Image Stabilization (IS) is a great benefit for low-light shooting. If it's in the lens it uses optical elements to minimize the effects of camera movement. If it's in the camera body it moves the sensor to achieve the same effect. It has no effect on subject motion though, so if you're shooting a moving subject at low shutter speeds, the subject will be blurred, IS or no IS.
The catch is that a photograph shot with IS will sometimes show motion blur even if the subject is stationary. The camera is most likely to be handheld (camera makers recommend against using IS if the camera is mounted on a tripod), so it's possible that the cause is a shaky camera. It's also possible that the IS mechanism itself is in motion even if the camera and subject aren't. I suspect this is because it takes a moment or two for the IS to stabilize the image. This effect is visible if IS is in the lens. If it's in the body, you can see it only if your camera has Live View.
Notice how the blur is to the right of the pole and parallel to it? Human-induced shake is seldom this symmetrical, hence my suspicion that IS is the cause. What do you think?
Yet another possibility is that I don't know what I'm talking about and that every now and then I get a jumpy shutter finger. Given that I've been at this for 40 years or so, I doubt it. So my question today is, has anyone else noticed this phenomenon? If so, has it made you any less inclined to use IS? I'm on the fence. IS has delivered enough sharp photos in low light that I fully appreciate its value, yet there have been enough situations where it's done more harm than good that I don't entirely trust it. The bottom line is that I don't plan to sell my tripod any time soon.
Gordon,
I would definitely concur with your judgment here. I shot with the K-7 and had a really tough time with blurring at modest shutter speeds with the IS enabled. I would turn the IS off in some instances and find that my own movement would at least be predictable and result in acceptable photos. I remember Ctein wrote a very thoughtful (as usual) article on OP about this topic. It's definitely worth a read to see how it meshes with your results: Link to article on The Online Photographer.
Good article, I'll be interested to see if you can determine the nature of the motion blur.
Posted by: Rich | June 25, 2010 at 11:30 AM
Gordon, I too have had strange results with the IS. I had a Pentax K-x which seemed to give me shakier results with IS than without. It was unfortunate and I had to return it. I have an Olympus E-P1 which consistently gives better pictures with IS enabled. I truly value this feature but I think there are different flavors of IS and you need test them before we settle on one.
Posted by: Jaladhi | June 25, 2010 at 01:14 PM
Fascinating. I had never considered that. But then, I've never seen an automatic system that didn't have its downside. Hmmm, maybe that's part of why I'm going back to film.
Posted by: Wes Medlin | June 25, 2010 at 03:15 PM
I think part (although probably not all) of the reason the K-7 may have this problem is that the autofocus of the camera locks faster than the antishake mechanism does. In my case, I forget to wait for the antishake icon to come on all the time when I take shots. Perhaps what you're seeing is shake introduced by the IS system in mid correction.
What's helped me some is to decouple the AF from the shutter button. I then prime the shake reduction before I focus by compressing the shutter half way, then focus with the AF button on the back.
I've also read that with shake reduction systems it sometimes helps to take photos in bursts of threes. When I've tried this, the second or third shot is usually sharper than the first. I don't know the exact technical reason for why this occurs, but it seems to work. Whenever I'm trying to get a shot at slower than 1/60 I usually can get more keepers with the three shot method.
Posted by: Stephen | June 25, 2010 at 10:08 PM
Here's a long essay by Thom Hogan on the topic: All About VR: Nikon's VR system explained. Although nominally about Nikon lenses, on the whole it's mostly brand-independent.
The boldfaced beginning — “The first and most important rule of VR is this: never turn VR on unless it's actually needed” — initially give me the impression that it was going to be a rant against VR, but it's actually not, and is worth reading.
Posted by: Gary Brown | June 26, 2010 at 08:54 AM
I've noticed this myself with my K7, but while there appears to be anecdotal verification I would hasten caution. Even if we are correct, there is no guarantee that the cause is the same.
Having said that though, I really wish they kept the switch SR switch from the K10D. I used to make good use of it. Now I rarely remember to turn it off. I think I will make the default off, and turn it on when needed as Thom Hogan suggests.
Posted by: Nikhil Ramkarran | June 26, 2010 at 06:33 PM
Nikhil,
I agree that the evidence is anecdoctal and that we shouldn't jumpt to conclusions--but there's enough evidence to suggest that one should use IS only when the risk of not using it is greater than the risk of leaving it on. Exactly when this is will be different for each photographer and will vary with the camera and lens. Personally, I've decided to use it only when the shutter speed is slower than I can reliable shoot without IS. Given the high ambient light level in the example photo, I could have easily shot it at 1/125 second or faster. If I'm ever faced with a similar situation again, trust me, I will.
Posted by: Gordon Lewis | June 26, 2010 at 09:50 PM
Falk Lumo has a very carefully done white paper on the very question of when the K7's IS works, and when it does not.
Posted by: Semilog | June 27, 2010 at 03:19 AM