In-Camera Image Stabilization Can’t Trump Lens Technology
Newer digital SLRs like this A100 from Sony have included an image stabilization feature inside the camera itself. Why aren’t offerings from Nikon and Canon following suit?
With camera manufacturers releasing their next generations of digital SLRs, one feature that’s gaining a lot of attention is in-camera image stabilization. While it’s one of the big selling points in Sony Digital SLRs like the A100, both Canon and Nikon continue to eschew in-camera image stabilization in their own new offerings. Are both Canon and Nikon simply behind the times? Perhaps surprisingly, the answer is no.
Here’s why. In-camera image stabilization works by shifting the image sensor to compensate for camera movement. This is effective at reducing camera shake, allowing the photographer to operate at lower shutter speeds than he would otherwise be able. There’s one big advantage to putting the image stabilization inside the camera - every lens becomes image stabilized to a degree. This can be a huge savings, as image-stabilized versions of lenses by both Canon and Nikon tend to cost hundreds of dollars more over their non-stabilized counterparts.
But there’s one big problem - in-camera image stabilization simply isn’t as effective. Part of the problem with an in-camera solution is that the vibration reduction becomes less effective in the telephoto range. This is exactly when image stabilization is needed the most, because camera shake is more and more pronounced at larger focal lengths. Vibration reduction elements inside the lens itself don’t suffer from this shortcoming.
Should you avoid purchasing a camera with native image stabilization? No - it’s still relatively effective at reducing camera shake and unwanted vibrations. But Canon and Nikon aren’t stupid when they opted not to include this feature in their latest models. Both camera manufacturers recognize that vibration reduction in the lens itself is a more effective - albeit a bit costlier - solution.
Like what you see? Join our newsletter.
More from this category
- Face Detection on Digital Cameras: Empty Marketing or Must-Have Feature?
- Canon 40D Focusing Problems
- The Benefits of a Digital Camera Card Reader
- 4 of the Strangest (and Coolest) Camera Lenses You’ll See
More from this author
- Face Detection on Digital Cameras: Empty Marketing or Must-Have Feature?
- Canon Shooters: Good Low-Light 50mm Lens for Cheap with Coupon
- Resizing and Resampling for the Web - How to Best Show Your Work Off Online
- Tips and Business Insight from a Sports Illustrated Photographer
Subscribe to this author
Canon has designed a standard lens element group which will allow them to add image stabilization to many more lenses without adding much to the cost.
Example EF-S 55-250mm IS lens from canon.
True - that’s a good point. Image stabilization costs are coming down. I’m curious, though, whether that same group will be used in the next generation of L lenses.
The 55-250 isn’t available in the US, at least yet. For more info:
Click here.
I agree with the article completely. I recently attended a seminar sponsered by a photo magazine and Sony, and was given the opportunity to compare the A100 to my Canon EOS 20D. The difference in the image stabilization capability is visible even at moderately higher focal lengths.
I respectfully disagree with your central point — that in-camera image stabilization is less effective than in-lens image stabilization.
Canon published a study that seemed to indicate it might be (not surprising, since there was the perception you noted that Canon and Nikon were not keeping up). But that was far from conclusive, and has been argued against from those who understand the technology better than I do. So I’m not sure how you are making this conclusion — have you done a fair comparison test?
The bigger point is this: if in-body image stabilization allows a photographer 2 or 3 additional shake-free stops, but in-lens gives 4 or 5, then we are really getting into a range that is effectively overkill. If antishake or image stabilization lets me handhold at 1/15th of a second for a shot as sharp as would have required 1/125th of a second previously, that’s great — but it also means that people moving in the shot will be blurred. Image stabilization does not freeze action — only higher shutter speeds do.
So why would I want to shoot with a telephoto and depend on antishake? If I’m shooting action, image stabilization doesn’t freeze motion anyway — it just counteracts the normal shake that happens when one pushes the shutter.
The in-body systems used in Sony, Pentax and other brands are very helpful for photographers when used appropriately. Why make the implication that they aren’t valuable, and why make the further claim that they don’t measure up to in-lens systems without showing any proof?
I would be interested in seeing proof that in-body systems in the new generation of cameras — for example the camera Sony will announce this week — are not as effective as in-lens systems. As well, I’m not sure I see any evidence why this would be the case at longer focal lengths — could you provide some evidence for that?
Please note: I’m a fan of the site, so please take this in the positive manner it is intended — I think the real reason Canon and Nikon don’t have in-body image stabilization is that the VR and IS lenses sell very well, and that the systems are moving to full-frame sensors, which will require improved anti-shake systems to match the lens-based systems.
Ted
Hey Ted,
Great comment. The basis of the little blurb I wrote here is the Canon study you cite in your comment. The study read, in part:
I wouldn’t argue that in-camera vibration reduction isn’t useful - as I tried to point out, it serves its purpose and is a great way of adding image stabilization - at least to some degree - to any lens. But I don’t think it can be viewed as a substitute for an in-lens solution.
If you have the time, I’d love to see some comments from those rebutting Canon’s findings. Obviously, Canon’s going to want to show that their IS technology is superior - this isn’t a non-partisan study. But the central premise makes sense logically to me - sensor movements in an in-camera vibration reduction system would have to be quite pronounced at a higher telephoto range.
The Online Photographer cites the same Canon article I do here and argues that since a 300mm IS lens is out of the range for most photographers, the conclusion of the Canon study is moot. It’s an interesting point. He also highlights the fact that IS is often neglected on shorter lenses because camera manufacturers see less of a need for it. While I think this attitude is changing, particularly with Canon’s announcement of the dirt-cheap 18-55mm IS version, camera-based vibration reduction has its benefits here.
Great points, Dan. I think the really amazing thing, in the bigger picture, is that this technology is available at all. It’s just not something I would have imagined when I got my first Pentax K1000….
It will be interesting to see if the new Sony pushes this even further — I believe Sony is claiming an improvement in SSS, and sometimes in the next month I expect will be seeing reviews….
I have never used Nikon or Canon, but have used Pentax K10D and K100d with in camera stabilisation. I have bad arthritis and have difficulty holding a camera steady at any shutter speed. The Pentax cameras work extremely well with a 300 mm lens and with a 28 mm shift lens designed for 35 mm cameras. I definitely prefer to use it with older design lenses, rather than pay more for fancier cameras, results of which do not provide any better examples of image stabilisation by fellow camera club members.