Fuji X100: the dirt on the dirt
Updated: May 24, 2011 (moved the close-range AF discussion to the end to accommodate some illustrations).
One does not need to comment on the X100’s picture-taking ability. If you want to see performance evaluations, some good, some half-baked, some systematic, some useless, and some by people who saw a spec sheet and freaked out (note that the last two are separate categories…) you can go to the usual sites. There, you can also see, variously, sample pictures of liquor bottles, small-town riff-raff, anodyne Scandinavian scenes, flowers, and felines. On this last point, with all of the other strange auto modes showing up on cameras, why no “cat” mode?!
I would shortcut the performance issue by looking at DxOMark, concluding that the camera is fine for most purposes, and move on. Your application may be different and more demanding, covering things we humans can’t even believe. Perhaps taking pictures of attack ships on fire off the Shoulder of Orion. Or capturing sea beams glittering at the Tannhauser Gate. You should probably wait for the perfect camera and herniate, while your photographic prospects are lost in time, like tears in rain.
But what might merit some discussion is correcting (or supplementing) some misinformation about what the X100 is like in person in terms of how it handles. I was able to pick this much up from a couple days. It is not certain whether any of the information or opinion below will impel you to buy an X100 (or not), but there are many things that just don’t appear to be the way people describe them. So here, labeled in a Snopes-style fashion, we have what we hear versus what we see.
1. “The X100 is as big as a Leica M [Hexar AF, etc.].” False. In overall volume, it’s only about 2/3 the size and less than 80% of the weight. In terms of feel, it seems a lot smaller and lighter than it is. It’s still about triple the volume of a GR Digital I/II/III, though. It’s a little bit bigger than a late-model Canonet 28.
2. “The X100’s build quality is fantastic.” Unclear. Expectations regarding build quality must have slipped a lot. The X100 is made of metal, but it has a very lightweight feel in the hand that is suggestive of a Bessa R or a compact SLR of the 1980s. Magnesium does not create the same heft in the hand that the typical silicon-aluminum alloy does. The top-deck switchgear and the aperture ring are excellent quality.
Finishing of the filter trim ring is unremarkable, as is the finishing of the lens cap. These are revelations if you grew up with Sony’s plastic accessories, but they are pretty mediocre even by the standards of third-tier SLRs of the 1970s. All of that said, this feels a bit more solid than the Olympuses and a little less solid than an NEX-5.
By the way, this camera comes with the flimsiest triangular strap rings ever. You don’t need the included spreader to get them on – just your fingernail. By the way, other than some bizarre nostalgia, why not just put a modern strap loop on both sides of the camera (and two on one end, in Fuji fashion)? It seems strange to go crazy making a metal top cover only to make a strap attachment mechanism (rings) that is likely to damage it. And speaking of nostalgia, I am not sure how I feel about the Leica M2/M3 styling. This could have looked a bit cleaner and more modern, like the Hexar RF. And why does the viewfinder selector lever look so much like the self-timer lever on an Olympus 35RC?
3. “The X100’s OVF is nothing to write home about.” False. Are you kidding? It is bright, clear, undistorted, and has the most visible framelines every put into a camera. It also has the ability to project a lot of data that has never been seen in a rangefinder camera, such as depth of field against a distance scale, a digital level, and composition lines. It even has parallax and field corrections. If you use Leica Ms, the X100 ‘s optical viewfinder is like something that came back through a time warp (naked), stole a biker’s clothes, and then started punching holes in brick walls.
4. “The X100’s image-review-in-viewfinder is annoying.” Unclear. You can always turn it off, but it is impressive to be able to see – for a second or three – what you just shot, without taking your eye out of the viewfinder. You take the shot, the blind closes, you see the image, blind opens, next shot. Suum cuique… but do consider that if you set auto-review and b/w simulation, it looks just like in a movie: color image (seen through the OVF) followed by momentary b/w freeze-frame (instant EVF-in-finder review).
5. “The X100’s flash is a joke.” False. Well, it will light up a white wall to white at 15 feet away at ISO 200, so it is not exactly impotent. If you think that’s insufficient, bolt another flash onto the hot shoe. After all, you won’t need that shoe for an accessory viewfinder. Just watch the sync voltage. But bear in mind the following: (1) the hot shoe contact must be activated through a menu – and that option shuts down the internal flash; and (2 ) the locking pin is in the same position as on a Nikon or Leica flash (following the ISO standard).
6. “The X100’s charger is poorly designed.” True. Get the Digipower TC-55F compact charger for about $25. Very small; plugs in without a separate cord; runs on 220; seems to have fully charged a Delkin NP-95 (1800 mah) from new in about 3 hours. It also has a nice, clean display and has a USB port to charge your phone.
7. “You can screw in a 49mm filter upside-down, right onto the filter threads.” Unclear. This is physically possible in a pinch, but (a) it is not clear whethe alloy filter threads will bind to the accessory threads and (b) you can’t use a hood easily. A 49mm Hoya filter, for example, does not seat completely and may end up falling off. Also be aware that if you get too close and try to autofocus, the lens front will bump into the filter and prompt you to restart the camera. There is also an issue with scratching up what is normally the inside glass surface of filters (B+W MRC, for example, is not hydrophobically coated, to my knowledge, except on the front side). Get the double-female 49mm spacer ring from photofilters.com for $5. It’s ugly, it’s crudely finished (like many Sonia products), but it works perfectly (and it’s black).
8. “The X100’s menus are confusing.” Depends. An adult of moderate intelligence (I’ll be the guinea pig) can figure out where the various controls are in half an hour or less and have the camera completely set up in two. What is confusingis having different settings apply in different modes (M, A, S, P), and it is the same thing the Ricoh GR digitals have done forever (MY1, MY2, and then settings associated with the individual exposure modes). Is there only one firmware writer in Japan?
9. “The X100’s manual focus is useless.” Depends or False. It beggars belief that people think to replicate a mechanical rangefinder mechanism with an electronic one. But it is true that the manual focus “ring” is vestigial on the X100. They could make it more functional – but the question is, why?
The theory of a prism rangefinder is that you point the camera at what you want to be in focus, you use a defined focus aid to line things up in a go/no-go fashion, and you shoot. Since rangefinders have no groundglass to illustrate what is happening in the rest of the field, most rangefinder users are doing manually the exact same thing that single-point AF is. Sorry, guys, but if you’re a slave to the RF, you can be replaced by contrast-detect or phase-detect autofocus. View camera users can, of course, move on to management… if you can find that elusive 8×10 film in your favorite variety.
The X100 actually does the same thing in MF mode when you point the focusing spot at the subject you want and hit the AE/AF button. The camera focuses on the selected item and stops focusing when you let go of that button. That cuts the lag and prevents refocusing. It’s the functional equivalent of using a rangefinder, but it just requires different hand movements.
10. “The “shutter sound” is for tyros.” Depends or False. You can tell that Fuji must have acquired some people from Konica-Minolta (or Fuji management spent a lot of time watching Airwolf) because the camera has a “silent mode.” This mode cuts off the AF assist light and all sounds (on the Hexar AF, it slowed down all of the camera’s motors to half-speed). You can also dial down individual sounds. The default shutter sound is a cheezy p/s noise, but there are two others. I am a big fan of sound #2, which sounds like a louder version of the clicks and whirrs that occur in the process of focusing and taking pictures – because at least you know you’ve taken the picture (unless you have instant-review turned on, it’s not always obvious that you shot the picture). And the practical reality of camera noises is that no one really hears them unless they are listening for them.
11. It would look better in black. False. For the first three-quarters of the twentieth century, black-painted (or -finished) cameras were viewed as workmanlike, plebeian, cheap – and with good reason. Chrome was a tougher, more elegant finish – and photographers rarely felt the need to sneak around.
Some people think tha having a black camera makes it more unobtrusive. Arguably, this is false; people are least freaked-out by camera phones and silver point-and-shoots. A black camera signifies professional (or spy). And you lose any element of surprise when you actually try to shoot with anything. Sometimes you never had an element of surprise. To put it in a crude and potentially politically incorrect way, a western tourist in Bangkok shooting is still a western tourist- regardless of whether the camera he holds is black or silver.
Seriously, though, there is a case to be made for chrome cameras where heat dissipation is needed. You never really think about the 140º F maximum operating temperature of your black, electronic camera until you are in Luxor in June at 11 in the morning – with no shade in sight. You will be surprised at how fast a camera can start acting strangely.
Also, the silver color is very popular with infants. Had I kept that little Hi-Matic F, I may have been able to deflect unwanted baby attention.
12. “The menus are too hard to use/inconsistent/anti-intuitive.” Depends. If my 14-month-old son can figure out how to press buttons on a stereo receiver until music comes out, you can probably make it through the menus on the X100 to take a picture. Once you settle on how you want the camera set up, many of the most commonly-accessed controls are buttons and dials. The only two things that are somewhat buried are the on/off controls for external flash and Auto ISO (whose “on” should be popping up in the menu of ISOs when you hit FN).
The menus in general could be a little “flatter,” and that is something that is true of most digital cameras. The Leica M8/M9 menu structure is simpler in that regard – but those camera have intentionally limited settings. Perhaps the easier solution is to do what Kodak used to do on its Nikon-based SLRs – and have a selectable one screen “basic” mode that shows commonly used functions and a complete hierarchy of menus for everything else.
The relationship between the “view mode” button and finder selection lever on the front requires some practice. The switch is actually electronic, not manual, so it only has a resting position and an activate position.
13. “You can’t hit the OK button on the control dial without selecting something else.” True but pointless. It is true that you need microscopic fingers to reliably press the OK button. The catch, however, is that you don’t need the OK button – all you need is the directional control. The way the Fuji menus are set up, selecting an item and pressing left (back) selects it just the same way OK would. And for the controls accessed while using the camera, you don’t need an OK. Whatever option the cursor comes to rest on automatically is selected when you release the particular buttton. The exception is in deleting files, but that should be difficult anyway.
14. “It doesn’t work with Lightroom.” False. But for some strange reason, you first see the JPG preview (film simulation) on importing the files. But when you click on the Lightroom preview, the picture reverts to the raw file. Frustrating, to be sure.
15. “The X100’s OVF has inaccurate AF at close range.” False. Actually, it is fine down to the minimum focus possible in OVF mode, which is not particularly close (in the film era, compact film cameras’ closest AF focus was always about two feet). There are three things at work here. First, what people don’t seem to grasp is that it is an AF zone – so if everything in the zone is at the same distance from the camera (or you get closer to fill up more of the zone), it all works out fine. If you don’t fill up the zone, the camera does not know the focusing intent. This is true of every AF camera ever made, whether a DSLR using TTL phase-detection or a viewfinder-type using external active or passive AF.
Second, there is also parallax error for which to account. You might be totally ignorant of how AF works with viewfinder cameras, but you need to recognize that focusing misses are largely a lack of practice on your part, not a product or design defect. Fuji did a pretty reasonable job at attacking a very tricky problem – what do you do when you are using one optical path to frame and another to focus?
- On a camera where the focusing is external to the optical path (Leica rangefinder, infrared on a point-and-shoot, hybrid on a Fuji GA-series camera), the AF sensing target/point is simply looking to find range to an object and make sure the lens is set, arbitrarily (through a coupling or encoder), to that distance.
- The basic system is one in which the focusing mechanism (zone, square, crosshairs) shown in the viewfinder in the same relative position, no matter what the focusing distance. So when you focus, if your framelines shift or shrink (depending on the camera) to account for parallax and field size, all you have to do is recompose – because the accuracy of focusing has zero to do with what is going on at the imaging (or film) plane.
- Leica-style rangefinders will also move the focusing spot along with the framelines as you close in. But the idea is the same: the you are sensing a distance to which to set the lens, through a mechanical coupling – and as the RF spot shifts, you naturally and continuously adjust the focus.
But the real pain comes when you use a focusing target in the (offset) viewfinder but put the focus mechanism on the image plane. Here is an illustration of what happens:
This is the base case. At infinity (or some preselected distance – the Fuji appears to be set for about 2m), the field shown in a separate viewfinder and captured on-frame is identical
This is field shrinkage (seriously exaggerted so that you can see where the focusing zone ends up). This happens because you are getting closer to the subject than infinity or the preselected distance, effective magnification increases, and the viewfinder or framelines are showing too much. Because Fuji’s preselected distance is relatively close, you don’t see this as much as you might expect (but it’s there). Shrinkage alone doesn’t have a huge effect on focusing accuracy in the center spot, because the position relative to the frame is the same. Other focusing spots change their radial distance from the center – as well as their size.
This is parallax error. Closer than the preselected distance, the fact that the finder is above and to one side of the lens comes into play more. Framelines move to compensate.
This is parallax error plus field shrinkage. Note that the focusing spot (in the middle of the frame) is different from where it appears to be in the viewfinder. You can see how both moving the spot and changing its size could cause a little bit of a problem.
Fuji’s solution was to disable the AF at very close distances when the OVF was activated. This does not entirely solve the problem, but it is not unlike what Leica did in making the M3 stop focusing at 1 meter. If you can’t do something well, don’t do it at all. Think of SLR lenses. There is nothing that theorectically stops you from infinitely extending a lens barrel to get closer focus – but size, weight, and piled-up optical aberrations get in the way.
Finally, as you may have gathered from some of the discussion above, all cameras have mismatches between the focus indicators and the actual focusing sensing. DSLRs are a great case in point – those brackets in the VF can be off a very close distances. Other mismatches occur because focusing mechanisms assume that the lens throws a perfectly flat plane of focus. Most lenses do not.
The bottom line – as it always has been, with every camera, is this: if you find yourself shooting pictures with subjects smaller than the focusing zone or inside of a meter, you might want to switch to the EVF so that you get a true TTL representation of the picture. And as you approach a meter, you should always make sure that the focused subject more than fills the selected focusing zone. Fuji could have gone the extra mile by having the framelines shrink and grow in real time with the camera in AF-C mode (where it continuously focuses, even without the shutter pressed) – and that may be something for firmware.
* * * * *
This is by no means an exhaustive list of things that seem a little off in other accounts of the X100. The camera looks fairly functional so far, and it is a nice piece of design and engineering work by Fujifilm. We’ll see how it goes and report back in a couple of months.