GUEST POST FEATURE
Write Your Articles Directly On FujiRumors!
There are a number of videos and manuals around which cover the Fujifilm autofocus system and settings in detail. The big problem isn’t the things they cover, it’s what they don’t cover. Guides often talk about how to configure and setup your camera, but with this guide, we’ll focus on the hidden things, how to use it to get the best outcomes.
I’m not going to try to cover the basic configurations here, for example, what AF-C or AF-S is. If you want to know how to use the basic settings, I’d recommend the Fujifilm autofocus site or this video from Fujifilm. Now please bare in mind, this is only my experience with the Fujifilm autofocus system, and more specifically my experience with the Fujifilm X-T2 and X-E3 systems. This isn’t an official Fujifilm guide, but in the absence of such a guide, I’m giving it a go.
To give you some context, I’ve come from a Nikon background, shooting mainly sports and charity events, along with family shots of my kids (5 and 7) who have enough energy that we’re secretly convinced they wake up in the morning and each snort a can of Red Bull. If you ask me which is harder to photograph, I’ll tell you it’s the children because at least a sportsman is predictable…not a child.
What is my claim to fame? Absolutely nothing. View my images on my website, make the call as to whether I know what I am talking about or not. If you think I don’t, stop reading and ignore me. All I can say from a personal perspective is I have reached a point where I don’t lose many images as a result of missed autofocus. Bad composition? Sure. Bad photos? Sure. Missed Autofocus? About 10% if I am lucky, normally a lot less.
In this guide, I’ll try to cover some of the things no one ever covers, my learnings coming from DSLRs and moving to Fujifilm. One of the reasons I wrote this, is I often hear that DSLR users say that they find their Fujifilm cameras vastly inferior to their DSLR, and I think part of the problem is that they think DSLR whilst using mirrorless and they expect it to work exactly the same. If you have ever used Mac and PC, you’ll know that you have similar frustrations switching from one to the other and a lot of it just comes down to how you use it…not better, not worse, just different.
From a personal perspective, I don’t think there is a big difference, I just think there is a lack of information that translates the use cases and as a result, people just don’t use them in the right way. DSLR is a little like an automatic on a formula one car, you can set it to drive mode, get high performance and forget about it, but it still has some limitations. Mirrorless on the other hand has two modes: learner driver which is the sort of automatic you find on a family car; and a manual mode (not referring to manual focus, but rather to the autofocus custom modes) suited to advanced users, you need to know what gear to have it in to get the right performance.
This guide is going to seem a little disorganised, in part because these are small random and in some cases unrelated items, all of which can have an impact on the autofocus system. Without wasting to much more of your time, lets get moving with the first item.
The fastest autofocus mode isn’t single point, use the smallest zone for the fast stuff
This is probably the most confusing part for DSLR users. Most DSLR’s work best with single point focus and many DSLR users assume that mirrorless works the same, after all, why wouldn’t it? The reality is you can’t treat them the same. Mirrorless still needs some contrast, even with phase detection , so if it can’t find some form of contrast to use as a point of reference, it battles. Single point doesn’t give it much area to look for contrast, so small single point will be worse than big single point. Whilst this method is the most accurate, it’s also the slowest.
Zone allows the camera to identify an area of contrast within the zone (because it’s a larger area) which it can do quickly, but there is a catch: Bigger zone doesn’t equal better or faster. It’s a little like 9 point vs 51 point. The more points, the more options the camera has to calculate and the longer it will take. In high speed land where milliseconds make a difference, that’s enough to miss focus. In conjunction to this, the larger the zone, the higher the risk that it focuses on the wrong thing.
The big zone vs small zone may be in contradiction to what a particular controversial you-tuber says, but it is correct. This is why it isn’t enough to simply know gear, you actually have to work with it in real situations to understand how to apply it.
Where to focus
I know this seems like something strange to bring up but this is something DSLR users often miss because they take it for granted. It is something simple that can often result in a big difference in the accuracy of autofocus. DSLR is point and shoot from an AF perspective, it’s easy, you point your DSLR at something and it focuses.
With mirrorless, it isn’t quite the same, particularly if you are using single point or the smallest zone. Mirrorless is a little more reliant on contrast so you have to think about where you focus. The easiest example to use is a plain smooth white wall. With a DSLR you can point at the centre of the wall and push the button, and it will probably focus, most of the time (although on some occasions even DSLR will battle). With a mirrorless, the centre of the wall may not have much contract if it’s a single colour, so you have to pick something on the wall where there is a little contrast, maybe a nail in the wall.
Some would argue this is a reason to go with DSLR but even DSLR suffers a similar fate, it’s just less exaggerated. If the wall was white enough, with very little contrast, DSLR will also battle. This is one of the reasons zone seems to acquire focus faster, because its not constrained to the selected point for focus, so if it can’t find contrast, it has a broader area to look for it. To put that into perspective, if you had your centre point on the centre of the wall, one of the points in the zone would probably pick up the nail in the wall, or something of high contrast on the wall and you get the focus.
Does this mean DSLR always has the upper hand? Yes and no. The challenge with DSLR is the strength can sometimes become a weakness. If it can’t find focus on the centre point, it hunts and hunts and hunts where Fujifilm’s zone will often find an alternative point that may or may not be enough to get you one picture. Yes, it may not be the right point, but it also may be correct because often it’s on the same focal plane.
What Focus mode to use in low light
I know this sounds strange, but the use of AF-S and AF-C actually makes a big difference to the autofocus in low light. In low light, if you use AF-C, it will often hunt for focus, because it’s not just trying to focus, but maintain focus. Getting focus once is easy, maintaining focus for any camera in low light is difficult, so sometimes the solution is simpler than most people realise…use AF-S.
I know this seems counterintuitive, but it’s one of these occasions where using AF-S and getting the shot off quickly works well. I know this isn’t an ideal solution, but it’s a solution that works if you’re not working with something moving fast that has paper thin DOF, and ultimately, until we have an autofocus system that is completely infallible, we have to find compromises that give us the best results.
Zone will also help considerably in low light as it gives the focus system more options to find focus. In short, if you want to have the highest chance of getting autofocus to work in low light and you aren’t playing with paper thin DOF or an object moving towards you at high speed where AF-S won’t work, use AF-S and Zone.
AF-C custom modes
Don’t think about the autofocus custom modes, think about what the individual settings are doing and think about how you want to apply them. For example, if I’m shooting kitesurfing in a waterhousing, I’m shooting someone coming towards me from into the sun, where they jump about 2-3m from me, jump over me, at the same time covering me in water spray, and this while I’m treading water holding a heavy camera in a housing whilst trying to keep the subject in frame.
Trust me when I say that keeping a single point on the kitesurfer is near impossible, but you also don’t want the autofocus focusing on the kite behind him which you want in the frame. Why am I telling you this? Because the custom mode you select has a big impact, in particular the option to give preference to what is in front or what is in the centre. If it’s giving preference to centre and the kite is in the centre, it will get the focus, if it’s giving preference what’s in front, the rider will be in focus. It’s a hit or miss difference in focus land.
What problems still exist that Fujifilm can address?
Zone isn’t perfect
There are a couple of problems with the current implementation if zone autofocus that I really don’t like. To understand what I mean I’m going to compare it to the Nikon 3D system and 9/21/51 point system in what I think Nikon actually gets right. With Nikon’s 3D system, if you were using the 3D system you put you put the middle autofocus point on top of something, it locks to that object under the focus point. I.e. If I am looking at a face and I put the middle of the 9 point on the eyeball and press down the focus button, it will lock focus on the eyeball, not the nose or any other object it thinks has better contrast within the 9 point area.
I’ve create an example below to show you what I mean: In this example, with centre as preference, I would expect it to always pick up the eyeball, but it doesn’t. It can often randomly select another point (which may have a higher contrast which I have highlighted in red) and this can be extremely frustrating making the zone unusable in many situations. If I use single point, it’s hard to keep the single point on the eyeball.
This is where the challenge lies with the current Fujifilm autofocus system. I’d like to see Fujifilm add an option to force the cross hair as the starting focus point for AF-C. If I put the centre on something, I want my Fujifilm to stay with that and not decide it has something it thinks is better. I believe this one item will provide valuable improvements to the autofocus.
Yes, it will have negative impacts on speed on occasion, but if it’s something you can switch on or off, I’d be happy. Some may argue that this is exactly what Face and eye detection is designed to do, but the reality is they are not quick enough to do that, or at least not at this point in time. It will improve and maybe one day it will be there, but until then, Fujifilm needs a better zone approach, even if it’s just an option you can select.
Automatic the way pro’s use automatic
What do I meant by this? Well, with DSLR, I could use 9 point autofocus for pretty much everything involving action and most of the pros I know work this way. I’d like to see an option for Fujifilm mirrorless that works the same, a 9 point or 21 point option that follows the guidelines mentioned in “Zone isn’t perfect” and the center point being the starting focus point. Whether it automatically switches between custom based on the way the object is moving, and whether it’s technically even possible is a different question.
I’m trying to establish ways it can work better, and a one size fits all option similar to DSLR will have two major benefits:
1. It will make an easy transition for DSLR users
2. It will simplify the autofocus workflow for those who don’t want the complexity but need an advanced autofocus. In all honesty, I think Fujifilm have made the current system a little overly complex for amateurs when it doesn’t need to be.
I don’t use faces often unless it’s only a single person, basically never if it’s more than one person but when I have tried it, one of the things that has really confused me is Fujifilm’s implementation of faces. It decides which face and you can’t change it! If you have 3 or 4 people and you want the person in the middle of the DOF selected, you can’t manually select it which can make it completely unusable at times.
What I would like to see is the ability to change faces with a front or rear dial, maybe push in a putton to scroll through the faces. There is a reason you can’t use it much currently and that’s the way Fujifilm implemented it. I think this would improvement the usability substantially.
Eye detection in AF-C
One of the annoyances. The moment you go to AF-C you lose eye detection which I actually quite like for close up portraits. We’re not taking photos of statues.
Yes, it can always be faster. No matter how good it gets, Fujifilm should be striving for faster. DSLR still has the edge in most situations so any improvements are welcome.
Low light autofocus
When you compare mirrorless to the current crop of DSLR cameras, it’s clear that mirrorless cameras aren’t’t up to scratch. Now, I’m going to be honest and say, I don’t understand the technical reason why that is, but mirrorless is getting better and so is DSLR. All I’m saying here is that it has to continue to get better and mirrorless manufacturers have to continue to find a way to make that work. Whether it’s some obscure technology like they use in night vision goggles, I don’t know, and don’t really care, but we’d all like to see autofocus work in ridiculously low light. I suspect this will improve in time so I won’t mention much more than that.
Where does mirrorless beat DSLR?
People will tell you that DSLR kills mirrorless and for some situations like low light, it does. But, there are also situations where mirrorless rules.
Front vs Centre Focus
I used the kitesurfing example previous about front vs center and that’s exactly what of the situations where mirrorless works well. The problem is whenever there is a scenario where the person goes out of frame, it’s how they acquire them again. If the kite is on the focus point, the kite is in focus and the rider out of focus…bad outcome and it’s a regular occurrence.
So how does mirrorless get this right? Zone with front preference. Zone with front preference gives preference to the object closest to you, not the object under the centre focus point. So from an autofocus perspective, it will look at all the points and find the one where the object is closest to you, and that’s what you want.
In reality, DSLR will find the item on the middle point when it reacquires focus and that’s the wrong item. In many cases, the centre point might actually be on sky because the rider is on one side and the kite on the other, and I can’t setup the autofocus for which side that will be because it changes on each jump.
DSLR autofocus often requires micro adjustments of lenses
DSLR relies on the lens and camera being within a specific tolerance, but sometimes you can have a situation where a lens and camera combined are outside tolerances and need to be micro adjusted to resolve a front focusing or back focusing error. This can be a time consuming task, often requiring adjustments with each lens, and in the case of zooms, at different focal lengths. With mirrorless the adjustment occurs on the sensor itself so this is no longer required.
Rear Display autofocus isn’t as good
DSLR autofocus is great if you are looking through the viewfinder, but if you have to use the rear screen, the autofocus generally becomes a lot slower. When I say slow, I mean in many cases, unusable for moving objects. Whilst this may not seem like a big issue, it is for kids (shooting at eye level) and sports where you need to keep the camera low to the ground (skateboarding as an example).