Fujifilm Optical Image Stabilisation (OIS)
One of the most important developments in photography in modern times is Optical Image Stabilisation (OIS). Different camera manufacturers call this technology different things - Nikon calls this Vibration Reduction (VR) and Canon calls this Image Stabilisation (IS). The way all of these systems work is very similar.
Their aim is to reduce camera vibration from hand-holding and generally, allow us to use slower shutter speeds than ever before. This means that we are able to use lower ISOs for low-light shots of stationary objects. This is important - there's no point shooting action with OIS because it only freezes camera shake, it does not freeze motion. For hand-held landscape, social and street shooters, this technology is particularly important.
Their aim is to reduce camera vibration from hand-holding and generally, allow us to use slower shutter speeds than ever before. This means that we are able to use lower ISOs for low-light shots of stationary objects. This is important - there's no point shooting action with OIS because it only freezes camera shake, it does not freeze motion. For hand-held landscape, social and street shooters, this technology is particularly important.
Probability Rehash
Whether you get a sharp shot or not is dependent on probability - at every focal length and shutter speed there is a certain chance of you getting a sharp shot and a certain chance of you not getting a sharp shot. This probability depends on many factors including your hand-holding skill and even your balance.
For example, if you're shooting at 18mm, you'd expect that most shots, e.g. 99%, would be sharp at 1/60s. If you go down to 1/30s, this number might decrease to around 85%. If you go down to 1/15s, then you might only be getting 50% shots sharp. These numbers are completely made up, but they illustrate the point. The higher the shutter speed, the higher the chance of you getting a sharp shot. The only way to get an accurate estimate of these probabilities is to conduct a test - for example, to shoot 10 shots at each shutter speed and count how many are sharp.
OIS does not guarantee that you will get a sharp shot, it simply increases the probability of you getting a sharp shot. For example, with OIS, the chance of you getting a sharp shot at 1/15s might now increase to 80%.
For example, if you're shooting at 18mm, you'd expect that most shots, e.g. 99%, would be sharp at 1/60s. If you go down to 1/30s, this number might decrease to around 85%. If you go down to 1/15s, then you might only be getting 50% shots sharp. These numbers are completely made up, but they illustrate the point. The higher the shutter speed, the higher the chance of you getting a sharp shot. The only way to get an accurate estimate of these probabilities is to conduct a test - for example, to shoot 10 shots at each shutter speed and count how many are sharp.
OIS does not guarantee that you will get a sharp shot, it simply increases the probability of you getting a sharp shot. For example, with OIS, the chance of you getting a sharp shot at 1/15s might now increase to 80%.
The Test
This test aims to determine those probabilities and see how much OIS actually helps. I will be taking a shot at 18mm, at each of the following shutter speeds - 1/15s, 1/2s, 1s. My aim is to see if I can possibly perform a sharp shot at each of these shutter speeds. Just for reference, this is the scene I will be photographing and this is the sharpness that will be required (the following shot was taken at 1/15s). I will then repeat the test, except this time with OIS enabled.
As expected, the shot at 1/15s was sharp. I continued to take 10 more shots at 1/15s, which were all sharp. At 1/15s, we can conclude that without OIS, we can expect at least most of our shots to be sharp. That said, I do have reasonable practice with hand-holding, but 1/15s isn't particularly long for an 18mm focal length, so these results should be very repeatable. We can see that there is some noise due to high ISO. Let's now try 1/4s.
At 1/4s, the shooting becomes more difficult. I had to take several shots to get one sharp one and this was the sharpest image I got. Even though this looks acceptably sharp to the eye, cropping in at 100%, we can see that it is not sharp, with some blurring of finer details. If we count "sharp to the eye" as sharp, then I got 7/10 images sharp at 1/4s. If we count "100% crop sharp" as sharp, then I got 0/10 images sharp at 1/4s. Even though this sharpest 1/4s image is less sharp than the 1/15s image before at 100% crop, we can see that at normal sizes, it actually looks better due to having less noise. Let's try 1s.
I took 10 shots at 1s too, this was the best I could get - clearly unsharp even at this size. I couldn't get any sharp, so it's a 0 for shots at 1s without OIS. We'll now try 1/15s with OIS. As expected, I got all the shots sharp and all of them are sharp at 100% crops too. I'm going to try 1/4s now. The following image is typical of what I got - since most were good, I couldn't pick a "best" per se.
This wasn't really much of a challenge - I expected most of them to be sharp. I'll include a 100% crop here too. They were almost all sharp at 100% crops as well - so this is already showing how effective OIS can be. Let's see what happens at 1s. This will be interesting.