The first part of this review looks at the physical aspects of the X100. Here, I take a look at the pictures captured with the camera, and there’s a bit more discussion on how I found it to use.
Firstly, I’ll say again that the quality of the images captured from this camera are superb. I’ve not used a micro four thirds camera, but I understand that you have to be a bit careful above about ISO800. With this camera, I’d be happy using it at ISO3200 when it needed to without any worries.
So, image quality is brilliant. Job’s a good ‘un, right? Well, not quite…
This post has lots of images in and is split into a few sections:
The first set of photos on here are shot with the X100 (in RAW+Fine JPEG mode), but I processed the RAWs as I would my DSLR files. Initially, my focus was on seeing whether I could use the results in the same way as I use my DSLR images, so I went straight into processing images that looked like they had potential. So they are not straight out of camera, they have had varying amount of processing.
I was starting to get the hang of the controls – there’s a lot of them and in all sorts of places, but I’m a technical sort of chap and I’m happy to spend a week or two letting a complex set of controls become natural. And I’m also more than happy to admit that in the first few hours I’m not going to be as adept as using the camera as someone with a lot more experience. So in my quest to decide whether I want an X100, a complex set of controls cannot be seen as a down-side.
From these shots, you can see that you can get decent shots. At least I’m happy with them! The shot above was deliberately focussed on Timmeh the Cat (mainly ’cause I couldn’t quickly remember how to adjust the AF point, and on a DLSR I almost never shoot with the centre AF point so it was off to the left in the shot above).
Here’s one that shows a bit of a problem for me. I’ve learnt an awful lot from shooting my kids – partly how to capture fast moving people without interrupting the flow of action. I have my ways of doing it – a bit of camera setup, a bit of camera technology to help me out and a bit of prediction and situation manipulation to guide them into the right places. Here, my youngest was pulling funny faces and giggling lots… a shot I could have easily caught with a DSLR, but with the X100 I missed it due to the AF hunting around. By the time the camera fired, he had fallen half out the frame (and moved forward a bit, so he was out of the focussed position anyway):
One issue I notice with this camera is that to focus less than about 1m, you need to use the macro mode (Graham had warned me about this). Whilst this is not a massive problem (it’s a couple of button presses to change the mode), when the camera focusses in macro mode you do not get a live view of the focus – the image just locks on the view you had before you pressed the button. If you’re trying to capture a child moving fast in front of you, after the 0.5-1sec or so it takes to focus, when you’re presented with the live view image again it’s all changed! In reality, showing the image wouldn’t make a lot of difference as you can’t really chase a moving target when doing contrast detect AF as it just isn’t fast enough.
How to get around this…? Pose him against the wall and get him to say silly words
– and it’s more or less fine. A point to note – I shot most of these pictures at f/2, partly because it is the most difficult aperture for AF on this camera, and partly because I love shallow DoF shots.
So you can take shots of kids, but you might have to pose them or catch them when they’re pretty much stationary. For me, this is an absolute killer. I’m convinced that with a few weeks of ownership of an X100 I could get significantly better at focussing and change my shooting style to get a lot more shots I’m happy with, but I just love the ability of a decent DSLR to be able to lock focus in an instant even at ultra shallow DoF’s like you get with the aperture at f/1.4.
The shot above shows something else too – the fill flash. It fired for this shot, and you can just see some flash shadow behind his right ear against the wall. The fill flash really is rather good – it fired on lots of the shots I took and several of them you’d be hard pressed to notice there’s any flash assistance. The two issues I noticed were (a) that you have to be careful as it’s close to the lens, so you can end up with red-eye, and (b) you can’t use it with the lens hood. The hood blocks a big chunk of the flash light and rather messes the picture up! You also have to watch out for those flash shadows, but the flash is very close to the axis of the lens so they often do disappear behind the subject.
Incidentally, all the processing I’ve done on all these photos are using Lightroom 3.4.1 (the latest at the time of writing).
As well as being rather taken by trains, my son is also a gadget freak and has a thing for my Galaxy S2. Again, in this shot below there’s a bit of flash shadow but on the whole the lighting of the shot is very pleasant – a mix of natural light and gentle fill:
We finally got out and about so I could take some outdoor shots and see how the camera handles when situations are a bit more static. As I was starting to get more of a feel of the controls now, I was quite enjoying myself. It was really nice to have a lightweight, compact camera to use rather than a big DSLR hanging off my shoulder, and the technology in the X100 made it feel like there were a lot of possibilities for shooting. Out in this sort of light, the AF seemed to lock reliably and generally in a few hundred milliseconds, especially if there wasn’t much movement between shots.
Focus seemed accurate to me when it locked – the great thing about using the imaging sensor to perform the AF is that a lock of the AF system will mean that focus is achieved at the image sensor - no worries about front or back focus here! There were a few instances where the lock was obviously on the wrong thing (a bit of background usually), but it was rarer than I was led to believe from other reviews and didn’t seem to be any more of a problem than using the phase-detect AF in a DSLR.
Sometimes, it can go a bit awry – the shot below is in aperture priority mode with auto ISO at 1/680s, f/2.0, ISO400 with flash doing a fair amount of the lighting and leading to a strong and ugly flash shadow:
In essence, when you capture a shot that’s in focus, at the right time and without any dodgy flash because you’ve left the setting in the wrong place, the files are as versatile as anything from any of my DSLRs.
The last shot I took when out was this one in the car. Ooops – it went a bit wrong here. I’m not quite sure what happened – the strong backlight obviously threw the exposure algorithm, but it’s massively underexposed – even the brightest point in the image is dull. (Camera settings: Av: 1/850s, f/2, ISO800):
In order to sort this image out, I had to up the exposure by over 2 stops in Lightroom. Or clicking “Auto” in Lightroom actually gave a nicely balanced image with an exposure increase of +1.2EV and a brightness increase from +50 to +118.
But… I later found a reason for the problem in this shot, which I’ve explained a bit below.
Now I wanted to take some shots that pushed the ISO up to see how the images seemed. There were some bottles on the side, so I lined them up and took this shot – it’s at ISO1250 (1/30s, f/2.0). It shows some noise but not a considerable amount – I’d say it’s pretty much on a par with my 7D. This image is generated from the RAW in Lightroom, but with no change in the default settings, so it hasn’t had extra noise reduction applied. The JPEGs from the Fuji (in the default setting) have some noise reduction applied so I would expect them to look a fraction better.
Now on to ISO3200. This is a shot down the street at about 11pm (1/3s, f/2, iso3200). A few things stand out. Firstly, at 1/3s there was no way it was going to be sharp! Second, the white balance does what I would tend to do when processing the image in Lightroom – it balances out the orange from the sodium vapour streetlights (unlike my Canon DSLRs where you end up with a very orange cast to the image). This does have a disadvantage, though: the colour rendition index of a sodium vapour streetlight is very low, which means if you remove the orange cast there are certain colours that will look quite unnatural as they wouldn’t have reflected any of the orange from the lights – blue for example. The final image is an attempt to look natural in a very unnatural lighting situation, so in some ways it would be better to leave the orange cast to show the real view. Whatever the case, the whitebalance settings in the X100 would allow you to choose which way you wanted the image captured if you were shooting JPEG.
These next two images show something of an issue with the whitebalance under some fairly difficult circumstances. Both shots were taken a moment apart at 0.5s, f/2 and ISO3200, lit by streetlight. In the first, camera chose a colour temperature of 2450K and a (Lightroom) Tint of -17 (towards green), and I would say it’s going in the right direction here:
I’m not quite sure what was going on there. Like I say, it is an incredibly difficult shooting situation for the camera – lighting from a street lamp with a few spectral peaks around orange, and in the dark at ISO3200. I wouldn’t consider this a fault of the camera, but it might be something to watch out for in low light.
On a more positive note, the AF performance in the dark was surprisingly good. When I say dark, I mean 1/3s at f/2, iso3200! The shot below was focussed with the AF system and it locked immediately. To be honest, even dark there’s some quite high contrast edges so it’s a reasonable target to lock on to (I focussed somewhere around the writing), but I’m not sure my 5Dmk2 or 7D would do a good job in this sort of light. The colour balance is a little off, but even so for the shot settings it’s not a bad image.
And this was a final ISO3200 shot of my Clavinova with a few tweaks to the whitebalance and clarity in Lightroom, but no extra noise reduction applied. I’d be more than happy with this sort of shot from any camera:
The camera has a macro mode which I have talked about using to capture people when close. It does allow you to get close to subjects, but the minimum focus distance feels very big for a 35mm lens, so in reality it’s a “close focus” mode rather than macro. It is important to use this mode under about 1m, so the macro mode would get a lot of use with any kind of close portrait work – but it does have that disadvantage that all EVF/back screen updates stop during focussing in this mode. Having said that, to me true macro is a dedicated subject that requires both forward planning and fairly specialised kit, so I wouldn’t assume that the X100 is any kind of replacement in a true macro sense. But you can get some pretty pictures nonetheless:
While I was sorting out all the pictures for this review, I was thinking back to the view of the picture of my son in the car above that came out very dark, and I didn’t remember it looking that dark on the viewfinder of the X100. I had a quick fiddle around, and as I was shooting in RAW+JPEG I could compare the JPEGs from the camera. I was pretty surprised by what I saw. The 4 pictures below show a split of the JPEG from the camera, and the RAW image processed in Lightroom 3.4.1 with all the default settings. In all cases, the darker (generally lower) region is the Lightroom output:
While the second picture is very close between the camera and Lightroom, the last picture took over 2 stops of exposure increase in Lightroom to come close to the X100 JPEG!
Now, I can think of a few possible explanations for this situation. The first is that Lightroom might be doing a somewhat strange job of processing X100 images (which I find quite possible). The second is that the X100 takes a guess as the exposure for the shot, exposes as predicted then analyses the RAW image and makes further adjustments in “slow time” to create the JPEG. The latter worries me a bit if it’s true, because it means the camera is artificially producing what look like decent exposures, but on careful viewing would show significantly increased noise compared to other shots, and that would make the low light capability of the camera inconsistent. And unreliability is, in my eyes, a complete and utter killer.
Firstly, I want to say a massive thank you again to Graham Fry for loaning me his much loved X100 for a couple of days so I could find out for myself what it’s like to have one. It’s been a thoroughly enjoyable experience, and also one that has (at least for the moment) saved me a lot of money! You see, I’m not going to get an X100.
Going back to the very beginning, what I was after was a camera that replaced my “big and bulky” DLSR. I put that in quotes, because up to the time when the smaller quality cameras came about (GF1/EP-1/X100 etc), I considered the DSLR weight the price to pay for quality images and didn’t think any more about it. So, I was hoping the X100 would be the “out and about” replacement for my DSLR kit, but it just isn’t.
I use my DSLRs for family and friend portraits, quite a bit of macro, some weddings, and some sports/aircraft/general shots as well. Currently, if I just want to grab a camera and go out somewhere for general photos, it’s my 5Dmk2 and 35 f/1.4L that go. It’s not tiny but it’s not massive either, and it matches the focal length capabilities of the X100 hence my initial interest.
So, why not an X100? For me, it’s all down to the AF.
AF is the biggest issue for me. I’m not knocking the capabilities – I’ve personally written contrast detect AF routines for automated lens calibration processing software, and I know how much processing is involved. The X100 does a fine job of autofocus with the tools at hand, but there’s a reason that phase detect AF is still present in high-end cameras – it uses dedicated sensors and processors and runs several orders-of-magnitude faster than a similar contrast detect AF system. As an example, suppose an AF system can sample at 1,000 times per second and quickly react to a change of 5mm at the focal length of interest – your subject could be moving at over 10mph and the focus system will track it without issue. For contrast detect (like the X100) current top speed is around 10 times slower than that, so if there is significant movement in the subject between samples points the AF system needs to restart and you end up with hunting.
I did a bit of shooting for a client over the last couple of days using my 5Dmk2 and the 85mm f/1.2L II lens which has a slow focus, but I found that the combination of camera and lens just locked quickly without me thinking, and 95% of the shots I took were in focus (they were of fast moving children again – one on a wooden bike coming at me at some pace!) Yes, it’s bigger and bulkier, but I think I’m convinced that – at the moment, at least – that’s still a price worth paying.