 and ISO 200.
UPDATE: each bracket was shot separately, and manually focused on one of the three: the M button, the accessory shoe, or the LENS text at the very bottom of the Canon lens. I did NOT shoot the LENS text and expect the accessory shoe to be in focus. It is more ridiculous to imagine that a person that shoots this sort of thing day in day out for a living would expect all those items to be in focus at the same time.
Here are crops of the focus bracket taken of the M button:
Here are crops of the focus bracket taken of the accessory shoe serial number:
Here are crops of the focus bracket taken of the CANON LENS 'LENS' text:
Going forward: I hope that Fujifilm will tell its fly-by-wire-happy engineers to fuck off. Making something properly, something that doesn't beep beep boop like a point & shoot camera, resetting its bits and bytes every chimp, every on/off cycle, and that doesn't do what it wants when specifically told not to - that's something I could get behind.
But again, I'm in manifoldly in the minority. To everyone else, Fujfilm can do no wrong. Fly by wire is the god's fucking knees. And look at that cute exposure dial! Damn!