Originally posted by nosnoop Human factor is the hardest variable to remove. Actually, just try taking shots at slow shutter speed without IS or SR, you will be surprised how variable the results would be. Different camera weight, hand grip, hand holding method, shutter release etc can all affect the result.
Yes, I understand that. When you can't eliminate a variable entirely in testing, you compensate by taking many tests and looking for the results that are most instructive. That's why I would make sure I took more than one shot at each setting with each camera. Two shots with each camera would be better than one; three shots would be better than two. I suspect that, with three shots, it would be possible to spot differences in result if they were significant.
Quote: Don't go too slow. SR can eliminate camera shake, but it cannot remove your own movement. Once the shutter speed is too slow, your own movement (fore & aft, side to side) would come into play.
Well, yes, but the thing you'd want to find out is, at what point, with each camera, would your own shaking become a problem for the photo. In tests I've done with the K100D itself - with SR on, and with SR off - I found that, with SR on, I had a reasonably good chance of geting a decent photo of a still subject with a shutter as slow as 1/4s. Without SR, I could not get close to that slow a shutter without blurring the photo.
Quote: Choose a longer focal length, distant object, then slow down the shuter speed to 2 stops, 2.5 stops, 3 stops.... below the 1/focal length*1.5 rule.
A variant of what I was thinking of, and a good idea, too.
The point, I should think, would be to spot a trend or an average or a tendency. I'd be looking for a result that could be expressed something like this: "49 of 70 images taken with in-the-lens IS were sharp, while only 40 of 70 images taken with in-camera SR were sharp". Or vice versa. Close results would be indecisive - except that the lack of a clear victor would be a victory for in-camera shake reduction, in my view.
My brother-in-law has a Canon. I've written him to ask if he has an image-stabilized lens. I think he may not. He's a landscape photographer and I know he uses the tripod all the time, so perhaps he's never bothered to pay for IS. I don't want to do something sneaky like buy a lens, do the tests, then return it for a refund.
Will