Close search results
Close search results

Canon Image Stabilizer (IS) test

Shaken, not blurred

April 5th, 2010

Many lenses today come with optical image stabilizers (in Canon's case, IS for short) to help producing sharp pictures when shooting in low-light conditions. Useful when a tripod is not an option. However, lenses with IS are typically more expensive than their non-IS counterparts. So how much does IS help, and is it worth the extra $$$?

Setup

All pictures were shot using a Canon 1000D with the kit lens (Canon 18-55 mm EF-S 3.5-5.6 IS). ISO was set to 100 and camera was set to time value (shutter priority). No special techniques were used (like sniper breathing), just holding the camera normally and looking through the optical viewfinder.

I started with the shutter set to 1/60, and then went through 1/30, 1/15, 1/10, 1/5 and 1/2, shooting 20 pictures using each setting; 10 with IS activated and 10 without IS.

Below is a sample shot, the crops at all settings follow.

A view of Frankfurt.
A view of Frankfurt.

Crops

For each setting, I selected one picture that I felt was the most representative one; the one with an average amount of blur. These are shown below.

1/60, with IS.
1/60, with IS.
1/60, without IS.
1/60, without IS.
1/30, with IS.
1/30, with IS.
1/30, without IS.
1/30, without IS.
1/15, with IS. Still sharp.
1/15, with IS. Still sharp.
1/15, without IS. Looks a bit shaky.
1/15, without IS. Looks a bit shaky.
1/10, with IS. Still nice and sharp.
1/10, with IS. Still nice and sharp.
1/10, without IS. This image is useless.
1/10, without IS. This image is useless.
1/5, with IS. A bit worse, but still looking okay.
1/5, with IS. A bit worse, but still looking okay.
1/5, without IS.
1/5, without IS.
1/2 second, with IS. Definitely not so sharp anymore, but still surprisingly good.
1/2 second, with IS. Definitely not so sharp anymore, but still surprisingly good.
1/2, without IS. I might just as well give up.
1/2, without IS. I might just as well give up.

Useful pictures

I evaluated each picture and put together a table showing how many of them I subjectively found "good enough" - that is, a picture that is not so badly blurred that I would normally just delete it.

ShutterWith ISWithout IS
1/60100%100%
1/30100%90%
1/15100%70%
1/10100%10%
1/590%0%
1/260%0%

Even at very slow shutter speeds (half a second), most of the pictures came out acceptable when IS was turned on. Without IS, the limit seems to be around 1/15th, after which most pictures become useless.

Conclusion

First of all: Optical image stabilization works. I had my doubts, and I was surprised by my own findings. Secondly, is it worth paying extra for? I did a quick check online, and I found that the non-IS version of my lens is only about 20 EUR cheaper than the IS one. In that case, it's a no-brainer, just get IS.

If you prefer to use a tripod anyway, no need to invest in IS. Apparently (I haven't tested it myself), when using a tripod, using IS may lead to worse results than not using it.

Leave a comment






This will just take a second.

Submitting your comment...
Erik Moberg  2019